WEBVTT
00:00:03.279 --> 00:00:10.880
You're listening to Built to Convert, the podcast where growth-minded marketers, founders, and creators come to level up.
00:00:11.119 --> 00:00:15.439
No hype, no guesswork, just bold insights that drive results.
00:00:15.679 --> 00:00:16.640
Let's go.
00:00:20.640 --> 00:00:27.600
Hey guys, welcome to Built to Convert, the podcast where we explore how strategy, analytics, and creativity come together to drive meaningful growth.
00:00:27.760 --> 00:00:30.399
I'm your host, Tyler Lane, founder of Session Interactive.
00:00:30.480 --> 00:00:36.799
And each episode we sit down with marketers, founders, and operators who are building real momentum inside today's most interesting companies.
00:00:37.039 --> 00:00:42.159
This episode is brought to you by Session Interactive, the performance marketing agency for brands that are serious about growth.
00:00:42.399 --> 00:00:55.759
Whether you're in B2B, SaaS, DTC, e-commerce, or anywhere in between, we help marketing teams cut through the noise of smarter SEO, GEO, AEO, all those acronyms, paid media, AI integration, and analytics that actually drive revenue.
00:00:55.920 --> 00:00:57.920
Learn more at Session Interactive.com.
00:00:58.399 --> 00:00:59.119
All right, guys.
00:00:59.280 --> 00:01:00.479
It's a fun one today.
00:01:00.560 --> 00:01:03.119
We've had a longstanding relationship with these guys.
00:01:03.200 --> 00:01:06.560
Uh we've got Steven No, director of growth marketing at Crazy Egg.
00:01:06.719 --> 00:01:19.040
Steven has spent his career on the go-to-market side of B2B software with stops at NIRA, PADL, and Profit Well before landing at a company that builds the exact tools marketers use to run experiments.
00:01:19.200 --> 00:01:20.239
He's talking about.
00:01:20.480 --> 00:01:22.319
It's a strange vantage point.
00:01:22.480 --> 00:01:31.840
He's not only doing CRO, he's selling infrastructure for it, which means he sees where testing culture actually breaks uh down inside real teams.
00:01:32.159 --> 00:01:33.120
Really exciting stuff.
00:01:33.280 --> 00:01:36.239
Steven's always been super progressive and always very generous with us.
00:01:36.480 --> 00:01:38.079
Stephen, I really appreciate your time, man.
00:01:38.319 --> 00:01:38.799
Absolutely.
00:01:39.040 --> 00:01:40.159
Yeah, thanks for joining us.
00:01:40.319 --> 00:01:43.840
Hey, we like to start out with these conversations with a little bit of kind of career origin.
00:01:43.920 --> 00:01:46.799
I know you just called out some of those companies that you were in.
00:01:46.879 --> 00:01:49.920
And hey, shout out to Heaton over at Crazy Egg2.
00:01:50.159 --> 00:01:54.400
He was, he was really um, you know, had some amazing things to say about you.
00:01:54.560 --> 00:02:02.079
Uh actually when Heaton and I first met a couple of years ago, and he was like, I think his quote honestly was like, Steven's the best marketer I've ever been around.
00:02:02.239 --> 00:02:02.959
So it was pretty cool.
00:02:03.040 --> 00:02:05.040
And that's it's high praise from Heaton, too.
00:02:05.280 --> 00:02:07.680
He's got uh he's got the notoriety in this space.
00:02:07.840 --> 00:02:18.080
But talk to me about kind of your experience, you know, as you you've matured in your career now, obviously director of growth marketing and such a reputable uh could CRO-related company.
00:02:18.159 --> 00:02:29.360
But talk to me about kind of how it shaped you and and certain things perhaps you've had to unlearn as you've you know, and then now as you've grown and developed into this uh growth marketing director at CrazyAgg.
00:02:29.919 --> 00:02:30.159
Yeah.
00:02:30.240 --> 00:02:34.240
So like a lot of people, I didn't start out wanting to be a marketer.
00:02:34.319 --> 00:02:36.080
I came through this through roundabout blends.
00:02:36.240 --> 00:02:38.639
In my particular case, I actually started out as a data analyst.
00:02:38.800 --> 00:02:42.159
So when I was in college, I had like data analyst um stints.
00:02:42.240 --> 00:02:45.520
I was working on the pricing algorithm at Wayfair, funnily enough.
00:02:45.680 --> 00:02:51.280
Um I was uh spent a little bit in finance, but I ended up coming back to the startup space, which I really loved.
00:02:51.439 --> 00:02:54.879
Um, you know, started out kind of doing a lot of revenue operations type stuff.
00:02:55.039 --> 00:03:00.560
I was the sales source admin, I was running all the numbers, I was like scraping all the leads for the sales team, you know.
00:03:00.719 --> 00:03:06.560
Um, and kind of when you're in startup, startup mode, just there's way too much work and not enough people.
00:03:06.800 --> 00:03:13.840
So kind of one day someone came up to me and said, Hey, you've been kind of like looking at the Google Analytics account for little nuts of information.
00:03:14.000 --> 00:03:15.840
Our paid media manager just left.
00:03:16.000 --> 00:03:17.599
You see much to do you want to take that?
00:03:17.759 --> 00:03:18.960
I said, Yeah, sure.
00:03:19.039 --> 00:03:23.039
I'm probably gonna make some screw ups along the way, but you know, always an opportunity to learn.
00:03:23.199 --> 00:03:24.800
And it kind of snowballed from there.
00:03:25.199 --> 00:03:30.319
Went over, ended up taking on paper media, um, expanding to kind of wholesale demand generation.
00:03:30.400 --> 00:03:40.560
When I came over to ProvalWell, I ended up taking over a lot of a business development team as well, as well as a lot uh doing a lot of like product marketing, um, conversion and optimization, copywriting, things like that.
00:03:40.800 --> 00:03:46.479
And so kind of I've always been very much a jack of all trades, but the thing I think I'm really good at is kind of two things.
00:03:46.639 --> 00:03:54.960
Number one is pattern recognition, especially when it comes to looking at data and kind of figure out like, okay, what can we learn from a given signal such that we can put that into action?
00:03:55.120 --> 00:04:00.479
And the second thing I'm probably really good at ultimately is connecting dots between different domains.
00:04:00.719 --> 00:04:12.240
I can kind of like both take that quantitative skill set and look at the numbers and figure out what this means, but also work on cop writing, kind of relate back to, okay, what does this mean in terms of which messages, which offers do really well?
00:04:12.400 --> 00:04:15.039
And ultimately, what is it about our customers really want?
00:04:15.360 --> 00:04:33.680
Um, and so coming over to Crazy Egg, it's kind of been a perfect opportunity just because I can kind of not only do that for Crazy Egg while we're dealing with kind of all the crazy change we're seeing in the world right now, but on top of that, I can provide some of that consultative skill set for our customers as well, which means I can be helpful to as many people as possible.
00:04:34.000 --> 00:04:34.879
I love that, Steven.
00:04:35.120 --> 00:04:36.639
Again, your your perspective's been awesome.
00:04:36.720 --> 00:04:42.800
And I love it, just the ability to kind of sink your teeth into all those different things and get that level of exposure.
00:04:42.959 --> 00:04:56.959
It reminds me of my career, not to that extent, but just like you were able to just be in the weeds and therefore shape and get your give give your perspective or get a perspective that's far more uh robust, if you will, than a lot of others.
00:04:57.439 --> 00:05:06.319
Walk me through a specific moment when you realized the growth playbook that worked at maybe Profit Well, for example, wouldn't necessarily work at Crazy Egg.
00:05:06.480 --> 00:05:10.079
Was there certain things you're like, oh, I'm gonna transition this over into Crazy Egg?
00:05:10.319 --> 00:05:18.240
And you quickly realize perhaps you're like, no, I need to maybe pivot separately of more of the subscription-based um component that Crazy Egg offers.
00:05:18.480 --> 00:05:20.959
You know what's really funny is um, you know, Prof.
00:05:21.199 --> 00:05:23.040
Well was a multi-product company.
00:05:23.120 --> 00:05:27.120
So kind of some parts of a company were already firing all cylinders by the time I got there.
00:05:27.279 --> 00:05:34.000
Um the Prop Well metrics analytics products is definitely one example of kind of like just a great freemium analytics product.
00:05:34.160 --> 00:05:47.439
Uh, we had our price intelligence elite um kind of service and offering there as well, which, you know, um Patrick Campbell, our old CEO, um, you know, very well known in the business uh of SaaS, at least as far as like modernization strategy, and that was already firing.
00:05:47.600 --> 00:05:56.240
Um, the kind of one win which I could potentially own is our Proper Retain product, which was essentially an automated retention marketing product.
00:05:56.480 --> 00:06:13.920
And what was great is kind of before we ended up being acquired, we ended up building a really replicable outbound business development strategy, um, where we basically came in and we said, hey, like we have this great automated retention marketing uh platform, and we believe in it so much that here's what we're gonna do.
00:06:14.160 --> 00:06:20.959
If, you know, um if you uh install this product, we're gonna run an A-B test at no charge.
00:06:21.199 --> 00:06:28.079
If we can measurably beat the control um group on this trial, then we're gonna um measure that uplift.
00:06:28.160 --> 00:06:31.519
We're gonna take a little bit of that for ourselves, but you get to keep most of it.
00:06:31.759 --> 00:06:34.879
And uh and if we don't beat it, it's no skin off your back, no risk.
00:06:34.959 --> 00:06:35.759
What do you say?
00:06:36.000 --> 00:06:40.399
And as it turns out, lots of people loved it, even big companies.
00:06:40.639 --> 00:06:48.319
Um, and so what I kind of realized about like, hey, this is a pretty unsexy kind of monetization lever, which is retention.
00:06:48.480 --> 00:06:57.360
But turns out when you have an offer that's really, really good and it comes down to brass tags pretty quickly, then we were able to replicate that and scale that really, really quickly.
00:06:57.519 --> 00:07:02.639
Um, you compare that to Crazy Egg right now, which is a bit more of a mature category, right?
00:07:02.800 --> 00:07:06.240
Website analytics, you know, Crazy Egg's been around since I think 2004.
00:07:06.480 --> 00:07:10.560
Heaton, you know, um and his co-founder like invented category way back in the day.
00:07:10.639 --> 00:07:12.480
So, in a sense, we're doing a refounding.
00:07:12.639 --> 00:07:22.560
And we're doing that in the context of just the entire AI shift, the entire jobs to be done of website analytics of experimentation is changing completely right now.
00:07:22.720 --> 00:07:29.439
So, frankly, I'm actually doing a lot more product stuff than I used to recently in terms of, hey, this is a marketing analytics product.
00:07:29.600 --> 00:07:34.879
This is something that I would have considered super cutting edge back in 2004.
00:07:35.120 --> 00:07:42.240
What we want in when it came on in 2024 is okay, what is cutting edge now looks totally different.
00:07:42.480 --> 00:07:46.720
So if I was using this as a marketer, what would I consider cutting edge?
00:07:46.800 --> 00:07:55.920
What's gonna actually do the job for me such that I could go to other markers and tell them with conviction, hey, this is really great for me, and I know what's gonna be really great for you.
00:07:56.879 --> 00:07:57.680
Love that, Steven.
00:07:57.839 --> 00:08:01.519
Yeah, amazing stuff to just and perspective again, you're you've been learning.
00:08:02.000 --> 00:08:07.920
Let's transition here into CRO specifically and analytics and kind of the experimentation discipline, if you will.
00:08:08.319 --> 00:08:17.199
Um you've written about the four-step approach to the first 90 days of a CRO engagement that starts with tools teams already have, right?
00:08:17.279 --> 00:08:21.279
And not necessarily requiring net new purchases for particular tools.
00:08:21.360 --> 00:08:30.000
What are some of the most common uh things a team discovers in the f that first pass that they have completely blind to before, perhaps?
00:08:30.240 --> 00:08:30.720
I'm curious.
00:08:31.040 --> 00:08:31.199
Right.
00:08:31.439 --> 00:08:41.600
I think you know, the funny part is, you know, data people like me often we get distracted by data because we love numbers, we feel like it's super robust, we think that, oh, this is a cutting edge thing.
00:08:41.759 --> 00:08:44.559
And so it becomes down to like, okay, what do these numbers say?
00:08:44.639 --> 00:08:46.799
Um, what is the pattern, all this stuff?
00:08:47.120 --> 00:08:53.200
But if we're talking as marketers, and ultimately like marketing comes down to customers, right?
00:08:53.360 --> 00:08:54.639
What is the customer experience?
00:08:54.720 --> 00:08:57.120
What do they want from us and what gets them actually by?
00:08:57.440 --> 00:09:08.559
And what's funny is that, you know, I could spin one analysis saying like the average, um, or maybe differently, the most popular element on this page is this button.
00:09:08.799 --> 00:09:15.519
But another way you could spin it is to say this button represents the most popular offer or product on our website.
00:09:16.159 --> 00:09:17.440
Completely shifting it around.
00:09:17.519 --> 00:09:35.279
I'm taking and taking metrics and a lot of this statistical jargon out of it because ultimately statistical jargon can be helpful for people who are more specialized to work with, but if you're removing the kind of human perspective of a customer out of a lot of this, then a lot of your analysis is going to be focusing on the wrong things altogether.
00:09:35.519 --> 00:09:40.960
So I feel like a lot of uh companies, when they're coming in, they already have data to work with.
00:09:41.200 --> 00:09:46.399
But what they need to be coming in through is not just someone who has those like data analytics chops.
00:09:46.559 --> 00:09:59.360
They can also relate that to the ICP, what the target customer actually wants, and then ask basically, okay, what would that, what does this reflect about who our customer is, what they actually want, and how can we optimize based off of that as well?
00:09:59.600 --> 00:09:59.840
Yeah.
00:10:00.000 --> 00:10:01.279
So that's number one.
00:10:01.360 --> 00:10:07.200
And I think ultimately, like we also get too distracted by the kind of brass tacks work, right?
00:10:07.360 --> 00:10:16.159
It's, you know, like I spent so many hours of my life kind of like instrumenting, setting up our like, you know, Salesforce to HubSpot connector and stuff like that.
00:10:16.399 --> 00:10:20.480
But then, you know, when I ultimately bring this back to like, hey, what does the customer want?
