EP 133: Uber Eats' First Marketing Hire On AI Search

WEBVTT
00:00:00.000 --> 00:00:01.760
Welcome to Embracing Marketing Mistakes.
00:00:01.760 --> 00:00:13.279
I'm Chris Norton and my guest today is Jenna Hannon, co-founder of Hatter, an agency in California, and the AI search platform I've been testing on our own website.
00:00:13.279 --> 00:00:16.719
She was also the first marketing hire at Uber Eats.
00:00:16.719 --> 00:00:18.559
And then there's the biggest miss of all.
00:00:18.559 --> 00:00:26.399
She spent hours in a car to the Sahara with Greg Brockman, the co-founder of OpenAI.
00:00:26.399 --> 00:00:30.079
She begged him to take her money and he said no.
00:00:30.079 --> 00:00:31.679
That was seven years ago.
00:00:31.679 --> 00:00:33.520
Imagine how much that would be worth now.
00:00:33.520 --> 00:00:35.039
This is one you're not gonna miss.
00:00:35.039 --> 00:00:36.560
This is one of my favourite shows ever.
00:00:36.560 --> 00:00:37.920
I guarantee you're gonna enjoy it.
00:00:37.920 --> 00:00:39.930
Let's get into it.
00:00:39.930 --> 00:00:42.993
Enjoy.
00:00:42.993 --> 00:00:50.719
Jenna Hannon, welcome to the show.
00:00:50.719 --> 00:00:52.338
Thanks for having me.
00:00:52.338 --> 00:00:53.354
Um, right.
00:00:53.354 --> 00:01:03.200
Well, this has been interesting because you're the one of the first guests that I've had on the show where I've actually gone off and before you've even come on the show, I was like, oh, this sounds really intriguing.
00:01:03.200 --> 00:01:05.920
Um the what you've done with Hatter.
00:01:05.920 --> 00:01:14.879
Like, because I meet so many fascinating people um and experts in their fields, but what you've done with Hatter really I was like, I've got to trust test this out for myself.
00:01:14.879 --> 00:01:28.560
So I've I have hands up on on this, I'm I am a test customer, I'd say, because I'm I'm gonna give it I've I've been working, I think I've been using your your your kit for about uh two and a half weeks, three weeks or something.
00:01:28.560 --> 00:01:37.760
And as a case study, live on the podcast, I can say that um yeah, in the the curve is we've got the we've got the uh the umbrella curve, what do you call that?
00:01:37.760 --> 00:01:38.640
It's a bell curve.
00:01:38.640 --> 00:01:39.670
Where it goes up.
00:01:39.670 --> 00:01:42.799
Yeah, the hockey stick effect, that's it.
00:01:42.799 --> 00:01:52.159
Um and so, yeah, so and this show's all about mistakes, and actually the whole concepts come out of a mistake, hasn't it, for Hatter?
00:01:52.159 --> 00:01:56.560
So do you want to explain a little bit about Hatter and where the idea came from?
00:01:57.200 --> 00:01:58.480
Yeah, absolutely.
00:01:58.480 --> 00:02:02.879
So we focus on what's now called AEO or GEO.
00:02:02.879 --> 00:02:06.319
Uh I think the non-technical term is AI search.
00:02:06.319 --> 00:02:12.502
So generative engine optimisation or answer engine optimisation is what the acronyms stand for.
00:02:12.502 --> 00:02:18.879
Uh, and we are an agentic platform that basically replaces an agency.
00:02:18.879 --> 00:02:23.759
So we we started as an agency just focused on organic search.
00:02:23.759 --> 00:02:27.520
And through that, we started building software to automate ourselves.
00:02:27.520 --> 00:02:28.639
We're tech backgrounds.
00:02:28.639 --> 00:02:30.319
So my co-founder is an engineer.
00:02:30.319 --> 00:02:34.240
He comes from Deliveroo and Microsoft, and I'm from the marketing side.
00:02:34.240 --> 00:02:37.013
Uh, I was at Uber for many years.
00:02:37.013 --> 00:02:42.240
Uh, but after Uber, I became a fractional CMO, and that's where the mistake happened.
00:02:42.240 --> 00:02:49.562
So I was working with several kind of mid-stage startups, Series B, Series C companies.
00:02:49.562 --> 00:02:52.639
Uh, and one of them was buzzing along.
00:02:52.639 --> 00:02:57.039
They were doing about 10 mil in revenue growing.
00:02:57.039 --> 00:03:00.319
And when I came in, most of that revenue was coming from ads.
00:03:00.319 --> 00:03:03.039
So we were spending a lot on ads.
00:03:03.039 --> 00:03:09.062
And the founder felt fairly confident that he'll just keep fundraising, there'll be more money, no problem.
00:03:09.062 --> 00:03:13.919
Uh, and I kept encouraging him, we should try to diversify our marketing channels.
00:03:13.919 --> 00:03:17.520
We should think about organic, which to me at the time was SEO.
00:03:17.520 --> 00:03:22.240
Uh, we should think about so at Uber, most of our business was actually organic.
00:03:22.240 --> 00:03:25.479
So we had so much demand for that business.
00:03:25.479 --> 00:03:31.439
Uh, we still ran ads, uh, but the majority of that business coming in was organic.
00:03:31.439 --> 00:03:35.295
People re people hearing about it, people finding it online.