00:10:20.639 --> 00:10:26.159
Turns out a lot of your metrics, you should collect the data anyway, because you may need it in the future.
00:10:26.240 --> 00:10:31.200
But ultimately it comes down to like, hey, um instrumenting is not the most important thing.
00:10:31.360 --> 00:10:35.840
The most important thing is number one, figure out which part of a page people are actually scrolling down to.
00:10:36.000 --> 00:10:38.799
And as it turns out, a lot of the time, people barely scroll down.
00:10:38.960 --> 00:10:42.720
So you got to get the right stuff at the top of a page, get the right offer in there.
00:10:42.879 --> 00:10:52.559
And I really figured that it's like, hey, get the right offer, because you know, the funny part is every time you go to say, you know, I've been to a bunch of like HopSpot inbound conferences in my life.
00:10:52.720 --> 00:10:59.360
And the example they would always give is, hey, we changed this button color, and that increased conversions by X percent.
00:11:00.080 --> 00:11:04.559
No one's ever seen a big increase in conversions from changing a button, right?
00:11:04.879 --> 00:11:10.159
A button placement can make sense because you just want to make that convenient and make that obvious.
00:11:10.399 --> 00:11:14.480
But changing a color doesn't make sense, but changing your offer absolutely does.
00:11:14.639 --> 00:11:23.519
So it's not just focusing on your customer, but also thinking about like, okay, instead of thinking of it as like optimizing every little lever, think about which levers make the most sense.
00:11:23.679 --> 00:11:33.679
And ultimately, a lot of the time this boils down to like two to five levers, a lot of the time, in terms of actually getting your customers to kind of see your offer and decide whether or not it's the right one for them.
00:11:35.600 --> 00:11:36.799
A lot to unpack there.
00:11:36.879 --> 00:11:48.879
I'm curious um around the, you know, obviously Crazy Egg, you know, well-known, very established, as you mentioned, Heaton really determined or really define the category.
00:11:49.360 --> 00:11:51.840
2004 brings me back for sure.
00:11:52.159 --> 00:11:55.440
Um you guys are running your own tests internally.
00:11:55.840 --> 00:12:10.720
Talk to me kind of maybe that first year or so in your role at Crazy Egg and some tests that you were you ran with the team and had some wins, but I'd also like to hear about maybe some some failures or certain things.
00:12:10.879 --> 00:12:23.519
Because I think this is a part when I not to talk about myself here, but when we our our clients, we want that rapport, that personal rapport, but we also want to prove ourselves through performance rapport.
00:12:23.679 --> 00:12:24.720
That's what I always say.
00:12:25.039 --> 00:12:39.039
But it's ultimately that relationship that they can trust us to essentially fail forward, to go out and test and to know that things aren't going to be perfect all the time, but we've got a strong conviction, we've got a hypothesis to move forward.
00:12:39.120 --> 00:12:48.080
So I'm curious, yeah, what what maybe some certain things that have worked well on Crazy Egg specifically, and maybe some things that were ultimately a massive failure.
00:12:48.480 --> 00:12:48.960
Yeah.
00:12:49.279 --> 00:13:02.320
So surprise, surprise, as soon as I came to Crazy Egg, the first thing we started doing is not only collecting heat maps on, you know, our homepage, our pricing page, I ran so many ABU tests, like so many different headlines, so many different things.
00:13:02.399 --> 00:13:13.200
And the funny part is that kind of um we ended up finding that for the existing product, that a lot of the existing kind of visuals and a lot of the same messaging already worked, right?
00:13:13.360 --> 00:13:17.279
Or at least um the test didn't improve what was already working.
00:13:17.519 --> 00:13:25.600
Um, but the thing is that you know your data is always gonna is always going to be biased towards the people who you're measuring or the people you're hearing from.
00:13:25.840 --> 00:13:29.039
And so I ended up jumping on some of these customer calls also, right?
00:13:29.120 --> 00:13:32.080
To um kind of figure out like, okay, what do people actually want?
00:13:32.240 --> 00:13:33.200
What's their experience been?
00:13:33.440 --> 00:13:34.240
How can we improve it?
00:13:34.399 --> 00:13:36.960
Those kind of like product um feedback calls.
00:13:37.279 --> 00:13:48.799
And the way that kind of these calls always went, right, is people we are talking about who are people who are already paying customers for crazy egg, who installed these heat maps on their page and are like getting a lot of data, right?
00:13:48.960 --> 00:14:03.519
And so we pull up a he map report in their account and kind of describe like, okay, this button's getting the most clicks, people are scrolling down this far, uh, people are navigating to this URL the most, and they say, okay, great.
00:14:04.240 --> 00:14:05.039
What do I do?
00:14:05.440 --> 00:14:06.399
Yeah, totally.
00:14:06.720 --> 00:14:10.720
Which is the classic kind of cliff almost of analytics, right?
00:14:10.799 --> 00:14:19.360
Which is like, hey, you can just most analytics are descriptive, but the real job to be done is just like tell me what to do to increase that performance.
00:14:19.600 --> 00:14:28.559
And historically, Betts needed someone like me to actually have the time to kind of like look at the data, hope that there's actually a pattern that's clear, and then make that jump.
00:14:28.799 --> 00:14:32.639
And as it turns out, is you know, we have kind of like a free trial model.
00:14:32.720 --> 00:14:38.000
Um, we promise that people can um kind of learn how to optimize their websites based off of it.
00:14:38.159 --> 00:14:55.600
And then we have people signing up, starting a free trial, going into the product, seeing the data, and still having that question of like, I can't justify kind of continuing to work with product when I, you know, uh some we have some people who are like full-time day alists, we have a lot of people who are like, you know, um small business owners, right?
00:14:55.759 --> 00:15:01.440
Who just want to get the answer because they need to go on to running the the freaking business, right?
00:15:01.600 --> 00:15:06.720
And they're telling me, like, I can see the data, I just don't know what I'm what it's supposed to mean or what I'm supposed to do with it.
00:15:06.879 --> 00:15:15.360
And so that kind of taught me like, oh, the job to be done, there's a there's a uh very clear job to be done that is still not being solved, right?
00:15:15.440 --> 00:15:29.600
And kind of if you think, imagine that kind of like the end result in terms of like improving engagement and uh reducing bounce rate of increasing conversions, that's a very end, but there's a whole bunch of stepping stones in between, and there's some pretty clear stepping stones that we're not solving here.
00:15:29.759 --> 00:15:36.639
And if we can't prove that, that's the most concrete thing that we're not showing in terms of proving that our product is doing what it's supposed to do.
00:15:36.879 --> 00:15:51.200
So we kind of went back to the drawing board and we're like, you know, if only there was a technology right now that could analyze the data and kind of, you know, find the patterns and all this and all this information and then make some recommendations or come up with some ideas.
00:15:51.360 --> 00:15:53.279
Um, and as it turns out, that exists now.
00:15:53.440 --> 00:16:04.080
So we kind of thought to, okay, what is the tool that I would want to do that job for me so that I can just go ahead and start running those tests that I know are actually gonna be um higher leverage.
00:16:05.519 --> 00:16:15.600
Yeah, it's it's uh and you're obviously we'll we're gonna we'll transition into kind of the the analysis piece that you guys have built out with crazy egg because I think it's wildly impressive.
00:16:16.720 --> 00:16:19.919
I love the um I think everyone does.
00:16:20.159 --> 00:16:25.200
I I told this story with Tahiton a number of years ago that I was on a plane and I had Crazy Egg up.
00:16:25.440 --> 00:16:32.399
This is a long time ago, and it was you know how sexy things look, you know, with the confetti in particular, one of my favorite reports by far.
00:16:32.720 --> 00:16:35.440
And I remember seeing the guy next to me, he's like, What is that?
00:16:35.519 --> 00:16:36.000
That's amazing.
00:16:36.080 --> 00:16:36.720
You gotta bring that.
00:16:36.879 --> 00:16:38.480
He's like, I'm going to Europe right now.
00:16:38.559 --> 00:16:41.279
I'm in marketing, we've never seen anything like that.
00:16:41.519 --> 00:16:48.879
And it was so funny because it does, it is that very attractive, uh compelling insight.
00:16:48.960 --> 00:16:55.200
But at the end of the day, or just kind of a view, and at the end of the day, it's like, what do I do with that information?
00:16:55.440 --> 00:16:59.840
And I think you're you're spot on in terms of the same thing with GA, right?
00:17:00.000 --> 00:17:02.559
And okay, I've configured some certain things.
00:17:02.799 --> 00:17:03.919
Okay, it tells a piece.
00:17:04.000 --> 00:17:12.319
And it isn't, uh I'm curious your thoughts and how and on how we uh articulate it to our clients and our prospects is like, great, GA is a portion of it.
00:17:12.400 --> 00:17:19.119
Now, layering on GTM with custom events, conversion tracking elements, uh, et cetera, that's more of it.
00:17:19.200 --> 00:17:26.640
Now getting a uh heat mapping software, click tracking software like Crazy Egg in particular, that's more of the pie in terms of the view.
00:17:26.799 --> 00:17:34.960
But we're still missing a lot of the qualitative insights, like survey-related technology or audience actual information, still more of it.
00:17:35.279 --> 00:17:39.920
Never perfect necessarily, but I think with a lot of those kind of directionally accurate.
00:17:40.000 --> 00:17:42.559
What are your thoughts in terms of that setup?
00:17:43.039 --> 00:17:52.240
What I've just started, which I mentioned, versus a lot with a lot of brands of like we've got basic tracking within GA, some within GTM, and now we're doing heat mapping.
00:17:52.400 --> 00:17:54.480
Are they limiting their cell themselves?
00:17:54.720 --> 00:17:57.279
But they have enough insight to move on.
00:17:57.519 --> 00:17:59.599
Um, talk to me about that.
00:18:00.960 --> 00:18:05.119
I would say that you know, the more data you have almost always the better.
00:18:05.359 --> 00:18:17.599
Um, what's really funny these days is that it used to be that kind of people talk about, oh, all your data is in these different silos and you can't compare it, so you gotta all put it into that one central CRM in order to compare all of it.
00:18:17.839 --> 00:18:21.759
And, you know, inevitably this conversation always goes to AI.
00:18:21.839 --> 00:18:32.640
But one funny use case that I found really good with AI is its ability to kind of actually analyze data that's in different silos that's not necessarily comparable one-to-one and actually relate these things.
00:18:32.880 --> 00:18:41.680
So if I have data that's you know, um in a CRM in one thing, and then I have behavioral data in another that can tell me like which elements are coming in.
00:18:42.079 --> 00:18:46.319
Ordinarily, a human analyst would say these are not kind of apples to apples comparisons.
00:18:46.480 --> 00:18:49.200
And so you kind of have to analyze things, things separately.
00:18:49.359 --> 00:18:59.680
But what I found helpful to do when uh an analyze with an AI is to just say, hey, like look through these data repositories and find me kind of patterns which are matching in these.
00:18:59.839 --> 00:19:05.359
And they is very often able to say, like, hey, this particular headline or this particular offer is doing really well.
00:19:05.519 --> 00:19:11.200
And I find in your CRM that this uh the this skew is being poached the most often by our customers.
00:19:11.359 --> 00:19:12.799
And these things are paralleling.
00:19:12.960 --> 00:19:28.480
Now, what's interesting is uh ordinarily BA would require a kind of um a really rare kind of human data analyst who can both look at numbers, look for trends, but also kind of relate, like, hey, buying the most of this product and relating the most of this offer, there's something qualitative that's in common here.
00:19:28.640 --> 00:19:33.039
That's kind of requiring you to cross both qualitative and quantitative data.
00:19:33.119 --> 00:19:35.759
But AI is actually surprisingly good at doing these.
00:19:35.920 --> 00:19:41.359
So I would say that having siloed data is not necessarily as big of a problem as it used to be.
00:19:41.440 --> 00:19:47.519
And certainly, if you're starting from zero, working with data silos can still get you much better than nothing.
00:19:47.759 --> 00:20:01.039
But if you're um a very high traffic brand, if you're trying to optimize for many different um products and many different segments, then analyzing this stuff becomes much more of a scalpel than a sledgehammer kind of um problem.
00:20:01.200 --> 00:20:08.240
And that's the point at which you start to need a lot more kind of like um data or previously siloed data that's talking to each other.
00:20:08.880 --> 00:20:09.039
Sure.
00:20:09.359 --> 00:20:09.599
Yeah.
00:20:09.839 --> 00:20:11.440
Yeah, and easier said than done too.
00:20:11.519 --> 00:20:18.079
I think there's always just like, oh, you know, that's uh I love the aspect of you know, crazy egg is so easy to get going.
00:20:18.640 --> 00:20:30.000
But I think in certain cases it's it's it's not necessarily helpful because some of these brands just haven't configured their GA or GTM instance to a place where Crazy Egg is already at.
00:20:30.880 --> 00:20:31.039
Yeah.
00:20:31.200 --> 00:20:55.680
And maybe one more thing to add is one maybe other big reason to kind of like um simplify your tech stack ultimately is because ultimately the more complexity you have your operation, be that, you know, like um, you know, kind of a lot of stakeholders or a lot of daily silos, is every additional silo um or a kind of layer of complexity adds more uncertainty to any given decision, which makes it harder to kind of like have any conviction around any of its stuff.
00:20:55.920 --> 00:21:04.000
So, you know, um certainly we have a lot of folks who use Crazy Egg in conjunction with a separate A-B testing tool or, you know, GA4 or things like that.
00:21:04.079 --> 00:21:09.839
And if that's your setup, then we are more than happy to kind of find a way to, you know, solve a certain point solution for you.
00:21:10.000 --> 00:21:24.240
But one advantage that a lot of our customers who do mainly use us for a lot of these different things is they have one source of truth for a lot of these things, even if it's different features, and therefore they kind of have a lot more conviction when it comes to making a decision fast.
00:21:24.480 --> 00:21:34.880
And these days it turns out that um information less and less is a barrier, but having conviction to actually take an action is really a thing that takes more time these days.
00:21:35.119 --> 00:21:35.920
Yeah, I love that.
00:21:36.000 --> 00:21:39.759
Just it's uh I couldn't have said it better myself, just the simplicity aspect.
00:21:39.920 --> 00:21:49.680
It reminds us, it reminds me of just the countless audits we've done of like a GTM instance, and there's just like inflated events all over the place.
00:21:49.839 --> 00:21:52.160
And oh yeah, that was configured a couple of years ago.
00:21:52.319 --> 00:21:53.279
No one really looks at that.
00:21:53.359 --> 00:22:01.759
It's like, okay, your historical event data is massively skewed, and we want to marry this with our heat mapping data and Unfortunately, it's not great.
00:22:01.920 --> 00:22:05.279
So and the the pages have changed ridiculous amount.
00:22:05.519 --> 00:22:16.880
Or honestly, we've we've actually onboarded clients in the past too that have Crazy Egg, for example, are tracking way too much, you know, and are like, this isn't I'm like what what benefit is this?
00:22:17.119 --> 00:22:24.559
So it is kind of really ratcheting it back and and uh honing in a more simplistic infrastructure before you expand.
00:22:24.880 --> 00:22:25.119
Yeah.
00:22:25.200 --> 00:22:30.960
I mean ultimately these days, you know, uh frankly, I think the days of tracking too much are kind of over.
00:22:31.119 --> 00:22:35.279
Um the funny part is that, you know, right, uh peep it's it's funny.
00:22:35.359 --> 00:22:40.480
I think 10, 15 years ago, people were going on about like, oh, data is the new oil, you know, it's gonna change everything.
00:22:40.640 --> 00:22:42.960
And but it managed to the bottleneck I mentioned earlier, right?
00:22:43.039 --> 00:22:49.519
It's like ultimately you need a guy like, you know, um young me to look actually look at this stuff and figure out what's gonna go on.
00:22:49.680 --> 00:23:02.799
And turns out that kind of what we might call a reasoning or intelligence bottleneck is no longer there because now you can have kind of a machine crawl through that data and find it for you, even to find patterns which you didn't even realize were bare.
00:23:03.119 --> 00:23:14.720
So, in a in a sense, kind of like that uh that old pattern of like, hey, we have more data than we know what to do with, so we're just kind of gonna throw it in a lake and then wait for a later data until we have someone who can actually take a look at it.
00:23:14.880 --> 00:23:18.400
Those days are over because like analysis is no longer a bottleneck.
00:23:18.480 --> 00:23:24.319
But figuring out what are we actually gonna do with that, that is kind of where we start to run into more bottlenecks these days.
00:23:24.559 --> 00:23:26.160
Yeah, excellent point, excellent point.
00:23:26.240 --> 00:23:33.039
And I think still, like, yeah, it's not a bottleneck, but I think it still does require a experienced mind too.
00:23:33.200 --> 00:23:34.559
Like you can get a lot of great insights.
00:23:34.640 --> 00:23:42.640
And look, we're gonna talk about some of the features, obviously, of Crazia here in a second, but it still requires like at least some a background a bit, right?
00:23:42.720 --> 00:23:47.440
And and some directional analysis, especially prompt structures, um leveraging AI.
00:23:47.680 --> 00:23:56.880
Steven, hey, you know, big fan of Crazia, obviously, and what you guys have been able to accomplish, big fan of yours and Heaton and and the rest of the team.
00:23:57.119 --> 00:24:01.519
Um, you know, it's a it's a saturated market in certain cases, right?
00:24:01.599 --> 00:24:05.920
You mentioned Heaton really starting the category, as I mentioned already in 2004.
00:24:06.079 --> 00:24:11.440
Um, Hotchard, Lucky Orange, Microsoft Clarity, the list goes on in certain things.
00:24:11.599 --> 00:24:14.559
I know Google at one point had their you know in-page analytics.
00:24:14.640 --> 00:24:15.599
That was pretty bad.
00:24:15.759 --> 00:24:18.799
Um, I'm surprised they didn't try to capitalize on that.
00:24:19.279 --> 00:24:25.359
You guys have made some significant strides um in the last year plus and continue to make those.
00:24:25.599 --> 00:24:26.400
Let's talk about that.
00:24:26.559 --> 00:24:30.720
I mean, I I love, you know, I love what you're able to do, obviously, on the A-B testing component.
00:24:30.799 --> 00:24:45.359
We're we're we've been alluding to it, but more of the AI rollout component of being able to empower individuals to become more of, as you mentioned before, of like their own professional services arm to analyze and make decisions on certain things.
00:24:45.599 --> 00:24:51.200
Talk to me, talk to us about that and how you see kind of the future of heat mapping and particularly around Crazy Egg.
00:24:51.680 --> 00:24:51.920
Yeah.
00:24:52.079 --> 00:24:55.599
So, I mean, obviously we have a lot of, you know, there's a lot of choice in the market, right?
00:24:55.680 --> 00:24:58.000
You've got your hot jars, you have your content squares.
00:24:58.319 --> 00:25:08.400
Um, and kind of whenever you look at a lot of these things, right, is, you know, every company starts very often starts out with one particular kind of role that it's supposed to play.
00:25:08.559 --> 00:25:22.000
And if we look at a hot jar content square kind of traditional crazy egg, what that role served was uh providing a data analyst with a product that will give them data and the ability to produce reporting, right?
00:25:22.160 --> 00:25:24.079
The end goal is reporting.
00:25:24.799 --> 00:25:28.720
And, you know, I when I was younger, kind of I that's the way I thought of my job as well.
00:25:28.880 --> 00:25:29.680
But like I just said, right?
00:25:29.759 --> 00:25:36.240
It's like ultimately, hey, it's like to drive more measurable uh impact, to drive more revenue, to make a better customer experience.
00:25:36.400 --> 00:25:41.519
So as it turns out, um, the way I see it these days is reporting is not the end goal.
00:25:41.759 --> 00:25:50.240
Um, and so when we looked at our competitors and we see the way that like what's capable with technology, what's uh what you can achieve with technology is expanding.
00:25:50.400 --> 00:26:03.680
But we look at a hot jar and a content square, and they're ultimately just um producing a lot of stuff that accelerates the job of reporting, but doesn't accelerate the real end result which you're trying to achieve, which is better business outcomes, better customer experiences.
00:26:04.000 --> 00:26:05.920
We think that falls a little bit short.
00:26:06.160 --> 00:26:19.519
And it's kind of an inert, they're kind of falling into an inertia that a lot of companies have, which is like going deeper into that same job we've done when really we should be going back to the drawing board and say, like, hey, what is the thing that we should be trying to achieve here?
00:26:19.759 --> 00:26:30.640
So the way we've gone back to the drawing board is kind of thought like, okay, we already have you know analytics in there, and we can slowly apply AI kind of in a very obvious way to bring insights out of that.
00:26:30.799 --> 00:26:32.559
We have an A-B testing tool.
00:26:32.799 --> 00:26:47.200
But um, and on top of that is, you know, we obviously kind of the capability to kind of generate copywriting really, really quickly is obviously out there, but but let's not get into that because kind of there's there's good there's good ways to do it.
00:26:47.279 --> 00:26:49.759
There's obviously a lot of terrible ways to do it, right?
00:26:50.160 --> 00:26:59.440
But ultimately, the way we thought about it is like, hey, if the end goal is to unify a lot of these jobs to be done and then take you to like, hey, how do I improve my customer experience?
00:26:59.680 --> 00:27:12.480
We're obviously gonna produce some of these analytical tools so that you know a marketing analyst or a marketer can kind of get in there, um, produce a report really quickly that they can take to their you know team and say, like, hey, here's what's happening.
00:27:12.640 --> 00:27:14.240
Let's discuss what we want to do.
00:27:14.400 --> 00:27:26.400
But you know, a marketer who's trying to kind of be a jack of all trades can come in, look at the data, immediately get a bunch of uh ideas for A-B tests or personalization, even that they can readily really quickly.
00:27:26.960 --> 00:27:32.799
And this is gonna get to a point where you can launch this with just a few clicks, or maybe even one click in the future.
00:27:33.039 --> 00:27:43.680
Um, and you know, the way that these things are going so quickly, you know, like we're obviously working on like, you know, shipping new dashboards, new forms of reporting for this as well.
00:27:43.839 --> 00:27:47.599
But you know, we're we've been working on audience targeting for A-B testing the last couple months.
00:27:47.759 --> 00:27:48.880
That's already live.
00:27:49.119 --> 00:27:53.920
Potentially, you know, by the end of this year, when's this gonna be full-blown personalization?
00:27:54.079 --> 00:28:00.160
But what really ultimately boiling down to is like, hey, let's go beyond just optimizing a website for everyone.
00:28:00.319 --> 00:28:04.960
What if we could deliver the right message to the right person at the right time?
00:28:05.200 --> 00:28:06.720
That technology exists now.
00:28:06.799 --> 00:28:11.359
So we should go and build that because ultimately I think that's where the puck is headed.
00:28:11.519 --> 00:28:14.079
And we shouldn't be thinking of what's possible three months down the line.
00:28:14.160 --> 00:28:16.880
We should be thinking what's going to be available two, three years down the line.
00:28:17.039 --> 00:28:17.279
Yeah.
00:28:17.359 --> 00:28:17.759
Yep.
00:28:18.079 --> 00:28:18.799
Personalization.
00:28:18.880 --> 00:28:21.279
How how many times have we heard that over the years, Steven?
00:28:21.519 --> 00:28:28.880
And I remember like uh, I forget, we were working with some.NET um CMSs, and they're like, oh, we do a ton of personalization.
00:28:29.039 --> 00:28:30.319
It was USL.
00:28:30.720 --> 00:28:31.680
Total bullshit.
00:28:31.759 --> 00:28:39.759
You know, it's like no, I mean, I've launched a lot of ABM campaigns in my life where we were uh, you know, using a variable to throw the customer's logo in there and stuff like that.
00:28:39.839 --> 00:28:42.400
And sure, that's technically personalization, right?
00:28:42.720 --> 00:28:44.480
Um, but that's service level.
00:28:44.720 --> 00:28:46.400
What and like I mentioned earlier, right?
00:28:46.480 --> 00:28:51.680
It's like it's one thing to change a headline, it's another to use a different value prop or a different offer.
00:28:51.839 --> 00:28:59.839
Um, and the challenge being that, like, well, with the current tools right now, you can certainly kind of like write five different variants of a headline and test these out.
00:29:00.000 --> 00:29:13.839
But what if, for example, the AI could look at what your um customer is track coming from as far as like a UTM parameter or you know, your ad headline, for example, and figure out like, oh, they want offer A, let's give them offer A.
00:29:14.160 --> 00:29:18.319
And then Ideally can do that like much faster, even programmatically.
00:29:18.559 --> 00:29:32.799
Um, you know, I live in it's weird to kind of like live in these kind of like weird tech worlds nowadays, but it's kind of like when you think beyond just automating a workflow to what does this mean for my actual customer experience, it starts to get pretty tetty.
00:29:33.279 --> 00:29:34.160
Yeah, yeah.
00:29:34.319 --> 00:29:35.440
Yeah, that's hilarious.
00:29:35.519 --> 00:29:39.359
That example of this uh infusing or including the logo.
00:29:39.519 --> 00:29:40.880
That's a classic.
00:29:41.200 --> 00:29:57.039
Uh let's talk just, you know, um, you mentioned the copywriting component, you know, we could talk about that for a long time, but how what how do you bridge the gap or just letting AI interpret behavioral data versus requiring a human to sign off?
00:29:57.200 --> 00:29:59.279
What is your thought process there?
00:29:59.759 --> 00:30:04.319
Um and opinion, I guess, there right now.
00:30:04.960 --> 00:30:05.440
Definitely.
00:30:05.599 --> 00:30:09.519
So I mentioned, you know, that one of the things I'm good at is pattern recognition, right?
00:30:09.599 --> 00:30:18.400
And I think ultimately a lot of data analysis, it doesn't numbers and running quantitative stuff is one way to do it, but qualitative is there as well.
00:30:18.559 --> 00:30:20.319
So what is the thing that unifies all of this?
00:30:20.400 --> 00:30:26.319
I think pattern recognition, frankly, is the if you would boil down the main job of data analysis.
00:30:26.799 --> 00:30:30.960
And, you know, when I I think the best example of this was actually when I was working at Wavefair.
00:30:31.119 --> 00:30:33.759
I was thinking I was there about a year before they IPO'd, right?
00:30:33.920 --> 00:30:40.000
Millions of products, hundreds of millions of pay of, you know, um, of pages and sessions, right?
00:30:40.160 --> 00:30:43.839
Do you think I as a data analyst, even back then, it's gotta be a lot bigger now?
00:30:44.000 --> 00:30:54.799
You know, if I were a looking at this in the most excruciating depth, I would be watching every single visitor, kind of taking notes on every single person, want to be clicking on all this stuff.
00:30:55.039 --> 00:30:56.160
But I don't do that.
00:30:56.400 --> 00:31:05.359
I run some queries, I take some uh take some averages, some counts, some sums, I do some segmentation analysis.
00:31:05.599 --> 00:31:11.759
All that really boils down to is simplifying data in such a way that a human being can actually process what's going on.
00:31:11.920 --> 00:31:13.839
Because ultimately, we human beings, we're limited.
00:31:13.920 --> 00:31:15.759
There's only so much information we can process.
00:31:15.920 --> 00:31:19.200
And if you look at a heat map, that's actually what's happening there as well, right?
00:31:19.279 --> 00:31:25.599
We are aggregating a whole bunch of click data and representing it in the visual format so that a human being can actually spot transfer.
00:31:26.400 --> 00:31:35.599
Whenever you simplify information, you're getting rid of some um information, getting rid of some signal in there because you're trying to also screen out noise.
00:31:36.000 --> 00:31:41.200
But the thing about AI is that you can actually throw a lot more processing power at the problem.
00:31:41.279 --> 00:31:49.920
And as a result, it can actually go through, check out every single click, and be like, hey, you know, um people from Indiana are buying a ton of this given product.
00:31:50.160 --> 00:31:56.240
Maybe we should be targeting uh more of more campaigns and offers around this particular product to folks in Indiana, right?
00:31:56.480 --> 00:32:01.599
That's a level of specificity, but you can only provide if you had previously a ton of human data analysts.
00:32:01.759 --> 00:32:06.240
But now you can actually have AI go through and look for a lot of that pattern recognition as well.
00:32:06.640 --> 00:32:16.480
And so the way I think about AI nowadays is that like that pat that job of pattern recognition, in a sense, frankly, I think a lot of the kind of like AI hype is a little bit overblown.