00:03:35.295 --> 00:03:39.199
Uh, and so this is the pitch that I made to the founder.
00:03:39.199 --> 00:03:42.439
Uh no, no, no, kept trying.
00:03:42.439 --> 00:03:43.520
No, no, no.
00:03:43.520 --> 00:03:46.719
And then he went to go fundraise.
00:03:46.719 --> 00:03:49.039
It seemed like it was looking good.
00:03:49.039 --> 00:03:52.719
And at the last minute, investors pulled out.
00:03:52.719 --> 00:03:55.360
So we ended up with no cash for ads.
00:03:55.360 --> 00:04:00.879
And so we we basically turned off the ad machine and revenue just dried up overnight.
00:04:01.360 --> 00:04:01.599
Right.
00:04:01.840 --> 00:04:06.639
And I felt like I felt like this is as a CMO, this is my fault, you know?
00:04:06.879 --> 00:04:13.680
Like I'm in charge of building the growth machine, and I let this happen.
00:04:13.919 --> 00:04:16.399
And that's kind of what inspired Hatter.
00:04:16.480 --> 00:04:22.160
There was a spark of going down this rabbit hole of organic, what's happening organic.
00:04:22.480 --> 00:04:33.759
My my co-founder and I were fascinated by this shift from traditional SEO that had been the same for 20 years, uh, to now AI search coming online and what that was going to do to organic.
00:04:33.920 --> 00:04:36.160
And and that's what started it.
00:04:36.959 --> 00:04:47.759
Yeah, and the fascinating thing in or the fascinating thing in in AEO GEO, let's call it AI Search, is when it came out, like we're talking 2022.
00:04:47.839 --> 00:04:52.000
So we're talking four years ago since ChatGPT 3, I think it was, that launched.
00:04:52.160 --> 00:05:02.319
Weirdly, I don't think anybody ever saw ChatGPT one or two, but ChatGPT 3 came out, and it's it's sort of I remember when it came out and everyone was like, whoo, there was suddenly a lot of noise.
00:05:02.480 --> 00:05:04.800
Oh my god, this is actually quite useful.
00:05:05.040 --> 00:05:08.639
And I saw things like, is this the Google killer?
00:05:08.879 --> 00:05:10.399
Google it Google's done.
00:05:10.560 --> 00:05:24.720
Google's been caught with its pants down, and it was like, um, you know, because it come from nowhere, and people are now going, oh, I can just ask it, and it will give me um proper thought-through results that are that I feel and you're trusting it, obviously.
00:05:24.959 --> 00:05:26.319
We'll get into that in a minute.
00:05:26.480 --> 00:05:46.000
Um, but actually, I had an interview with the head of uh, I think actually head of marketing at um Google in the UK or in Dublin or wherever they're based in the Europe, and she was saying that actually AI um the search for Google, the usage of Google's gone up since um which is amazing for them.
00:05:46.160 --> 00:05:52.399
I know search has changed on their platform, the amounts of different search, and they've embraced AI search.
00:05:52.560 --> 00:06:06.480
Um, but it's been fascinating to me because I I genuinely thought, oh my god, Google's got all this money, they've they've got the they've got the map, they've got the mass um reach and everything, and suddenly this Sam Altman and his team have just absolutely blown it out of the water.
00:06:06.639 --> 00:06:09.519
So, how how has it been for you guys, your side?
00:06:09.759 --> 00:06:15.040
Is it is it still as exciting and and do you think Google is still playing catch-up?
00:06:15.279 --> 00:06:16.720
Or do you think Google's fine?
00:06:16.800 --> 00:06:22.319
And they're just like, you guys go off and fight with your frontier models, but we're just gonna keep leading the way.
00:06:23.120 --> 00:06:25.040
We think Google's fine.
00:06:25.040 --> 00:06:29.199
We so we started in this game in 2023.
00:06:29.199 --> 00:06:31.839
So it right at the beginning.
00:06:31.839 --> 00:06:38.160
And at the beginning, it was hard to convince companies that this is where search was going.
00:06:38.160 --> 00:06:42.319
That the companies were like, why would I want to show up in ChatGPT?
00:06:42.319 --> 00:06:46.240
That's only a small percentage of traffic.
00:06:46.240 --> 00:06:50.720
And our hypothesis, exactly, and when we started, I think it was like 1%.
00:06:50.720 --> 00:06:58.560
Um, but our hypothesis was that Google was building models and that Google will start doing the AI summaries.
00:06:58.560 --> 00:07:02.319
If Google's getting beaten, they're a they're a giant machine.
00:07:02.319 --> 00:07:06.240
There's very smart people at Google, they will pivot quickly.
00:07:06.240 --> 00:07:07.600
And that that's what we've seen.
00:07:07.600 --> 00:07:14.456
So I think now it's been fun because we sometimes we'll talk to companies which are all bought in on SEO.
00:07:14.456 --> 00:07:16.560
No, my customer doesn't use AI.
00:07:16.560 --> 00:07:18.079
My customer's too old school.
00:07:18.079 --> 00:07:22.079
And we're like, your customer, if they're going to Google, is using AI.
00:07:22.079 --> 00:07:26.160
They're doing AI search on Google, whether they know it or not.
00:07:26.160 --> 00:07:30.720
The in the entire thing is shifted thanks to Google making that shift.
00:07:30.720 --> 00:07:37.519
Now, search still matters because when you go to Google, you're still seeing an AI summary and listings.
00:07:37.519 --> 00:07:46.000
But now you're you're kind of in this funny land of a combination of both, where we think probably the next stage is even more AI summaries.
00:07:46.000 --> 00:07:47.120
It's just so convenient.
00:07:47.120 --> 00:07:53.279
Getting an answer at the top for a lot of the questions that you're going to Google to answer is is just really nice.
00:07:53.680 --> 00:07:54.000
Yeah.
00:07:54.000 --> 00:08:02.160
I mean it's the set it's this because I did an event about four or five, maybe five or six years ago, and it was about the future of search.
00:08:02.160 --> 00:08:07.120
I don't do well, I'm a PR person, and um we do digital PR and all that, those sorts of things.
00:08:07.120 --> 00:08:12.160
But the future of search and the few they were talking about the fact that it was gonna be audio search, right?
00:08:12.160 --> 00:08:23.920
Because of Alexa, Alexa's gonna take over the world, she was gonna be the next Terminator, and actually she's been improved slightly, but unless she's just tricking us all, she's not as clever as I thought she was gonna be.
00:08:23.920 --> 00:08:40.480
And they were saying exactly what you're saying, which is be the first result, because um she's gonna go, Alexa, she, I say she's she or he can have either way, but anyway, and Alexa in my house would go, yeah, it'll give you the first result, won't it?
00:08:40.480 --> 00:08:49.840
And not the others, like Google used to always give you the 10, and the the the mission of an SEO was to be one to nine, preferably one to three, but one to nine.
00:08:49.840 --> 00:08:59.923
Whereas now you're not sure if you're one to nine, it might be one of five in of an LLM, and each and it's like trying to choose which one do you want to beat?
00:08:59.923 --> 00:09:10.080
Perplexity, uh, Gemini, ChatGPT, Claude, I mean, there's that many of them, and everybody's just going throwing their balls in the air and going, which one's gonna land?
00:09:10.159 --> 00:09:19.519
And so, yeah, I mean, I don't know what the question is there, but the good news there is it's it's simpler than you think.
00:09:19.519 --> 00:09:28.240
So a lot most of the AI tools under the hood, even though they're slightly different models, are generally working in a similar way.
00:09:28.240 --> 00:09:39.120
So if you're if you're optimising for ChatGPT, the tactics aren't going to be that different to show up in Perplexity Claude or Google AI Overviews.
00:09:39.120 --> 00:09:40.320
I think that's the good news.
00:09:40.320 --> 00:09:43.615
Still in our data, I think this is quite interesting.
00:09:43.615 --> 00:09:47.840
90% of the traffic we're seeing in AI is happening in ChatGPT.
00:10:38.059 --> 00:10:39.419
I would say that's true as well.
00:10:39.579 --> 00:10:48.779
I can confirm that my I look at my HubSpot data and my HubSpot traffic, and um, yeah, we come from uh most of them are chat, most of the AI search.
00:10:48.860 --> 00:10:53.419
We've been tracking it since AI has been out, uh, is ChatGPT, you're right, you're correct.
00:10:53.579 --> 00:10:57.339
But then I suppose it's getting confused with what Google traffic is now.
00:10:57.500 --> 00:10:59.099
Because it will be Google, won't it?
00:10:59.259 --> 00:11:06.539
Because Google is, you know, it's still, you know, you might get 200, 300 visits from AI, whereas you might get 5,000 from Google.
00:11:06.620 --> 00:11:12.459
So and that's like you've just said there, they're all like most of them are AI search, I think, if it if it's changed its way of working.
00:11:13.019 --> 00:11:13.419
Could be.
00:11:13.500 --> 00:11:14.459
I think that's blending.
00:11:14.620 --> 00:11:15.339
So I'm with you.
00:11:15.419 --> 00:11:17.179
I think Google is still people's default.
00:11:17.339 --> 00:11:22.779
When they think I need to search something, they go to Google, and now Google is pushing you to get an AI answer.
00:11:22.860 --> 00:11:24.779
So I think that traffic is blended.
00:11:25.179 --> 00:11:27.179
So having used your product, right?
00:11:27.259 --> 00:11:29.419
You've got you've got you've got an agency, haven't you?
00:11:29.500 --> 00:11:35.579
And they you you can do managed in-house and they can obviously take your products, and there'll be people listening to this thinking that's great.
00:11:35.819 --> 00:11:40.539
Um, I've had I've had a few things where I'm like, how does this how is this working?
00:11:40.779 --> 00:11:47.259
So because your the the content of the content that your application does is is excellent.
00:11:47.579 --> 00:11:48.939
That's the that's the key thing.
00:11:49.099 --> 00:11:52.620
That's the reason why I started to use it because I looked at it, I thought, this is excellent.
00:11:52.699 --> 00:12:00.139
The articles, they're not, they're sort of like more like guides and things that and they they complement the stuff that we're already doing.
00:12:01.099 --> 00:12:12.139
But my question is this SEO, AEO, AI Search is there's so many content marketers out there, people that have got jobs in content marketing, right?
00:12:12.219 --> 00:12:19.740
Obviously, a lot I'm I I worry for the copywriters, but the con the content writers, you you're writing um blog posts.