00:32:16.720 --> 00:32:26.960
AI, to me, first and foremost, is a master pattern recognition machine, which sounds like I'm downplaying what it's capable of, but actually that's a pretty world-changing capability.
00:32:27.359 --> 00:32:37.119
And so if you can use AI to like, hey, find me patterns in this engagement data and find me opportunities to sell more of this given product, it can find a lot more of that as well.
00:32:37.359 --> 00:32:45.039
Now, there's still room for human judgment in there, which say, hey, give me a bunch of ideas, um, which optimizations are on brand for us.
00:32:45.200 --> 00:32:49.839
But I can, as a human being, can look at it and say, oh, sure, those are the five most obvious ideas.
00:32:49.920 --> 00:32:53.200
But here's ideas six, seven, and eight that no one's thought of yet.
00:32:53.359 --> 00:32:55.119
Let's throw that into the mix as well.
00:32:55.359 --> 00:33:00.880
So, in a sense, it's going beyond just like kind of the roach um workflow automation.
00:33:01.039 --> 00:33:14.559
And when it comes to um performing jobs which are basically replicating and recognizing patterns, it's doing a ton of that as well, which means we human beings can kind of accelerate that process and get to the part where we apply our judgment about what's important.
00:33:15.119 --> 00:33:16.240
Yeah, really well said.
00:33:16.400 --> 00:33:17.039
Well said stuff.
00:33:17.119 --> 00:33:19.200
Pattern recognition is huge, obviously.
00:33:19.279 --> 00:33:30.960
And that's I think you you and I have just been relying on that for years because we've seen so many of these things and we can make some assumptions without having to look at all the data, but obviously now things are making it a little bit easier.
00:33:31.279 --> 00:33:31.839
Uh for sure.
00:33:32.000 --> 00:33:33.279
A lot easier, let's be honest.
00:33:33.920 --> 00:33:36.160
Definitely let's talk about some bets here.
00:33:36.559 --> 00:33:47.920
Um, I'm curious in terms of we talked about it before we hit record earlier in terms of just the um where behavioral analytics is is goes in the next two years.
00:33:48.480 --> 00:33:58.240
And where do you think um most of the industry industry potentially maybe disagrees with where it's heading versus maybe where you where you think are things are heading?
00:33:58.319 --> 00:33:59.920
Um I'm curious your thoughts.
00:34:00.240 --> 00:34:07.680
Yeah, I think, you know, um I think people are rightly uh kind of either burnt out or feeling the overhype of AI, right?
00:34:07.759 --> 00:34:11.119
Um, you know, I love on the East Coast, I believe you're an East Coaster too, right?
00:34:11.280 --> 00:34:18.960
Um, and you know, what's funny is, you know, I get exposed to lots of folks like Heaton, the Bay Area types who think that AI is kind of gonna change everything.
00:34:19.119 --> 00:34:20.639
Maybe they're right, maybe they're wrong.
00:34:20.800 --> 00:34:29.199
But when I talk to folks out here in the quote unquote real world, and most people are frankly using it as a glorified search engine, right?
00:34:29.599 --> 00:34:31.840
Um now I think that's gonna change.
00:34:31.920 --> 00:34:34.320
I think that's actually a generational change, right?
00:34:34.400 --> 00:34:45.119
Like um a lot of folks, you know, I'm a millennial, I still like to make large purchases on a desktop, but you talk to kind of younger kids, and I'm perfectly fine buying things on a mobile phone.
00:34:45.280 --> 00:34:49.440
And they're also the ones who are frankly using AI more to more on a day-to-day basis.
00:34:49.519 --> 00:34:59.679
So I think this kind of thing that's gonna creep up more so then, well, it's some things are gonna change very quickly, but then it's gonna make a kind of more macro change from the bottoms up.
00:35:00.719 --> 00:35:05.920
Um, but so I think kind of it's not gonna move quite as quickly as everyone thinks.
00:35:06.079 --> 00:35:08.320
But at the same time, again, like I mentioned, right?
00:35:08.400 --> 00:35:10.400
It I think it's still gonna change a ton.
00:35:10.559 --> 00:35:14.159
Um, I think when it comes to behavioral analytics in particular, right?
00:35:14.320 --> 00:35:24.639
The capability of bridging that gap from just collecting that data, making um actually looking at it and turning that into an actual insight, that's gonna happen super fast.
00:35:24.880 --> 00:35:29.679
Um, and frankly, we're kind of working on trying to accelerate that change as well, right?
00:35:29.840 --> 00:35:32.320
So obviously, of course, I believe that.
00:35:32.480 --> 00:35:37.599
Um, but on top of that is, you know, we mentioned personalization earlier, right?
00:35:37.760 --> 00:35:46.719
And I think that one thing I really believe, but I think people are focusing on the wrong things, is they're looking at workflow automation for a lot of these use cases.
00:35:46.880 --> 00:35:48.159
I'm not saying that's a bad thing.
00:35:48.320 --> 00:35:50.079
I'm saying you're not thinking big enough.
00:35:50.320 --> 00:35:53.920
Because ultimately you're kind of thinking in terms of like, hey, how can I produce faster reporting?
00:35:54.159 --> 00:35:56.719
How can I spit out blog posts more quickly?
00:35:56.960 --> 00:36:03.599
But you know, in the world we live in, right, it's like, is more spitting out blog posts more quickly going to get me more eyeballs?
00:36:03.760 --> 00:36:06.159
No, because of what's happening with all our channels.
00:36:06.400 --> 00:36:10.320
But we can get to a point really quickly where we can process data really quickly.
00:36:10.559 --> 00:36:17.440
We can um see people that are coming in through, say, a meta ad versus a you know, organic post versus an email.
00:36:17.599 --> 00:36:20.960
And we have an ability to kind of like adjust.
00:36:21.119 --> 00:36:33.199
Um, I can take my kind of brand guidelines, my main copywriting, and then adjust, pattern recognize um my creative or my offer or my copywriting for a specific um user.
00:36:33.280 --> 00:36:36.719
Um, the kind of analogy I use for it, frankly, is code switching, right?
00:36:36.800 --> 00:36:46.239
How do I um code switch a little bit to this customer so that they I can use the words that they would use in their own mind and that they get the uh the value back much more quickly.
00:36:47.519 --> 00:36:54.400
This is gonna become so quick, I think, that we're not just gonna be able to like push out social media content much more quickly.
00:36:54.639 --> 00:37:04.239
I think we can get to the point where your website very quickly can kind of figure out who's visiting it and then target bright personalized offer to the right person at the right time.
00:37:05.119 --> 00:37:09.440
And that's I think a few more steps ahead of where people are going.
00:37:09.519 --> 00:37:14.239
But I think it's where a lot of this technology is headed in a few years beyond simple workflow automation.
00:37:14.559 --> 00:37:21.519
And I think especially if we can get to the point where it's not just doing copywriting, but the other part is discipline, right?
00:37:21.599 --> 00:37:25.199
Can we discipline based off of what's actually getting engagement and conversions?
00:37:25.599 --> 00:37:30.079
That's how you get higher performance, where people will start to see the value of that particular technology.
00:37:30.960 --> 00:37:32.000
Excellent stuff.
00:37:32.880 --> 00:37:36.320
Steven, talk to me, unpack, and we'll wrap up here soon.
00:37:37.039 --> 00:37:40.880
This is uh this has been great and super um impactful items.
00:37:41.599 --> 00:38:03.599
I'm curious, a lot of CRO insights, incredibly powerful, deep discovery within analytics, obviously map mirroring or mapping out and um articulating heat mapping insights, GA data, um, and a lot of these CRO reports are proposing wireframes and things like that.
00:38:03.760 --> 00:38:10.159
None of all of this has been um happening for years, but now it's happening a lot faster, as we know.
00:38:10.239 --> 00:38:15.920
It's much easier to compile all that with cloud design, with all sorts of other things that have been really amazing.
00:38:16.079 --> 00:38:20.960
Uh that said, I'm curious your thoughts, because uh we've experienced this in other agencies as well.
00:38:21.039 --> 00:38:23.199
They really love the data, they love the insight.
00:38:23.599 --> 00:38:27.920
But when it comes to actually implementing tests, there's a bit of hesitation.
00:38:28.159 --> 00:38:46.079
And I think I love you know what Crazy Egg's been able to do, but it's it's a little bit, it's like, oh gosh, we have to engage the copywriter, we have engaged our developer, we have to engage a UX team, we have to engage all these other and it's and it's it's new, it's it's nuanced, it's intimidating, it's potentially time consuming.
00:38:46.559 --> 00:38:56.960
What would you say of how to kind of convince a skeptical CMO to kind of run their first or second test, if you will, for CRO that is perhaps again skeptical of of this whole plan?
00:38:57.440 --> 00:39:06.719
Yeah, I would say that you know, if you're expecting that either this is gonna produce as at least right now, this will probably change next year.
00:39:07.039 --> 00:39:20.559
But if you're expecting that, hey, I want this to be able to generate a finished product from a first bet, or as we would say, like four from a one-shot, then I think that's kind of an unreasonable expectation.
00:39:20.719 --> 00:39:25.760
And frankly, you don't need to have a finished product to benefit from AI, right?
00:39:25.920 --> 00:39:28.000
Um, we've talked about like wireframes, right?
00:39:28.159 --> 00:39:29.679
We talked about prototyping.
00:39:30.000 --> 00:39:35.199
The the funniest thing that I found as a copywriter is my biggest enemy is a blank page, right?
00:39:35.360 --> 00:39:42.800
Is it's very turns out I have a lot of opinions about what the design and copywriting looks like once I have a bad example in front of me.
00:39:43.119 --> 00:39:43.679
For sure.
00:39:44.000 --> 00:39:44.320
Right.
00:39:44.559 --> 00:40:01.039
And so what's funny is, you know, I can um, but as it turns out, it's really easy to get a bad but a stepping stone first draft by having the AI look at a lot of say my brand guidelines, look at a lot of my positioning, even feedback that I've gotten from customer quotes, right?
00:40:01.199 --> 00:40:07.599
To produce a first draft that is, let's call it a one-year out-of-school quality first draft.
00:40:07.760 --> 00:40:11.119
Like uh 23-year-old me, what my copywriting might have looked like.
00:40:11.599 --> 00:40:28.239
And it can produce something, like say a slide deck or a page prototype and say, hey, based off of uh all the training documents you provided me right now, this is what, say, a train a sales slide deck or a uh you know agency-facing landing page should look like.
00:40:28.400 --> 00:40:34.960
I can look at it and say, okay, there's some poor judgment here as far as like which um stuff that agencies want to look at.
00:40:35.119 --> 00:40:37.199
Um, but I can see the nuggets in here.
00:40:37.280 --> 00:40:40.079
I can say, hey, all right, let's move these sections further up.
00:40:40.239 --> 00:40:43.360
Let's rewrite these particular sections, but working upon the same theme.
00:40:43.519 --> 00:40:50.480
And as it turns out, it's like at least automating the uh the job of getting a bad first draft on.
00:40:50.559 --> 00:40:54.239
And almost every first draft of it has ever been made by humanity is bad, right?
00:40:54.400 --> 00:40:57.599
So so we don't need a good quality first draft.
00:40:57.679 --> 00:41:07.599
We need something to react to so that we can start slotting in our actual inspiration, our actual original ideas, and kind of factoring the this stuff in as well.
00:41:07.840 --> 00:41:19.039
So I would say, frankly, hey, if you're looking to just get started, you can use it for competitor research, for example, and say, like, hey, um, I'm saying I'm a marketer looking at Hotjar right now.
00:41:19.119 --> 00:41:20.320
What is hotjar good for?
00:41:20.400 --> 00:41:22.960
And where is Crazy Egg um a better alternative?
00:41:23.199 --> 00:41:29.440
That's super interesting because I can kind of tell, like, hey, if I had no information whatsoever, what are my customers going to see?
00:41:29.599 --> 00:41:44.800
Um, but conversely is, you know, if I produce a say sales collateral based off of my all my internal documents and produces something that is off message, that implies that the internal kind of documents that I have are off message and something's missing.
00:41:44.960 --> 00:41:50.079
So that highlights gaps basically in terms of what I'm actually even giving my team, but I can provide them as well.
00:41:50.320 --> 00:41:57.920
So what I would frankly say is like, hey, AI is kind of like a new kind of way of computing, but instead of programming, you're actually training it.
00:41:58.079 --> 00:42:06.000
But you know, sometimes if your employee is doing something new and they're doing something not the right way, that's a gap in my training, right?
00:42:06.159 --> 00:42:13.679
And as it turns out, if the AI is not quite getting 90% there, then hey, there's some training we can do, and that's something we can address.
00:42:14.000 --> 00:42:17.519
Yeah, that foundational component training up is so key.
00:42:17.760 --> 00:42:19.280
Steven, it's been an absolute pleasure.
00:42:19.440 --> 00:42:20.639
Really appreciate your time.
00:42:20.800 --> 00:42:23.920
Hey, where can people interact and engage with you?
00:42:24.000 --> 00:42:27.360
And then obviously Crazy Egg, talk talk about that a little bit.
00:42:27.599 --> 00:42:27.760
Yeah.
00:42:27.840 --> 00:42:30.480
So believe it or not, I actually take direct customer emails.
00:42:30.639 --> 00:42:33.679
So if you want to email me, I'm at steven at crazyag.com.
00:42:33.760 --> 00:42:34.880
That's Steven with a pH.
00:42:35.199 --> 00:42:37.119
Um, you can also find me on LinkedIn.
00:42:37.360 --> 00:42:37.840
Awesome.
00:42:38.000 --> 00:42:38.400
Awesome.
00:42:38.559 --> 00:42:47.360
Yeah, and obviously Crazy Egg is uh, you know, shout out to them and and you guys just wonder what you've been able to accomplish and uh continues to really lead the pack.
00:42:47.519 --> 00:42:48.719
So really appreciate it, Steven.
00:42:48.880 --> 00:42:49.360
Thanks again.
00:42:50.239 --> 00:42:50.719
Thanks so much.
00:42:50.880 --> 00:42:51.840
Awesome.
00:42:55.519 --> 00:42:58.000
Thanks for listening to Built to Convert.
00:42:58.159 --> 00:43:06.480
If you're serious about growth, don't just listen, subscribe, leave a review, and follow us for new episodes that drop straight into your feed.
00:43:06.639 --> 00:43:09.280
Until next time, keep building.