00:12:19.899 --> 00:12:25.980
Are blog posts still as important as they were uh like five, six years ago?
00:12:26.139 --> 00:12:31.179
Because it's always been like write a really well executed and you get crap blog posts, you can get excellent blog posts, right?
00:12:31.259 --> 00:12:37.500
Really well listicles, for example, are a great use of um but yours are slightly they're like a hybrid of that.
00:12:37.659 --> 00:12:44.939
They're like they're not quite that, they're more resources and uh insightful reviews of sectors and stuff.
00:12:45.099 --> 00:12:52.219
So is the is the are these two meant to go together as part of a collective strategy, is what I'm trying to say.
00:12:53.179 --> 00:12:53.659
Yes.
00:12:53.899 --> 00:13:05.579
So the way the way that we look at it is that uh a blog post, whether it's good or bad, if no one sees it, then it's not useful.
00:13:05.740 --> 00:13:12.219
So you can write the best blog post or the there could be the best page on your website that you spent tons of time crafting it.
00:13:12.299 --> 00:13:20.299
But if it doesn't have distribution, meaning it's not showing up in search and people aren't finding it, it's not gonna actually be useful to your business.
00:13:20.459 --> 00:13:25.740
So I think people overweight the quality of content versus the distribution.
00:13:25.899 --> 00:13:36.139
Both both matter, I would say equally, because once first you got to get someone to hit that piece of content or that page on your website, and then you need that person to read that and convert.
00:13:36.219 --> 00:13:37.500
And it needs to be the right person.
00:13:37.579 --> 00:13:39.419
So there's quite a few jobs to do there.
00:13:39.579 --> 00:13:46.860
But I would say to answer your question more simply, at a concept level, it it's fairly easy to understand.
00:13:47.099 --> 00:13:54.539
Basically, now we have these AIs that are training on all of the information that is on the internet.
00:13:55.500 --> 00:13:57.099
And ever, exactly.
00:13:57.259 --> 00:14:02.379
Yeah, as you can imagine, you previously had a website.
00:14:02.620 --> 00:14:10.779
For many years, people weren't blogging, people weren't updating their website because it wasn't helping you in in Google search.
00:14:11.019 --> 00:14:18.459
And so what we often see is that these models actually don't have that much information about your company.
00:14:18.779 --> 00:14:20.939
And that that's all-encompassing.
00:14:21.019 --> 00:14:32.620
It's your industry, your market, your competitors, the pricing, all the things that your customer would look up when deciding whether or not to buy your product or service.
00:14:32.939 --> 00:14:38.860
So the simple solution from an AI perspective is to give it that information.
00:14:39.099 --> 00:14:51.579
So, what our tool is doing is it's basically spinning up an external marketing guide, which is trying to answer every single question your customer asks along that buying journey.
00:14:51.740 --> 00:14:53.659
That could be who do you compete with?
00:14:53.819 --> 00:14:55.179
Who should they compare it with?
00:14:55.339 --> 00:15:00.779
Um, why you were better or what the differentiators are, how are you priced versus your competitors?
00:15:00.939 --> 00:15:03.819
Uh, what particular use cases are you for?
00:15:03.980 --> 00:15:06.139
And that all needs to sit on your website.
00:15:06.219 --> 00:15:09.659
Otherwise, AI doesn't know any of that exists.
00:15:09.819 --> 00:15:26.219
And so I would say 99% of the time, even with very mature companies, we will go to their website and we will quickly find out that there's just so many gaps in content, which is why they are not showing up for particular searches.
00:15:26.379 --> 00:15:29.259
So, concept level is quite easy to understand.
00:15:29.500 --> 00:15:39.099
The way that we execute that is we basically spin up a marketing hub for you, and our tool identifies what those gaps are.
00:15:39.259 --> 00:15:41.259
So, where are your competitors showing up?
00:15:41.339 --> 00:15:48.299
Uh, is it are they talking about pricing or comparisons or are they answering questions that your customers might be asking that you're not asking?
00:15:48.459 --> 00:15:50.779
And then we start building essentially guides.
00:15:50.939 --> 00:16:06.459
And so what our tool is doing is essentially helping you draft those guides, but it's all grounded in what your business does, your industry, and pretty much as much context as you want to feed it, because the the context is what makes the content good.
00:16:06.860 --> 00:16:07.099
Okay.
00:16:07.179 --> 00:16:16.779
And then the so the content you so so say I was you um, so you've got you've got your agency side, you've got a minimum fee, a minimum retainer of six months.
00:16:16.939 --> 00:16:19.419
And when we spoke initially, I was like, oh my god, this sounds amazing.
00:16:19.500 --> 00:16:20.059
We're gonna try it out.
00:16:20.219 --> 00:16:25.339
So I can interview you, I've got a bit of a contextual of how it works because I find it interesting.
00:16:25.579 --> 00:16:33.259
Um and you said to me, Oh, it can take about six to eight weeks to see some sort of tangible difference.
00:16:33.500 --> 00:16:39.339
What so what happens if someone's with you for like uh on your app and they're with you for nine months?
00:16:39.500 --> 00:16:43.899
Then if they leave, do they keep the content or is it your content, or how does it work?
00:16:44.459 --> 00:16:45.693
Yeah, great question.
00:16:45.693 --> 00:16:47.980
Yeah, so we we essentially give them the content.
00:16:47.980 --> 00:16:52.219
So you can you can export your content and put it on your CMS.
00:16:52.219 --> 00:16:59.339
One of the reasons that our tool hosts it is because lots of people have not touched their website in a long time.
00:16:59.339 --> 00:17:03.099
So their website might be slow, it's not very optimised.
00:17:03.099 --> 00:17:12.140
And so if we just start putting content on a website that loads slowly, that's not easy for crawlers to read, then it's not gonna pick up that content right away.
00:17:12.140 --> 00:17:19.019
So what we do is we basically give you an SEO optimised hub, which is we know is gonna get picked up right away.
00:17:19.019 --> 00:17:23.740
And anytime we create a new guide, we actually submit that directly to Google for you.
00:17:23.740 --> 00:17:27.420
So we remove that manual step and we say, hey, Google, go crawl this.
00:17:28.460 --> 00:17:30.460
And so what's the first sort of misunderstanding?
00:17:30.460 --> 00:17:33.259
What's the main misunderstanding you get with new customers?
00:17:33.259 --> 00:17:34.299
How many customers have you got?
00:17:34.299 --> 00:17:37.500
And what's the misunderstanding they have with with AI search?
00:17:37.500 --> 00:17:37.980
Do you think?
00:17:37.980 --> 00:17:39.019
Because is it quite hard?
00:17:39.019 --> 00:17:43.500
Because you I think you said to me, you're one of the first people that I've just spoken to and you've gone off and done it.
00:17:43.500 --> 00:17:47.259
Like you you said like people are really careful and don't understand it.
00:17:47.259 --> 00:17:48.860
I like trialling everything.
00:17:48.860 --> 00:17:50.860
I test anything, anything that works.
00:17:50.860 --> 00:17:52.220
I'm always like to try something.
00:17:53.659 --> 00:17:56.779
As you can imagine, our agency side is much bigger.
00:17:56.779 --> 00:18:08.539
So I think the the market is still nervous about about just using AI agents to do something that where the quality really matters.
00:18:08.539 --> 00:18:11.659
But on the software side, this is where the market is going.
00:18:11.659 --> 00:18:21.339
Uh so we we had this hypothesis very early that with AI, you in the old world of SaaS software, you were buying a workflow.
00:18:21.339 --> 00:18:25.740
You were buying a tool that would make your workflow faster, easier, more productive.
00:18:25.740 --> 00:18:29.180
But at the end of the day, it was still a human running that workflow.
00:18:29.180 --> 00:18:32.539
Now you can buy software that's delivering results.
00:18:32.539 --> 00:18:40.619
And this is a very new concept because before only service businesses could deliver results because you need a human.
00:18:40.619 --> 00:18:45.659
Or you're using a software and you're bringing it internally and then you're hiring someone to use that software.
00:18:45.659 --> 00:18:49.980
Now agents can run a lot of different workflows.
00:18:49.980 --> 00:18:54.859
And if you build specialised agents, you can lead to the end results.
00:18:54.859 --> 00:18:57.019
So that was that was our initial hypothesis.
00:18:57.019 --> 00:19:00.539
Now you have to wait for the market to catch up because that's a scary concept.
00:19:00.539 --> 00:19:06.059
It's just never existed before, where automation was at the level that it could deliver results.
00:19:06.059 --> 00:19:08.299
It was always just productivity.
00:19:08.299 --> 00:19:19.500
So on the software side, oddly enough, that's a harder sell than on the agency side, where a human is in the loop on all of those steps walking you through how that works.
00:19:19.899 --> 00:19:21.740
Because people want to speak to people.
00:19:22.379 --> 00:19:22.770
Yes.
00:19:22.770 --> 00:19:23.259
Yeah.
00:19:23.259 --> 00:19:27.099
And that makes sense to me because that's this has never existed before, right?
00:19:27.099 --> 00:19:30.619
You've never been able to trust agents.
00:19:30.619 --> 00:19:36.779
And I think now we're we're in an interesting place where everyone was like, oh, I can build my own AI agents.
00:19:36.779 --> 00:19:39.899
So we talked to a lot of companies where they're like, oh, why would we use you?
00:19:39.899 --> 00:19:41.659
We can use our own AI agents.
00:19:41.659 --> 00:19:47.180
But it kind of goes right back to why you would hire a specialist agency like Chris, your PR firm.
00:19:47.180 --> 00:19:51.899
It's because you have spent many, many years perfecting that methodology.
00:19:51.899 --> 00:19:54.859
And without that, it actually doesn't work.
00:19:54.859 --> 00:20:01.819
And that's the same thing people are realising with these agents that are just empty shells that they can plug, plug in.
00:20:01.819 --> 00:20:05.819
If you don't, if you don't know the methodology, that agent is not going to be effective.
00:20:05.819 --> 00:20:18.700
So I think we're seeing this move to specialised agents where people are realising, oh, I can get results, but I do need to trust that these agents are running a process that is effective or a workflow that is effective.
00:20:18.700 --> 00:20:20.518
So they do need to be trained.
00:20:20.518 --> 00:20:24.059
Uh, and I think the market is just starting to get there.
00:20:24.059 --> 00:20:31.500
So, especially with some of our, we have clients who have been in business for 40 years who are wrapping their head around this.