00:00:03.279 --> 00:00:10.880
You're listening to Built to Convert, the podcast where growth-minded marketers, founders, and creators come to level up.
00:00:11.119 --> 00:00:15.439
No hype, no guesswork, just bold insights that drive results.
00:00:15.679 --> 00:00:16.640
Let's go.
00:00:20.640 --> 00:00:27.600
Hey guys, welcome to Built to Convert, the podcast where we explore how strategy, analytics, and creativity come together to drive meaningful growth.
00:00:27.760 --> 00:00:30.399
I'm your host, Tyler Lane, founder of Session Interactive.
00:00:30.480 --> 00:00:36.799
And each episode we sit down with marketers, founders, and operators who are building real momentum inside today's most interesting companies.
00:00:37.039 --> 00:00:42.159
This episode is brought to you by Session Interactive, the performance marketing agency for brands that are serious about growth.
00:00:42.399 --> 00:00:55.759
Whether you're in B2B, SaaS, DTC, e-commerce, or anywhere in between, we help marketing teams cut through the noise of smarter SEO, GEO, AEO, all those acronyms, paid media, AI integration, and analytics that actually drive revenue.
00:00:55.920 --> 00:00:57.920
Learn more at Session Interactive.com.
00:00:58.399 --> 00:00:59.119
All right, guys.
00:00:59.280 --> 00:01:00.479
It's a fun one today.
00:01:00.560 --> 00:01:03.119
We've had a longstanding relationship with these guys.
00:01:03.200 --> 00:01:06.560
Uh we've got Steven No, director of growth marketing at Crazy Egg.
00:01:06.719 --> 00:01:19.040
Steven has spent his career on the go-to-market side of B2B software with stops at NIRA, PADL, and Profit Well before landing at a company that builds the exact tools marketers use to run experiments.
00:01:19.200 --> 00:01:20.239
He's talking about.
00:01:20.480 --> 00:01:22.319
It's a strange vantage point.
00:01:22.480 --> 00:01:31.840
He's not only doing CRO, he's selling infrastructure for it, which means he sees where testing culture actually breaks uh down inside real teams.
00:01:32.159 --> 00:01:33.120
Really exciting stuff.
00:01:33.280 --> 00:01:36.239
Steven's always been super progressive and always very generous with us.
00:01:36.480 --> 00:01:38.079
Stephen, I really appreciate your time, man.
00:01:38.319 --> 00:01:38.799
Absolutely.
00:01:39.040 --> 00:01:40.159
Yeah, thanks for joining us.
00:01:40.319 --> 00:01:43.840
Hey, we like to start out with these conversations with a little bit of kind of career origin.
00:01:43.920 --> 00:01:46.799
I know you just called out some of those companies that you were in.
00:01:46.879 --> 00:01:49.920
And hey, shout out to Heaton over at Crazy Egg2.
00:01:50.159 --> 00:01:54.400
He was, he was really um, you know, had some amazing things to say about you.
00:01:54.560 --> 00:02:02.079
Uh actually when Heaton and I first met a couple of years ago, and he was like, I think his quote honestly was like, Steven's the best marketer I've ever been around.
00:02:02.239 --> 00:02:02.959
So it was pretty cool.
00:02:03.040 --> 00:02:05.040
And that's it's high praise from Heaton, too.
00:02:05.280 --> 00:02:07.680
He's got uh he's got the notoriety in this space.
00:02:07.840 --> 00:02:18.080
But talk to me about kind of your experience, you know, as you you've matured in your career now, obviously director of growth marketing and such a reputable uh could CRO-related company.
00:02:18.159 --> 00:02:29.360
But talk to me about kind of how it shaped you and and certain things perhaps you've had to unlearn as you've you know, and then now as you've grown and developed into this uh growth marketing director at CrazyAgg.
00:02:29.919 --> 00:02:30.159
Yeah.
00:02:30.240 --> 00:02:34.240
So like a lot of people, I didn't start out wanting to be a marketer.
00:02:34.319 --> 00:02:36.080
I came through this through roundabout blends.
00:02:36.240 --> 00:02:38.639
In my particular case, I actually started out as a data analyst.
00:02:38.800 --> 00:02:42.159
So when I was in college, I had like data analyst um stints.
00:02:42.240 --> 00:02:45.520
I was working on the pricing algorithm at Wayfair, funnily enough.
00:02:45.680 --> 00:02:51.280
Um I was uh spent a little bit in finance, but I ended up coming back to the startup space, which I really loved.
00:02:51.439 --> 00:02:54.879
Um, you know, started out kind of doing a lot of revenue operations type stuff.
00:02:55.039 --> 00:03:00.560
I was the sales source admin, I was running all the numbers, I was like scraping all the leads for the sales team, you know.
00:03:00.719 --> 00:03:06.560
Um, and kind of when you're in startup, startup mode, just there's way too much work and not enough people.
00:03:06.800 --> 00:03:13.840
So kind of one day someone came up to me and said, Hey, you've been kind of like looking at the Google Analytics account for little nuts of information.
00:03:14.000 --> 00:03:15.840
Our paid media manager just left.
00:03:16.000 --> 00:03:17.599
You see much to do you want to take that?
00:03:17.759 --> 00:03:18.960
I said, Yeah, sure.
00:03:19.039 --> 00:03:23.039
I'm probably gonna make some screw ups along the way, but you know, always an opportunity to learn.
00:03:23.199 --> 00:03:24.800
And it kind of snowballed from there.
00:03:25.199 --> 00:03:30.319
Went over, ended up taking on paper media, um, expanding to kind of wholesale demand generation.
00:03:30.400 --> 00:03:40.560
When I came over to ProvalWell, I ended up taking over a lot of a business development team as well, as well as a lot uh doing a lot of like product marketing, um, conversion and optimization, copywriting, things like that.
00:03:40.800 --> 00:03:46.479
And so kind of I've always been very much a jack of all trades, but the thing I think I'm really good at is kind of two things.
00:03:46.639 --> 00:03:54.960
Number one is pattern recognition, especially when it comes to looking at data and kind of figure out like, okay, what can we learn from a given signal such that we can put that into action?
00:03:55.120 --> 00:04:00.479
And the second thing I'm probably really good at ultimately is connecting dots between different domains.
00:04:00.719 --> 00:04:12.240
I can kind of like both take that quantitative skill set and look at the numbers and figure out what this means, but also work on cop writing, kind of relate back to, okay, what does this mean in terms of which messages, which offers do really well?
00:04:12.400 --> 00:04:15.039
And ultimately, what is it about our customers really want?
00:04:15.360 --> 00:04:33.680
Um, and so coming over to Crazy Egg, it's kind of been a perfect opportunity just because I can kind of not only do that for Crazy Egg while we're dealing with kind of all the crazy change we're seeing in the world right now, but on top of that, I can provide some of that consultative skill set for our customers as well, which means I can be helpful to as many people as possible.
00:04:34.000 --> 00:04:34.879
I love that, Steven.
00:04:35.120 --> 00:04:36.639
Again, your your perspective's been awesome.
00:04:36.720 --> 00:04:42.800
And I love it, just the ability to kind of sink your teeth into all those different things and get that level of exposure.
00:04:42.959 --> 00:04:56.959
It reminds me of my career, not to that extent, but just like you were able to just be in the weeds and therefore shape and get your give give your perspective or get a perspective that's far more uh robust, if you will, than a lot of others.
00:04:57.439 --> 00:05:06.319
Walk me through a specific moment when you realized the growth playbook that worked at maybe Profit Well, for example, wouldn't necessarily work at Crazy Egg.
00:05:06.480 --> 00:05:10.079
Was there certain things you're like, oh, I'm gonna transition this over into Crazy Egg?
00:05:10.319 --> 00:05:18.240
And you quickly realize perhaps you're like, no, I need to maybe pivot separately of more of the subscription-based um component that Crazy Egg offers.
00:05:18.480 --> 00:05:20.959
You know what's really funny is um, you know, Prof.
00:05:21.199 --> 00:05:23.040
Well was a multi-product company.
00:05:23.120 --> 00:05:27.120
So kind of some parts of a company were already firing all cylinders by the time I got there.
00:05:27.279 --> 00:05:34.000
Um the Prop Well metrics analytics products is definitely one example of kind of like just a great freemium analytics product.
00:05:34.160 --> 00:05:47.439
Uh, we had our price intelligence elite um kind of service and offering there as well, which, you know, um Patrick Campbell, our old CEO, um, you know, very well known in the business uh of SaaS, at least as far as like modernization strategy, and that was already firing.
00:05:47.600 --> 00:05:56.240
Um, the kind of one win which I could potentially own is our Proper Retain product, which was essentially an automated retention marketing product.
00:05:56.480 --> 00:06:13.920
And what was great is kind of before we ended up being acquired, we ended up building a really replicable outbound business development strategy, um, where we basically came in and we said, hey, like we have this great automated retention marketing uh platform, and we believe in it so much that here's what we're gonna do.
00:06:14.160 --> 00:06:20.959
If, you know, um if you uh install this product, we're gonna run an A-B test at no charge.
00:06:21.199 --> 00:06:28.079
If we can measurably beat the control um group on this trial, then we're gonna um measure that uplift.
00:06:28.160 --> 00:06:31.519
We're gonna take a little bit of that for ourselves, but you get to keep most of it.
00:06:31.759 --> 00:06:34.879
And uh and if we don't beat it, it's no skin off your back, no risk.
00:06:34.959 --> 00:06:35.759
What do you say?
00:06:36.000 --> 00:06:40.399
And as it turns out, lots of people loved it, even big companies.
00:06:40.639 --> 00:06:48.319
Um, and so what I kind of realized about like, hey, this is a pretty unsexy kind of monetization lever, which is retention.
00:06:48.480 --> 00:06:57.360
But turns out when you have an offer that's really, really good and it comes down to brass tags pretty quickly, then we were able to replicate that and scale that really, really quickly.
00:06:57.519 --> 00:07:02.639
Um, you compare that to Crazy Egg right now, which is a bit more of a mature category, right?
00:07:02.800 --> 00:07:06.240
Website analytics, you know, Crazy Egg's been around since I think 2004.
00:07:06.480 --> 00:07:10.560
Heaton, you know, um and his co-founder like invented category way back in the day.
00:07:10.639 --> 00:07:12.480
So, in a sense, we're doing a refounding.
00:07:12.639 --> 00:07:22.560
And we're doing that in the context of just the entire AI shift, the entire jobs to be done of website analytics of experimentation is changing completely right now.
00:07:22.720 --> 00:07:29.439
So, frankly, I'm actually doing a lot more product stuff than I used to recently in terms of, hey, this is a marketing analytics product.
00:07:29.600 --> 00:07:34.879
This is something that I would have considered super cutting edge back in 2004.
00:07:35.120 --> 00:07:42.240
What we want in when it came on in 2024 is okay, what is cutting edge now looks totally different.
00:07:42.480 --> 00:07:46.720
So if I was using this as a marketer, what would I consider cutting edge?
00:07:46.800 --> 00:07:55.920
What's gonna actually do the job for me such that I could go to other markers and tell them with conviction, hey, this is really great for me, and I know what's gonna be really great for you.
00:07:56.879 --> 00:07:57.680
Love that, Steven.
00:07:57.839 --> 00:08:01.519
Yeah, amazing stuff to just and perspective again, you're you've been learning.
00:08:02.000 --> 00:08:07.920
Let's transition here into CRO specifically and analytics and kind of the experimentation discipline, if you will.
00:08:08.319 --> 00:08:17.199
Um you've written about the four-step approach to the first 90 days of a CRO engagement that starts with tools teams already have, right?
00:08:17.279 --> 00:08:21.279
And not necessarily requiring net new purchases for particular tools.
00:08:21.360 --> 00:08:30.000
What are some of the most common uh things a team discovers in the f that first pass that they have completely blind to before, perhaps?
00:08:30.240 --> 00:08:30.720
I'm curious.
00:08:31.040 --> 00:08:31.199
Right.
00:08:31.439 --> 00:08:41.600
I think you know, the funny part is, you know, data people like me often we get distracted by data because we love numbers, we feel like it's super robust, we think that, oh, this is a cutting edge thing.
00:08:41.759 --> 00:08:44.559
And so it becomes down to like, okay, what do these numbers say?
00:08:44.639 --> 00:08:46.799
Um, what is the pattern, all this stuff?
00:08:47.120 --> 00:08:53.200
But if we're talking as marketers, and ultimately like marketing comes down to customers, right?
00:08:53.360 --> 00:08:54.639
What is the customer experience?
00:08:54.720 --> 00:08:57.120
What do they want from us and what gets them actually by?
00:08:57.440 --> 00:09:08.559
And what's funny is that, you know, I could spin one analysis saying like the average, um, or maybe differently, the most popular element on this page is this button.
00:09:08.799 --> 00:09:15.519
But another way you could spin it is to say this button represents the most popular offer or product on our website.
00:09:16.159 --> 00:09:17.440
Completely shifting it around.
00:09:17.519 --> 00:09:35.279
I'm taking and taking metrics and a lot of this statistical jargon out of it because ultimately statistical jargon can be helpful for people who are more specialized to work with, but if you're removing the kind of human perspective of a customer out of a lot of this, then a lot of your analysis is going to be focusing on the wrong things altogether.
00:09:35.519 --> 00:09:40.960
So I feel like a lot of uh companies, when they're coming in, they already have data to work with.
00:09:41.200 --> 00:09:46.399
But what they need to be coming in through is not just someone who has those like data analytics chops.
00:09:46.559 --> 00:09:59.360
They can also relate that to the ICP, what the target customer actually wants, and then ask basically, okay, what would that, what does this reflect about who our customer is, what they actually want, and how can we optimize based off of that as well?
00:09:59.600 --> 00:09:59.840
Yeah.
00:10:00.000 --> 00:10:01.279
So that's number one.
00:10:01.360 --> 00:10:07.200
And I think ultimately, like we also get too distracted by the kind of brass tacks work, right?
00:10:07.360 --> 00:10:16.159
It's, you know, like I spent so many hours of my life kind of like instrumenting, setting up our like, you know, Salesforce to HubSpot connector and stuff like that.
00:10:16.399 --> 00:10:20.480
But then, you know, when I ultimately bring this back to like, hey, what does the customer want?
00:10:20.639 --> 00:10:26.159
Turns out a lot of your metrics, you should collect the data anyway, because you may need it in the future.
00:10:26.240 --> 00:10:31.200
But ultimately it comes down to like, hey, um instrumenting is not the most important thing.
00:10:31.360 --> 00:10:35.840
The most important thing is number one, figure out which part of a page people are actually scrolling down to.
00:10:36.000 --> 00:10:38.799
And as it turns out, a lot of the time, people barely scroll down.
00:10:38.960 --> 00:10:42.720
So you got to get the right stuff at the top of a page, get the right offer in there.
00:10:42.879 --> 00:10:52.559
And I really figured that it's like, hey, get the right offer, because you know, the funny part is every time you go to say, you know, I've been to a bunch of like HopSpot inbound conferences in my life.
00:10:52.720 --> 00:10:59.360
And the example they would always give is, hey, we changed this button color, and that increased conversions by X percent.
00:11:00.080 --> 00:11:04.559
No one's ever seen a big increase in conversions from changing a button, right?
00:11:04.879 --> 00:11:10.159
A button placement can make sense because you just want to make that convenient and make that obvious.
00:11:10.399 --> 00:11:14.480
But changing a color doesn't make sense, but changing your offer absolutely does.
00:11:14.639 --> 00:11:23.519
So it's not just focusing on your customer, but also thinking about like, okay, instead of thinking of it as like optimizing every little lever, think about which levers make the most sense.
00:11:23.679 --> 00:11:33.679
And ultimately, a lot of the time this boils down to like two to five levers, a lot of the time, in terms of actually getting your customers to kind of see your offer and decide whether or not it's the right one for them.
00:11:35.600 --> 00:11:36.799
A lot to unpack there.
00:11:36.879 --> 00:11:48.879
I'm curious um around the, you know, obviously Crazy Egg, you know, well-known, very established, as you mentioned, Heaton really determined or really define the category.
00:11:49.360 --> 00:11:51.840
2004 brings me back for sure.
00:11:52.159 --> 00:11:55.440
Um you guys are running your own tests internally.
00:11:55.840 --> 00:12:10.720
Talk to me kind of maybe that first year or so in your role at Crazy Egg and some tests that you were you ran with the team and had some wins, but I'd also like to hear about maybe some some failures or certain things.
00:12:10.879 --> 00:12:23.519
Because I think this is a part when I not to talk about myself here, but when we our our clients, we want that rapport, that personal rapport, but we also want to prove ourselves through performance rapport.
00:12:23.679 --> 00:12:24.720
That's what I always say.
00:12:25.039 --> 00:12:39.039
But it's ultimately that relationship that they can trust us to essentially fail forward, to go out and test and to know that things aren't going to be perfect all the time, but we've got a strong conviction, we've got a hypothesis to move forward.
00:12:39.120 --> 00:12:48.080
So I'm curious, yeah, what what maybe some certain things that have worked well on Crazy Egg specifically, and maybe some things that were ultimately a massive failure.
00:12:48.480 --> 00:12:48.960
Yeah.
00:12:49.279 --> 00:13:02.320
So surprise, surprise, as soon as I came to Crazy Egg, the first thing we started doing is not only collecting heat maps on, you know, our homepage, our pricing page, I ran so many ABU tests, like so many different headlines, so many different things.
00:13:02.399 --> 00:13:13.200
And the funny part is that kind of um we ended up finding that for the existing product, that a lot of the existing kind of visuals and a lot of the same messaging already worked, right?
00:13:13.360 --> 00:13:17.279
Or at least um the test didn't improve what was already working.
00:13:17.519 --> 00:13:25.600
Um, but the thing is that you know your data is always gonna is always going to be biased towards the people who you're measuring or the people you're hearing from.
00:13:25.840 --> 00:13:29.039
And so I ended up jumping on some of these customer calls also, right?
00:13:29.120 --> 00:13:32.080
To um kind of figure out like, okay, what do people actually want?
00:13:32.240 --> 00:13:33.200
What's their experience been?
00:13:33.440 --> 00:13:34.240
How can we improve it?
00:13:34.399 --> 00:13:36.960
Those kind of like product um feedback calls.
00:13:37.279 --> 00:13:48.799
And the way that kind of these calls always went, right, is people we are talking about who are people who are already paying customers for crazy egg, who installed these heat maps on their page and are like getting a lot of data, right?
00:13:48.960 --> 00:14:03.519
And so we pull up a he map report in their account and kind of describe like, okay, this button's getting the most clicks, people are scrolling down this far, uh, people are navigating to this URL the most, and they say, okay, great.
00:14:04.240 --> 00:14:05.039
What do I do?
00:14:05.440 --> 00:14:06.399
Yeah, totally.
00:14:06.720 --> 00:14:10.720
Which is the classic kind of cliff almost of analytics, right?
00:14:10.799 --> 00:14:19.360
Which is like, hey, you can just most analytics are descriptive, but the real job to be done is just like tell me what to do to increase that performance.
00:14:19.600 --> 00:14:28.559
And historically, Betts needed someone like me to actually have the time to kind of like look at the data, hope that there's actually a pattern that's clear, and then make that jump.
00:14:28.799 --> 00:14:32.639
And as it turns out, is you know, we have kind of like a free trial model.
00:14:32.720 --> 00:14:38.000
Um, we promise that people can um kind of learn how to optimize their websites based off of it.
00:14:38.159 --> 00:14:55.600
And then we have people signing up, starting a free trial, going into the product, seeing the data, and still having that question of like, I can't justify kind of continuing to work with product when I, you know, uh some we have some people who are like full-time day alists, we have a lot of people who are like, you know, um small business owners, right?
00:14:55.759 --> 00:15:01.440
Who just want to get the answer because they need to go on to running the the freaking business, right?
00:15:01.600 --> 00:15:06.720
And they're telling me, like, I can see the data, I just don't know what I'm what it's supposed to mean or what I'm supposed to do with it.
00:15:06.879 --> 00:15:15.360
And so that kind of taught me like, oh, the job to be done, there's a there's a uh very clear job to be done that is still not being solved, right?
00:15:15.440 --> 00:15:29.600
And kind of if you think, imagine that kind of like the end result in terms of like improving engagement and uh reducing bounce rate of increasing conversions, that's a very end, but there's a whole bunch of stepping stones in between, and there's some pretty clear stepping stones that we're not solving here.
00:15:29.759 --> 00:15:36.639
And if we can't prove that, that's the most concrete thing that we're not showing in terms of proving that our product is doing what it's supposed to do.
00:15:36.879 --> 00:15:51.200
So we kind of went back to the drawing board and we're like, you know, if only there was a technology right now that could analyze the data and kind of, you know, find the patterns and all this and all this information and then make some recommendations or come up with some ideas.
00:15:51.360 --> 00:15:53.279
Um, and as it turns out, that exists now.
00:15:53.440 --> 00:16:04.080
So we kind of thought to, okay, what is the tool that I would want to do that job for me so that I can just go ahead and start running those tests that I know are actually gonna be um higher leverage.
00:16:05.519 --> 00:16:15.600
Yeah, it's it's uh and you're obviously we'll we're gonna we'll transition into kind of the the analysis piece that you guys have built out with crazy egg because I think it's wildly impressive.
00:16:16.720 --> 00:16:19.919
I love the um I think everyone does.
00:16:20.159 --> 00:16:25.200
I I told this story with Tahiton a number of years ago that I was on a plane and I had Crazy Egg up.
00:16:25.440 --> 00:16:32.399
This is a long time ago, and it was you know how sexy things look, you know, with the confetti in particular, one of my favorite reports by far.
00:16:32.720 --> 00:16:35.440
And I remember seeing the guy next to me, he's like, What is that?
00:16:35.519 --> 00:16:36.000
That's amazing.
00:16:36.080 --> 00:16:36.720
You gotta bring that.
00:16:36.879 --> 00:16:38.480
He's like, I'm going to Europe right now.
00:16:38.559 --> 00:16:41.279
I'm in marketing, we've never seen anything like that.
00:16:41.519 --> 00:16:48.879
And it was so funny because it does, it is that very attractive, uh compelling insight.
00:16:48.960 --> 00:16:55.200
But at the end of the day, or just kind of a view, and at the end of the day, it's like, what do I do with that information?
00:16:55.440 --> 00:16:59.840
And I think you're you're spot on in terms of the same thing with GA, right?
00:17:00.000 --> 00:17:02.559
And okay, I've configured some certain things.
00:17:02.799 --> 00:17:03.919
Okay, it tells a piece.
00:17:04.000 --> 00:17:12.319
And it isn't, uh I'm curious your thoughts and how and on how we uh articulate it to our clients and our prospects is like, great, GA is a portion of it.
00:17:12.400 --> 00:17:19.119
Now, layering on GTM with custom events, conversion tracking elements, uh, et cetera, that's more of it.
00:17:19.200 --> 00:17:26.640
Now getting a uh heat mapping software, click tracking software like Crazy Egg in particular, that's more of the pie in terms of the view.
00:17:26.799 --> 00:17:34.960
But we're still missing a lot of the qualitative insights, like survey-related technology or audience actual information, still more of it.
00:17:35.279 --> 00:17:39.920
Never perfect necessarily, but I think with a lot of those kind of directionally accurate.
00:17:40.000 --> 00:17:42.559
What are your thoughts in terms of that setup?
00:17:43.039 --> 00:17:52.240
What I've just started, which I mentioned, versus a lot with a lot of brands of like we've got basic tracking within GA, some within GTM, and now we're doing heat mapping.
00:17:52.400 --> 00:17:54.480
Are they limiting their cell themselves?
00:17:54.720 --> 00:17:57.279
But they have enough insight to move on.
00:17:57.519 --> 00:17:59.599
Um, talk to me about that.
00:18:00.960 --> 00:18:05.119
I would say that you know, the more data you have almost always the better.
00:18:05.359 --> 00:18:17.599
Um, what's really funny these days is that it used to be that kind of people talk about, oh, all your data is in these different silos and you can't compare it, so you gotta all put it into that one central CRM in order to compare all of it.
00:18:17.839 --> 00:18:21.759
And, you know, inevitably this conversation always goes to AI.
00:18:21.839 --> 00:18:32.640
But one funny use case that I found really good with AI is its ability to kind of actually analyze data that's in different silos that's not necessarily comparable one-to-one and actually relate these things.
00:18:32.880 --> 00:18:41.680
So if I have data that's you know, um in a CRM in one thing, and then I have behavioral data in another that can tell me like which elements are coming in.
00:18:42.079 --> 00:18:46.319
Ordinarily, a human analyst would say these are not kind of apples to apples comparisons.
00:18:46.480 --> 00:18:49.200
And so you kind of have to analyze things, things separately.
00:18:49.359 --> 00:18:59.680
But what I found helpful to do when uh an analyze with an AI is to just say, hey, like look through these data repositories and find me kind of patterns which are matching in these.
00:18:59.839 --> 00:19:05.359
And they is very often able to say, like, hey, this particular headline or this particular offer is doing really well.
00:19:05.519 --> 00:19:11.200
And I find in your CRM that this uh the this skew is being poached the most often by our customers.
00:19:11.359 --> 00:19:12.799
And these things are paralleling.
00:19:12.960 --> 00:19:28.480
Now, what's interesting is uh ordinarily BA would require a kind of um a really rare kind of human data analyst who can both look at numbers, look for trends, but also kind of relate, like, hey, buying the most of this product and relating the most of this offer, there's something qualitative that's in common here.
00:19:28.640 --> 00:19:33.039
That's kind of requiring you to cross both qualitative and quantitative data.
00:19:33.119 --> 00:19:35.759
But AI is actually surprisingly good at doing these.
00:19:35.920 --> 00:19:41.359
So I would say that having siloed data is not necessarily as big of a problem as it used to be.
00:19:41.440 --> 00:19:47.519
And certainly, if you're starting from zero, working with data silos can still get you much better than nothing.
00:19:47.759 --> 00:20:01.039
But if you're um a very high traffic brand, if you're trying to optimize for many different um products and many different segments, then analyzing this stuff becomes much more of a scalpel than a sledgehammer kind of um problem.
00:20:01.200 --> 00:20:08.240
And that's the point at which you start to need a lot more kind of like um data or previously siloed data that's talking to each other.
00:20:08.880 --> 00:20:09.039
Sure.
00:20:09.359 --> 00:20:09.599
Yeah.
00:20:09.839 --> 00:20:11.440
Yeah, and easier said than done too.
00:20:11.519 --> 00:20:18.079
I think there's always just like, oh, you know, that's uh I love the aspect of you know, crazy egg is so easy to get going.
00:20:18.640 --> 00:20:30.000
But I think in certain cases it's it's it's not necessarily helpful because some of these brands just haven't configured their GA or GTM instance to a place where Crazy Egg is already at.
00:20:30.880 --> 00:20:31.039
Yeah.
00:20:31.200 --> 00:20:55.680
And maybe one more thing to add is one maybe other big reason to kind of like um simplify your tech stack ultimately is because ultimately the more complexity you have your operation, be that, you know, like um, you know, kind of a lot of stakeholders or a lot of daily silos, is every additional silo um or a kind of layer of complexity adds more uncertainty to any given decision, which makes it harder to kind of like have any conviction around any of its stuff.
00:20:55.920 --> 00:21:04.000
So, you know, um certainly we have a lot of folks who use Crazy Egg in conjunction with a separate A-B testing tool or, you know, GA4 or things like that.
00:21:04.079 --> 00:21:09.839
And if that's your setup, then we are more than happy to kind of find a way to, you know, solve a certain point solution for you.
00:21:10.000 --> 00:21:24.240
But one advantage that a lot of our customers who do mainly use us for a lot of these different things is they have one source of truth for a lot of these things, even if it's different features, and therefore they kind of have a lot more conviction when it comes to making a decision fast.
00:21:24.480 --> 00:21:34.880
And these days it turns out that um information less and less is a barrier, but having conviction to actually take an action is really a thing that takes more time these days.
00:21:35.119 --> 00:21:35.920
Yeah, I love that.
00:21:36.000 --> 00:21:39.759
Just it's uh I couldn't have said it better myself, just the simplicity aspect.
00:21:39.920 --> 00:21:49.680
It reminds us, it reminds me of just the countless audits we've done of like a GTM instance, and there's just like inflated events all over the place.