00:20:31.500 --> 00:20:34.460
But it's interesting, we have some of those clients trying the agentic tool.
00:20:34.460 --> 00:20:38.079
Like we have a commercial linen service that just signed up last week.
00:20:38.079 --> 00:20:45.980
Uh, and it's it's magical to walk them through it and it's it just blows their mind because this is such a big piece of the growth of their business.
00:20:45.980 --> 00:20:48.278
And so they're putting their trust in us.
00:20:48.278 --> 00:20:48.642
Yeah.
00:20:48.642 --> 00:20:54.220
Uh, but I think the market is still catching up to that, which is why our agency is still the bigger part of the business.
00:20:54.539 --> 00:20:59.659
You could you just reminded me of when uh because I've been in PR for that long, 25 years, God.
00:20:59.659 --> 00:21:05.339
And um, when SEO came out, I remember we we I was in a meeting and we had we had a brand in the UK.
00:21:05.339 --> 00:21:07.180
It's used it was quite a big brand in the UK.
00:21:07.180 --> 00:21:10.379
It was called the SOC the Sock Shop, I guess what they did.
00:21:10.379 --> 00:21:15.899
And um I I remember meeting the head of SEO there, and they were talking about articles.
00:21:15.899 --> 00:21:22.460
They were like, Oh yeah, we have to do loads of articles about the different set, and they had all these different copywriters writing all these articles.
00:21:22.460 --> 00:21:28.940
And I was like, Are you so are you telling me that you write 17 articles a day on socks?
00:21:28.940 --> 00:21:33.579
Because that to make sense, and I think that's probably where we've moved away from.
00:21:33.579 --> 00:21:40.859
We're probably moving away from humans writing articles about socks to better quality, well-researched articles.
00:21:40.859 --> 00:21:43.899
I think there's gonna be a I think there's gonna be a slight shift.
00:21:43.899 --> 00:21:48.220
We'll have we'll go automated, then there'll be the human-opinionated stuff.
00:21:48.220 --> 00:22:07.500
I think there's gonna be like the bit in the middle of the the diagram where people want like you just said there, people want to speak to people, they still trust people, they'd rather deal with, they'd rather pay a bit more and have a person on the end of a an email than I'll just click enter and let this rogue agent go, you know, hack Hugging Face or whatever.
00:22:07.500 --> 00:22:09.019
I know you won't do that, but you're not a vague.
00:22:09.180 --> 00:22:17.980
I I think the concept that's tough for people to grasp right now is that what you write about is more important than how you write it.
00:22:18.220 --> 00:22:26.460
So when we first started, everyone, the the biggest pushback we got, the biggest objection was is Google gonna accept AI writing?
00:22:26.700 --> 00:22:33.259
And and Google published, and we talked to Google, and they said, we don't care how you write.
00:22:33.420 --> 00:22:35.099
We care what you write about.
00:22:35.339 --> 00:22:42.539
And so that's that's the big sticking point in AI search that most people don't understand is that the precision of what you write about.
00:22:42.700 --> 00:22:44.220
Now, here's the change from SEO.
00:22:44.379 --> 00:22:47.980
In SEO, the tactic was oh, get as many keywords as you can.
00:22:48.700 --> 00:22:50.460
You can grab as much traffic as you can.
00:22:50.619 --> 00:22:56.700
A lot of those keywords were they didn't lead to a buyer coming to your website.
00:22:56.859 --> 00:22:59.420
They might lead to someone who's just looking for information.
00:22:59.579 --> 00:23:02.859
They might read a blog post, but they're just looking for that piece of information.
00:23:02.940 --> 00:23:04.539
They're not actually looking to buy something.
00:23:04.700 --> 00:23:11.579
And so the important difference now is in the what you write about, all of the content should have commercial intent.
00:23:11.740 --> 00:23:20.299
It's just like building sales materials where you are trying to walk a prospect through the decision to buy.
00:23:20.460 --> 00:23:20.779
Yeah.
00:23:20.940 --> 00:23:22.619
And that's the what you write about.
00:23:22.700 --> 00:23:29.500
It's all commercial, which means if that brings someone into your website, it's more likely the right target, which is much higher intent.
00:23:29.819 --> 00:23:31.579
But I've got a question here then.
00:23:31.659 --> 00:23:39.180
So it used to be like, okay, so the any platform, it's like LinkedIn posts three times a week, or you can post once a week.
00:23:39.339 --> 00:23:44.059
You should write blog posts that are well done three times a week to keep your blog refreshed.
00:23:44.299 --> 00:23:49.099
Now, with this sorts of content that I've read of your stuff, like I'll give you an example.
00:23:49.259 --> 00:23:53.579
Two of them are like 17-minute reads, they're like long, long pieces.
00:23:53.819 --> 00:24:04.059
How often, like I've been with it for three weeks, and I think I've I think I've got 15 15 articles at which is three weeks, five five a weeks, well one a day, really.
00:24:04.220 --> 00:24:13.659
Um but obviously there must come a point where I've run out of guides and stuff to write about what my business does and who it can help.
00:24:13.740 --> 00:24:14.539
And do you know what I mean?
00:24:14.700 --> 00:24:20.619
So what what sort of what sort of what's this sort of cadence and and amount that we should be doing, do you think?
00:24:20.700 --> 00:24:22.940
And or is there any guidance on that or not?