00:21:49.839 --> 00:21:52.160
And oh yeah, that was configured a couple of years ago.
00:21:52.319 --> 00:21:53.279
No one really looks at that.
00:21:53.359 --> 00:22:01.759
It's like, okay, your historical event data is massively skewed, and we want to marry this with our heat mapping data and Unfortunately, it's not great.
00:22:01.920 --> 00:22:05.279
So and the the pages have changed ridiculous amount.
00:22:05.519 --> 00:22:16.880
Or honestly, we've we've actually onboarded clients in the past too that have Crazy Egg, for example, are tracking way too much, you know, and are like, this isn't I'm like what what benefit is this?
00:22:17.119 --> 00:22:24.559
So it is kind of really ratcheting it back and and uh honing in a more simplistic infrastructure before you expand.
00:22:24.880 --> 00:22:25.119
Yeah.
00:22:25.200 --> 00:22:30.960
I mean ultimately these days, you know, uh frankly, I think the days of tracking too much are kind of over.
00:22:31.119 --> 00:22:35.279
Um the funny part is that, you know, right, uh peep it's it's funny.
00:22:35.359 --> 00:22:40.480
I think 10, 15 years ago, people were going on about like, oh, data is the new oil, you know, it's gonna change everything.
00:22:40.640 --> 00:22:42.960
And but it managed to the bottleneck I mentioned earlier, right?
00:22:43.039 --> 00:22:49.519
It's like ultimately you need a guy like, you know, um young me to look actually look at this stuff and figure out what's gonna go on.
00:22:49.680 --> 00:23:02.799
And turns out that kind of what we might call a reasoning or intelligence bottleneck is no longer there because now you can have kind of a machine crawl through that data and find it for you, even to find patterns which you didn't even realize were bare.
00:23:03.119 --> 00:23:14.720
So, in a in a sense, kind of like that uh that old pattern of like, hey, we have more data than we know what to do with, so we're just kind of gonna throw it in a lake and then wait for a later data until we have someone who can actually take a look at it.
00:23:14.880 --> 00:23:18.400
Those days are over because like analysis is no longer a bottleneck.
00:23:18.480 --> 00:23:24.319
But figuring out what are we actually gonna do with that, that is kind of where we start to run into more bottlenecks these days.
00:23:24.559 --> 00:23:26.160
Yeah, excellent point, excellent point.
00:23:26.240 --> 00:23:33.039
And I think still, like, yeah, it's not a bottleneck, but I think it still does require a experienced mind too.
00:23:33.200 --> 00:23:34.559
Like you can get a lot of great insights.
00:23:34.640 --> 00:23:42.640
And look, we're gonna talk about some of the features, obviously, of Crazia here in a second, but it still requires like at least some a background a bit, right?
00:23:42.720 --> 00:23:47.440
And and some directional analysis, especially prompt structures, um leveraging AI.
00:23:47.680 --> 00:23:56.880
Steven, hey, you know, big fan of Crazia, obviously, and what you guys have been able to accomplish, big fan of yours and Heaton and and the rest of the team.
00:23:57.119 --> 00:24:01.519
Um, you know, it's a it's a saturated market in certain cases, right?
00:24:01.599 --> 00:24:05.920
You mentioned Heaton really starting the category, as I mentioned already in 2004.
00:24:06.079 --> 00:24:11.440
Um, Hotchard, Lucky Orange, Microsoft Clarity, the list goes on in certain things.
00:24:11.599 --> 00:24:14.559
I know Google at one point had their you know in-page analytics.
00:24:14.640 --> 00:24:15.599
That was pretty bad.
00:24:15.759 --> 00:24:18.799
Um, I'm surprised they didn't try to capitalize on that.
00:24:19.279 --> 00:24:25.359
You guys have made some significant strides um in the last year plus and continue to make those.
00:24:25.599 --> 00:24:26.400
Let's talk about that.
00:24:26.559 --> 00:24:30.720
I mean, I I love, you know, I love what you're able to do, obviously, on the A-B testing component.
00:24:30.799 --> 00:24:45.359
We're we're we've been alluding to it, but more of the AI rollout component of being able to empower individuals to become more of, as you mentioned before, of like their own professional services arm to analyze and make decisions on certain things.
00:24:45.599 --> 00:24:51.200
Talk to me, talk to us about that and how you see kind of the future of heat mapping and particularly around Crazy Egg.
00:24:51.680 --> 00:24:51.920
Yeah.
00:24:52.079 --> 00:24:55.599
So, I mean, obviously we have a lot of, you know, there's a lot of choice in the market, right?
00:24:55.680 --> 00:24:58.000
You've got your hot jars, you have your content squares.
00:24:58.319 --> 00:25:08.400
Um, and kind of whenever you look at a lot of these things, right, is, you know, every company starts very often starts out with one particular kind of role that it's supposed to play.
00:25:08.559 --> 00:25:22.000
And if we look at a hot jar content square kind of traditional crazy egg, what that role served was uh providing a data analyst with a product that will give them data and the ability to produce reporting, right?
00:25:22.160 --> 00:25:24.079
The end goal is reporting.
00:25:24.799 --> 00:25:28.720
And, you know, I when I was younger, kind of I that's the way I thought of my job as well.
00:25:28.880 --> 00:25:29.680
But like I just said, right?
00:25:29.759 --> 00:25:36.240
It's like ultimately, hey, it's like to drive more measurable uh impact, to drive more revenue, to make a better customer experience.
00:25:36.400 --> 00:25:41.519
So as it turns out, um, the way I see it these days is reporting is not the end goal.
00:25:41.759 --> 00:25:50.240
Um, and so when we looked at our competitors and we see the way that like what's capable with technology, what's uh what you can achieve with technology is expanding.
00:25:50.400 --> 00:26:03.680
But we look at a hot jar and a content square, and they're ultimately just um producing a lot of stuff that accelerates the job of reporting, but doesn't accelerate the real end result which you're trying to achieve, which is better business outcomes, better customer experiences.
00:26:04.000 --> 00:26:05.920
We think that falls a little bit short.
00:26:06.160 --> 00:26:19.519
And it's kind of an inert, they're kind of falling into an inertia that a lot of companies have, which is like going deeper into that same job we've done when really we should be going back to the drawing board and say, like, hey, what is the thing that we should be trying to achieve here?
00:26:19.759 --> 00:26:30.640
So the way we've gone back to the drawing board is kind of thought like, okay, we already have you know analytics in there, and we can slowly apply AI kind of in a very obvious way to bring insights out of that.
00:26:30.799 --> 00:26:32.559
We have an A-B testing tool.
00:26:32.799 --> 00:26:47.200
But um, and on top of that is, you know, we obviously kind of the capability to kind of generate copywriting really, really quickly is obviously out there, but but let's not get into that because kind of there's there's good there's good ways to do it.
00:26:47.279 --> 00:26:49.759
There's obviously a lot of terrible ways to do it, right?
00:26:50.160 --> 00:26:59.440
But ultimately, the way we thought about it is like, hey, if the end goal is to unify a lot of these jobs to be done and then take you to like, hey, how do I improve my customer experience?
00:26:59.680 --> 00:27:12.480
We're obviously gonna produce some of these analytical tools so that you know a marketing analyst or a marketer can kind of get in there, um, produce a report really quickly that they can take to their you know team and say, like, hey, here's what's happening.
00:27:12.640 --> 00:27:14.240
Let's discuss what we want to do.
00:27:14.400 --> 00:27:26.400
But you know, a marketer who's trying to kind of be a jack of all trades can come in, look at the data, immediately get a bunch of uh ideas for A-B tests or personalization, even that they can readily really quickly.
00:27:26.960 --> 00:27:32.799
And this is gonna get to a point where you can launch this with just a few clicks, or maybe even one click in the future.
00:27:33.039 --> 00:27:43.680
Um, and you know, the way that these things are going so quickly, you know, like we're obviously working on like, you know, shipping new dashboards, new forms of reporting for this as well.
00:27:43.839 --> 00:27:47.599
But you know, we're we've been working on audience targeting for A-B testing the last couple months.
00:27:47.759 --> 00:27:48.880
That's already live.
00:27:49.119 --> 00:27:53.920
Potentially, you know, by the end of this year, when's this gonna be full-blown personalization?
00:27:54.079 --> 00:28:00.160
But what really ultimately boiling down to is like, hey, let's go beyond just optimizing a website for everyone.
00:28:00.319 --> 00:28:04.960
What if we could deliver the right message to the right person at the right time?
00:28:05.200 --> 00:28:06.720
That technology exists now.
00:28:06.799 --> 00:28:11.359
So we should go and build that because ultimately I think that's where the puck is headed.
00:28:11.519 --> 00:28:14.079
And we shouldn't be thinking of what's possible three months down the line.
00:28:14.160 --> 00:28:16.880
We should be thinking what's going to be available two, three years down the line.
00:28:17.039 --> 00:28:17.279
Yeah.
00:28:17.359 --> 00:28:17.759
Yep.
00:28:18.079 --> 00:28:18.799
Personalization.
00:28:18.880 --> 00:28:21.279
How how many times have we heard that over the years, Steven?
00:28:21.519 --> 00:28:28.880
And I remember like uh, I forget, we were working with some.NET um CMSs, and they're like, oh, we do a ton of personalization.
00:28:29.039 --> 00:28:30.319
It was USL.
00:28:30.720 --> 00:28:31.680
Total bullshit.
00:28:31.759 --> 00:28:39.759
You know, it's like no, I mean, I've launched a lot of ABM campaigns in my life where we were uh, you know, using a variable to throw the customer's logo in there and stuff like that.
00:28:39.839 --> 00:28:42.400
And sure, that's technically personalization, right?
00:28:42.720 --> 00:28:44.480
Um, but that's service level.
00:28:44.720 --> 00:28:46.400
What and like I mentioned earlier, right?
00:28:46.480 --> 00:28:51.680
It's like it's one thing to change a headline, it's another to use a different value prop or a different offer.
00:28:51.839 --> 00:28:59.839
Um, and the challenge being that, like, well, with the current tools right now, you can certainly kind of like write five different variants of a headline and test these out.
00:29:00.000 --> 00:29:13.839
But what if, for example, the AI could look at what your um customer is track coming from as far as like a UTM parameter or you know, your ad headline, for example, and figure out like, oh, they want offer A, let's give them offer A.
00:29:14.160 --> 00:29:18.319
And then Ideally can do that like much faster, even programmatically.
00:29:18.559 --> 00:29:32.799
Um, you know, I live in it's weird to kind of like live in these kind of like weird tech worlds nowadays, but it's kind of like when you think beyond just automating a workflow to what does this mean for my actual customer experience, it starts to get pretty tetty.
00:29:33.279 --> 00:29:34.160
Yeah, yeah.
00:29:34.319 --> 00:29:35.440
Yeah, that's hilarious.
00:29:35.519 --> 00:29:39.359
That example of this uh infusing or including the logo.
00:29:39.519 --> 00:29:40.880
That's a classic.
00:29:41.200 --> 00:29:57.039
Uh let's talk just, you know, um, you mentioned the copywriting component, you know, we could talk about that for a long time, but how what how do you bridge the gap or just letting AI interpret behavioral data versus requiring a human to sign off?
00:29:57.200 --> 00:29:59.279
What is your thought process there?
00:29:59.759 --> 00:30:04.319
Um and opinion, I guess, there right now.
00:30:04.960 --> 00:30:05.440
Definitely.
00:30:05.599 --> 00:30:09.519
So I mentioned, you know, that one of the things I'm good at is pattern recognition, right?
00:30:09.599 --> 00:30:18.400
And I think ultimately a lot of data analysis, it doesn't numbers and running quantitative stuff is one way to do it, but qualitative is there as well.
00:30:18.559 --> 00:30:20.319
So what is the thing that unifies all of this?
00:30:20.400 --> 00:30:26.319
I think pattern recognition, frankly, is the if you would boil down the main job of data analysis.
00:30:26.799 --> 00:30:30.960
And, you know, when I I think the best example of this was actually when I was working at Wavefair.
00:30:31.119 --> 00:30:33.759
I was thinking I was there about a year before they IPO'd, right?
00:30:33.920 --> 00:30:40.000
Millions of products, hundreds of millions of pay of, you know, um, of pages and sessions, right?
00:30:40.160 --> 00:30:43.839
Do you think I as a data analyst, even back then, it's gotta be a lot bigger now?
00:30:44.000 --> 00:30:54.799
You know, if I were a looking at this in the most excruciating depth, I would be watching every single visitor, kind of taking notes on every single person, want to be clicking on all this stuff.
00:30:55.039 --> 00:30:56.160
But I don't do that.
00:30:56.400 --> 00:31:05.359
I run some queries, I take some uh take some averages, some counts, some sums, I do some segmentation analysis.
00:31:05.599 --> 00:31:11.759
All that really boils down to is simplifying data in such a way that a human being can actually process what's going on.
00:31:11.920 --> 00:31:13.839
Because ultimately, we human beings, we're limited.
00:31:13.920 --> 00:31:15.759
There's only so much information we can process.
00:31:15.920 --> 00:31:19.200
And if you look at a heat map, that's actually what's happening there as well, right?
00:31:19.279 --> 00:31:25.599
We are aggregating a whole bunch of click data and representing it in the visual format so that a human being can actually spot transfer.
00:31:26.400 --> 00:31:35.599
Whenever you simplify information, you're getting rid of some um information, getting rid of some signal in there because you're trying to also screen out noise.
00:31:36.000 --> 00:31:41.200
But the thing about AI is that you can actually throw a lot more processing power at the problem.
00:31:41.279 --> 00:31:49.920
And as a result, it can actually go through, check out every single click, and be like, hey, you know, um people from Indiana are buying a ton of this given product.
00:31:50.160 --> 00:31:56.240
Maybe we should be targeting uh more of more campaigns and offers around this particular product to folks in Indiana, right?
00:31:56.480 --> 00:32:01.599
That's a level of specificity, but you can only provide if you had previously a ton of human data analysts.
00:32:01.759 --> 00:32:06.240
But now you can actually have AI go through and look for a lot of that pattern recognition as well.
00:32:06.640 --> 00:32:16.480
And so the way I think about AI nowadays is that like that pat that job of pattern recognition, in a sense, frankly, I think a lot of the kind of like AI hype is a little bit overblown.