00:24:23.339 --> 00:24:24.619
Yeah, absolutely.
00:24:24.619 --> 00:24:27.723
So AI search is not a game of volume.
00:24:27.723 --> 00:24:36.539
Meaning, I think when it first started and people started realising that it was content-based, people were just shipping all these programmatic pages on their website.
00:24:36.539 --> 00:24:38.379
Websites were getting huge.
00:24:38.379 --> 00:24:44.059
Uh, you're not going to be rewarded from the algorithms by having volume.
00:24:44.059 --> 00:24:48.907
The algorithms care about recency, which is basically consistency.
00:24:48.907 --> 00:24:59.740
So, just like you do with social media, the best practice is to post three to five times a week, not more, not less, because the algorithms are looking for how recent are they posting.
00:24:59.740 --> 00:25:05.420
So when we are building your guides, it's kind of a metered cadence.
00:25:05.420 --> 00:25:08.619
And what happens is older pieces of content get stale.
00:25:08.619 --> 00:25:10.460
And they get stale for two reasons.
00:25:10.460 --> 00:25:16.059
One is that they're actually just aged and the algorithms stop looking at it after a certain point in time.
00:25:16.059 --> 00:25:19.899
Google's kind of rule of thumb is a year and a half is stale.
00:25:19.899 --> 00:25:21.579
AI is is much quicker.
00:25:21.579 --> 00:25:23.819
We're saying it looks more like six months.
00:25:24.059 --> 00:25:24.220
Right.
00:25:24.379 --> 00:25:29.259
Uh so well the second reason that articles get stale is that your business changes.
00:25:29.500 --> 00:25:31.099
Businesses evolve.
00:25:31.259 --> 00:25:38.139
Uh, so your your value props evolve, what's happening in your market evolves, your product evolves, your service evolves.
00:25:38.299 --> 00:25:40.779
And so all of that content needs to be maintained.
00:25:40.940 --> 00:25:47.500
So part of what part of the way the tool works is that it's constantly updating content.
00:25:47.659 --> 00:25:54.539
So it's looking at content that's working, that's bringing in traffic and bringing in leads, and it's looking at content that's not working.
00:25:54.700 --> 00:25:58.379
And it's saying we're gonna do more of what's working, less of what's not working.
00:25:58.539 --> 00:26:03.740
And so it's kind of this self-reinforcing cycle, which is what we do as an agency.
00:26:03.899 --> 00:26:04.460
Same thing.
00:26:04.700 --> 00:26:09.180
Every two weeks we're looking at how things are performing, and then we're adjusting the strategy.
00:26:09.339 --> 00:26:11.099
So the tool is built the same way.
00:26:11.259 --> 00:26:21.099
So part of what makes AI search tough to commit to as a business is its consistency, is you just need to keep writing, keep updating your website.
00:26:21.259 --> 00:26:23.740
You need to make sure that content is fresh.
00:26:24.460 --> 00:26:26.220
So keep writing then as well.
00:26:26.460 --> 00:26:27.339
That's the moral of the story.
00:26:27.420 --> 00:26:29.579
Keep looking for ops and keep writing.
00:26:29.899 --> 00:26:33.579
Um, I mean, like it's like social media, right?
00:26:33.659 --> 00:26:37.500
Like it once you stop on your social media account, you're it's not going to be effective.
00:26:37.740 --> 00:26:41.019
Social media just wants you to continue to feed the algorithm.
00:26:41.259 --> 00:26:43.180
Same thing is happening in AI search.
00:26:43.500 --> 00:26:44.539
Yeah, it's crazy.
00:26:44.700 --> 00:26:45.259
It's crazy.
00:26:45.339 --> 00:26:59.259
Um and outside of the the the guides and the reports and things like that, what other things can you do to ensure that your visibility is, you know, what's the what's the hot tip for other marketers that don't that can't afford?
00:26:59.339 --> 00:27:05.019
Maybe the there'll be people in the UK who can't who've got little two one-man, two-man marketing teams.
00:27:05.180 --> 00:27:09.099
Maybe they can't afford to have a you know uh that amount of investment.
00:27:09.259 --> 00:27:16.220
What can they do outside of doing a guide on their sector or a a review of the sales type stuff?
00:27:16.379 --> 00:27:21.099
What other things can they do to be to be more visible, I suppose, on on AI?
00:27:21.259 --> 00:27:23.019
Should they keep doing their blog content?
00:27:23.180 --> 00:27:24.940
Or is that is that dead in the water?
00:27:25.019 --> 00:27:26.859
That's what I was trying to say earlier.
00:27:28.779 --> 00:27:29.819
Yeah, I think so.
00:27:29.980 --> 00:27:33.259
So so AI search is basically two big pieces.
00:27:33.500 --> 00:27:38.619
Piece number one is the content that you're training AI on your business.
00:27:39.099 --> 00:27:52.299
99% of businesses we see do not have fully fleshed out their content, meaning even their homepage doesn't answer it the customer questions that need to be answered to get someone to buy, which is very common.
00:27:52.460 --> 00:28:06.220
So that if you if you don't have a lot to invest, I would say step one, go to your existing website and make sure that there is enough information on there that your customer could make a buying decision without ever talking to you.
00:28:06.379 --> 00:28:11.899
So if you if if you don't have a lot of cash to invest, I would say make sure that content exists on your website.