00:32:16.720 --> 00:32:26.960
AI, to me, first and foremost, is a master pattern recognition machine, which sounds like I'm downplaying what it's capable of, but actually that's a pretty world-changing capability.
00:32:27.359 --> 00:32:37.119
And so if you can use AI to like, hey, find me patterns in this engagement data and find me opportunities to sell more of this given product, it can find a lot more of that as well.
00:32:37.359 --> 00:32:45.039
Now, there's still room for human judgment in there, which say, hey, give me a bunch of ideas, um, which optimizations are on brand for us.
00:32:45.200 --> 00:32:49.839
But I can, as a human being, can look at it and say, oh, sure, those are the five most obvious ideas.
00:32:49.920 --> 00:32:53.200
But here's ideas six, seven, and eight that no one's thought of yet.
00:32:53.359 --> 00:32:55.119
Let's throw that into the mix as well.
00:32:55.359 --> 00:33:00.880
So, in a sense, it's going beyond just like kind of the roach um workflow automation.
00:33:01.039 --> 00:33:14.559
And when it comes to um performing jobs which are basically replicating and recognizing patterns, it's doing a ton of that as well, which means we human beings can kind of accelerate that process and get to the part where we apply our judgment about what's important.
00:33:15.119 --> 00:33:16.240
Yeah, really well said.
00:33:16.400 --> 00:33:17.039
Well said stuff.
00:33:17.119 --> 00:33:19.200
Pattern recognition is huge, obviously.
00:33:19.279 --> 00:33:30.960
And that's I think you you and I have just been relying on that for years because we've seen so many of these things and we can make some assumptions without having to look at all the data, but obviously now things are making it a little bit easier.
00:33:31.279 --> 00:33:31.839
Uh for sure.
00:33:32.000 --> 00:33:33.279
A lot easier, let's be honest.
00:33:33.920 --> 00:33:36.160
Definitely let's talk about some bets here.
00:33:36.559 --> 00:33:47.920
Um, I'm curious in terms of we talked about it before we hit record earlier in terms of just the um where behavioral analytics is is goes in the next two years.
00:33:48.480 --> 00:33:58.240
And where do you think um most of the industry industry potentially maybe disagrees with where it's heading versus maybe where you where you think are things are heading?
00:33:58.319 --> 00:33:59.920
Um I'm curious your thoughts.
00:34:00.240 --> 00:34:07.680
Yeah, I think, you know, um I think people are rightly uh kind of either burnt out or feeling the overhype of AI, right?
00:34:07.759 --> 00:34:11.119
Um, you know, I love on the East Coast, I believe you're an East Coaster too, right?
00:34:11.280 --> 00:34:18.960
Um, and you know, what's funny is, you know, I get exposed to lots of folks like Heaton, the Bay Area types who think that AI is kind of gonna change everything.
00:34:19.119 --> 00:34:20.639
Maybe they're right, maybe they're wrong.
00:34:20.800 --> 00:34:29.199
But when I talk to folks out here in the quote unquote real world, and most people are frankly using it as a glorified search engine, right?
00:34:29.599 --> 00:34:31.840
Um now I think that's gonna change.
00:34:31.920 --> 00:34:34.320
I think that's actually a generational change, right?
00:34:34.400 --> 00:34:45.119
Like um a lot of folks, you know, I'm a millennial, I still like to make large purchases on a desktop, but you talk to kind of younger kids, and I'm perfectly fine buying things on a mobile phone.
00:34:45.280 --> 00:34:49.440
And they're also the ones who are frankly using AI more to more on a day-to-day basis.
00:34:49.519 --> 00:34:59.679
So I think this kind of thing that's gonna creep up more so then, well, it's some things are gonna change very quickly, but then it's gonna make a kind of more macro change from the bottoms up.
00:35:00.719 --> 00:35:05.920
Um, but so I think kind of it's not gonna move quite as quickly as everyone thinks.
00:35:06.079 --> 00:35:08.320
But at the same time, again, like I mentioned, right?
00:35:08.400 --> 00:35:10.400
It I think it's still gonna change a ton.
00:35:10.559 --> 00:35:14.159
Um, I think when it comes to behavioral analytics in particular, right?
00:35:14.320 --> 00:35:24.639
The capability of bridging that gap from just collecting that data, making um actually looking at it and turning that into an actual insight, that's gonna happen super fast.
00:35:24.880 --> 00:35:29.679
Um, and frankly, we're kind of working on trying to accelerate that change as well, right?
00:35:29.840 --> 00:35:32.320
So obviously, of course, I believe that.
00:35:32.480 --> 00:35:37.599
Um, but on top of that is, you know, we mentioned personalization earlier, right?
00:35:37.760 --> 00:35:46.719
And I think that one thing I really believe, but I think people are focusing on the wrong things, is they're looking at workflow automation for a lot of these use cases.
00:35:46.880 --> 00:35:48.159
I'm not saying that's a bad thing.
00:35:48.320 --> 00:35:50.079
I'm saying you're not thinking big enough.
00:35:50.320 --> 00:35:53.920
Because ultimately you're kind of thinking in terms of like, hey, how can I produce faster reporting?
00:35:54.159 --> 00:35:56.719
How can I spit out blog posts more quickly?
00:35:56.960 --> 00:36:03.599
But you know, in the world we live in, right, it's like, is more spitting out blog posts more quickly going to get me more eyeballs?
00:36:03.760 --> 00:36:06.159
No, because of what's happening with all our channels.
00:36:06.400 --> 00:36:10.320
But we can get to a point really quickly where we can process data really quickly.
00:36:10.559 --> 00:36:17.440
We can um see people that are coming in through, say, a meta ad versus a you know, organic post versus an email.
00:36:17.599 --> 00:36:20.960
And we have an ability to kind of like adjust.
00:36:21.119 --> 00:36:33.199
Um, I can take my kind of brand guidelines, my main copywriting, and then adjust, pattern recognize um my creative or my offer or my copywriting for a specific um user.
00:36:33.280 --> 00:36:36.719
Um, the kind of analogy I use for it, frankly, is code switching, right?
00:36:36.800 --> 00:36:46.239
How do I um code switch a little bit to this customer so that they I can use the words that they would use in their own mind and that they get the uh the value back much more quickly.
00:36:47.519 --> 00:36:54.400
This is gonna become so quick, I think, that we're not just gonna be able to like push out social media content much more quickly.
00:36:54.639 --> 00:37:04.239
I think we can get to the point where your website very quickly can kind of figure out who's visiting it and then target bright personalized offer to the right person at the right time.
00:37:05.119 --> 00:37:09.440
And that's I think a few more steps ahead of where people are going.
00:37:09.519 --> 00:37:14.239
But I think it's where a lot of this technology is headed in a few years beyond simple workflow automation.
00:37:14.559 --> 00:37:21.519
And I think especially if we can get to the point where it's not just doing copywriting, but the other part is discipline, right?
00:37:21.599 --> 00:37:25.199
Can we discipline based off of what's actually getting engagement and conversions?
00:37:25.599 --> 00:37:30.079
That's how you get higher performance, where people will start to see the value of that particular technology.
00:37:30.960 --> 00:37:32.000
Excellent stuff.
00:37:32.880 --> 00:37:36.320
Steven, talk to me, unpack, and we'll wrap up here soon.
00:37:37.039 --> 00:37:40.880
This is uh this has been great and super um impactful items.
00:37:41.599 --> 00:38:03.599
I'm curious, a lot of CRO insights, incredibly powerful, deep discovery within analytics, obviously map mirroring or mapping out and um articulating heat mapping insights, GA data, um, and a lot of these CRO reports are proposing wireframes and things like that.
00:38:03.760 --> 00:38:10.159
None of all of this has been um happening for years, but now it's happening a lot faster, as we know.
00:38:10.239 --> 00:38:15.920
It's much easier to compile all that with cloud design, with all sorts of other things that have been really amazing.
00:38:16.079 --> 00:38:20.960
Uh that said, I'm curious your thoughts, because uh we've experienced this in other agencies as well.
00:38:21.039 --> 00:38:23.199
They really love the data, they love the insight.
00:38:23.599 --> 00:38:27.920
But when it comes to actually implementing tests, there's a bit of hesitation.
00:38:28.159 --> 00:38:46.079
And I think I love you know what Crazy Egg's been able to do, but it's it's a little bit, it's like, oh gosh, we have to engage the copywriter, we have engaged our developer, we have to engage a UX team, we have to engage all these other and it's and it's it's new, it's it's nuanced, it's intimidating, it's potentially time consuming.
00:38:46.559 --> 00:38:56.960
What would you say of how to kind of convince a skeptical CMO to kind of run their first or second test, if you will, for CRO that is perhaps again skeptical of of this whole plan?
00:38:57.440 --> 00:39:06.719
Yeah, I would say that you know, if you're expecting that either this is gonna produce as at least right now, this will probably change next year.
00:39:07.039 --> 00:39:20.559
But if you're expecting that, hey, I want this to be able to generate a finished product from a first bet, or as we would say, like four from a one-shot, then I think that's kind of an unreasonable expectation.
00:39:20.719 --> 00:39:25.760
And frankly, you don't need to have a finished product to benefit from AI, right?
00:39:25.920 --> 00:39:28.000
Um, we've talked about like wireframes, right?
00:39:28.159 --> 00:39:29.679
We talked about prototyping.
00:39:30.000 --> 00:39:35.199
The the funniest thing that I found as a copywriter is my biggest enemy is a blank page, right?
00:39:35.360 --> 00:39:42.800
Is it's very turns out I have a lot of opinions about what the design and copywriting looks like once I have a bad example in front of me.
00:39:43.119 --> 00:39:43.679
For sure.
00:39:44.000 --> 00:39:44.320
Right.
00:39:44.559 --> 00:40:01.039
And so what's funny is, you know, I can um, but as it turns out, it's really easy to get a bad but a stepping stone first draft by having the AI look at a lot of say my brand guidelines, look at a lot of my positioning, even feedback that I've gotten from customer quotes, right?
00:40:01.199 --> 00:40:07.599
To produce a first draft that is, let's call it a one-year out-of-school quality first draft.
00:40:07.760 --> 00:40:11.119
Like uh 23-year-old me, what my copywriting might have looked like.
00:40:11.599 --> 00:40:28.239
And it can produce something, like say a slide deck or a page prototype and say, hey, based off of uh all the training documents you provided me right now, this is what, say, a train a sales slide deck or a uh you know agency-facing landing page should look like.
00:40:28.400 --> 00:40:34.960
I can look at it and say, okay, there's some poor judgment here as far as like which um stuff that agencies want to look at.
00:40:35.119 --> 00:40:37.199
Um, but I can see the nuggets in here.
00:40:37.280 --> 00:40:40.079
I can say, hey, all right, let's move these sections further up.
00:40:40.239 --> 00:40:43.360
Let's rewrite these particular sections, but working upon the same theme.
00:40:43.519 --> 00:40:50.480
And as it turns out, it's like at least automating the uh the job of getting a bad first draft on.
00:40:50.559 --> 00:40:54.239
And almost every first draft of it has ever been made by humanity is bad, right?
00:40:54.400 --> 00:40:57.599
So so we don't need a good quality first draft.
00:40:57.679 --> 00:41:07.599
We need something to react to so that we can start slotting in our actual inspiration, our actual original ideas, and kind of factoring the this stuff in as well.
00:41:07.840 --> 00:41:19.039
So I would say, frankly, hey, if you're looking to just get started, you can use it for competitor research, for example, and say, like, hey, um, I'm saying I'm a marketer looking at Hotjar right now.
00:41:19.119 --> 00:41:20.320
What is hotjar good for?
00:41:20.400 --> 00:41:22.960
And where is Crazy Egg um a better alternative?
00:41:23.199 --> 00:41:29.440
That's super interesting because I can kind of tell, like, hey, if I had no information whatsoever, what are my customers going to see?
00:41:29.599 --> 00:41:44.800
Um, but conversely is, you know, if I produce a say sales collateral based off of my all my internal documents and produces something that is off message, that implies that the internal kind of documents that I have are off message and something's missing.
00:41:44.960 --> 00:41:50.079
So that highlights gaps basically in terms of what I'm actually even giving my team, but I can provide them as well.
00:41:50.320 --> 00:41:57.920
So what I would frankly say is like, hey, AI is kind of like a new kind of way of computing, but instead of programming, you're actually training it.
00:41:58.079 --> 00:42:06.000
But you know, sometimes if your employee is doing something new and they're doing something not the right way, that's a gap in my training, right?
00:42:06.159 --> 00:42:13.679
And as it turns out, if the AI is not quite getting 90% there, then hey, there's some training we can do, and that's something we can address.
00:42:14.000 --> 00:42:17.519
Yeah, that foundational component training up is so key.
00:42:17.760 --> 00:42:19.280
Steven, it's been an absolute pleasure.
00:42:19.440 --> 00:42:20.639
Really appreciate your time.
00:42:20.800 --> 00:42:23.920
Hey, where can people interact and engage with you?
00:42:24.000 --> 00:42:27.360
And then obviously Crazy Egg, talk talk about that a little bit.
00:42:27.599 --> 00:42:27.760
Yeah.
00:42:27.840 --> 00:42:30.480
So believe it or not, I actually take direct customer emails.
00:42:30.639 --> 00:42:33.679
So if you want to email me, I'm at steven at crazyag.com.
00:42:33.760 --> 00:42:34.880
That's Steven with a pH.
00:42:35.199 --> 00:42:37.119
Um, you can also find me on LinkedIn.
00:42:37.360 --> 00:42:37.840
Awesome.
00:42:38.000 --> 00:42:38.400
Awesome.
00:42:38.559 --> 00:42:47.360
Yeah, and obviously Crazy Egg is uh, you know, shout out to them and and you guys just wonder what you've been able to accomplish and uh continues to really lead the pack.
00:42:47.519 --> 00:42:48.719
So really appreciate it, Steven.
00:42:48.880 --> 00:42:49.360
Thanks again.
00:42:50.239 --> 00:42:50.719
Thanks so much.
00:42:50.880 --> 00:42:51.840
Awesome.
00:42:55.519 --> 00:42:58.000
Thanks for listening to Built to Convert.
00:42:58.159 --> 00:43:06.480
If you're serious about growth, don't just listen, subscribe, leave a review, and follow us for new episodes that drop straight into your feed.
00:43:06.639 --> 00:43:09.280
Until next time, keep building.