00:28:12.059 --> 00:28:24.220
Because actually 99% of the time, the algorithms are pulling from first-party websites and there's just not enough content there for AI to understand your industry, how you compete, who you compete with.
00:28:24.460 --> 00:28:28.379
Number two is AI is trained on the entire internet.
00:28:28.940 --> 00:28:41.500
So where else your brand shows up will matter, although it won't matter if the AI doesn't even know you exist, meaning the content is not there in the first place, which is why we tell most companies start with your website.
00:28:41.659 --> 00:28:54.059
Because when we look at the data and we look at your market, most of the time we see that the algorithms are pulling from your website and your competitor websites.
00:28:54.379 --> 00:29:00.139
But in very competitive markets, so Chris, you're in a competitive market, PR is a competitive space.
00:29:00.299 --> 00:29:06.859
In these types of markets, you'll you'll sometimes see that AI is using other sources for credibility.
00:29:07.019 --> 00:29:07.259
Yeah.
00:29:07.500 --> 00:29:12.139
So other sources will be things like online reviews, showing up in press.
00:29:12.299 --> 00:29:15.500
So press is a piece of this, especially exactly.
00:29:15.740 --> 00:29:17.579
And I think that's exactly.
00:29:17.659 --> 00:29:29.420
So a lot of a lot of clients come to us through PR agencies or we work with PR agencies because that is the second key piece to winning AI search, is you want that credibility.
00:29:29.420 --> 00:29:38.299
The thing that I will say is that the AI is is more precise in analysing these third-party sources than Google is.
00:29:38.299 --> 00:29:42.379
Google would basically say a backlink is a backlink and we'll give some weighting.
00:29:42.379 --> 00:29:53.579
So if you have a backlink from a credible uh publication, we will weight that more, but overall volume of links that are coming to you matter more.
00:29:53.579 --> 00:29:56.940
The AI is saying, no, the context of that matters.
00:29:56.940 --> 00:30:03.099
So we notice that the type of press that's moving the needle tends to be more specific to the industry.
00:30:03.099 --> 00:30:05.980
So trade associations pull a lot of weight.
00:30:05.980 --> 00:30:06.940
Niche, exactly.
00:30:06.940 --> 00:30:07.740
That's the right word.
00:30:07.740 --> 00:30:11.740
So niche press matters more than press in itself.
00:30:11.740 --> 00:30:14.700
And the backlink game, I would say, is totally over.
00:30:14.700 --> 00:30:22.700
Meaning AI is very good at saying none of these backlinks are useful or they don't have the right context, and therefore I'm not bringing this in.
00:30:23.099 --> 00:30:25.659
I literally had link building written here.
00:30:25.659 --> 00:30:26.715
I was going to ask you about that.
00:30:26.715 --> 00:30:28.059
That was one of my questions.
00:30:28.059 --> 00:30:30.220
You've just said link building instead.
00:30:30.220 --> 00:30:34.059
Because um I I had a chat yesterday with a guy, right?
00:30:34.059 --> 00:30:42.700
Um it was quite timely for this conversation, and then uh we uh I used to work with a guy and and he he had run he runs an agency.
00:30:42.700 --> 00:30:48.859
There was about 40 of them, SEO agency, uh since 2006, I think.
00:30:48.859 --> 00:30:53.740
Uh he's got a friend who ran one that had 300 staff, and he had 30 odd staff.
00:30:53.740 --> 00:31:08.972
They were global and they did they work mainly in the SaaS you can see where this is going in the SaaS uh space, and they had copywriters, and their their quality was they did quality link building and article writing and placement.
00:31:08.972 --> 00:31:26.072
Literally, ChatGPT's come out and just gone his mate who had 300 um staff shut his agency, and he said, I because I I spoke to him yesterday, and I was like, How is business like because of what your the stuff that we've been talking about here with tools and things?
00:31:26.072 --> 00:31:29.259
Well, his staff did that like your team do manually.
00:31:29.259 --> 00:31:30.779
They did it manually first, didn't they?
00:31:30.779 --> 00:31:32.220
And they were excellent at it.
00:31:32.220 --> 00:31:36.940
Well, he's got four staff now, and he was like, It's just decimated my business.
00:31:36.940 --> 00:31:43.339
And it I think the world of link building is is fascinating because it's not something that as a PR I've ever got into.
00:31:43.339 --> 00:31:53.980
I've got into content writing and helping clients write excellent blog posts and uh guest posts and press releases and get them featured and get them on TV, get them on broadcast.
00:31:53.980 --> 00:32:05.019
But link building, where we go out and we get links uh off the back of X an excellent campaign and we'll we'll have an excellent asset as part of the campaign, and it'll be about the campaign and about the brand.
00:32:05.019 --> 00:32:13.899
We weren't the sort of agency that would go out and buy a hundred links for 20, 30 bucks or 50 bucks and then point them all to a page.
00:32:13.899 --> 00:32:16.807
Those companies that they're finished, aren't they?
00:32:16.807 --> 00:32:17.180
Really?
00:32:17.180 --> 00:32:26.539
Which is it's just it looks it looks to me like SEO has moved to AI, it has moved to, and I saw another thing that another thing for this context.
00:32:26.539 --> 00:32:31.980
I saw someone on LinkedIn yesterday say, I've now changed my job title to head of organic search.
00:32:31.980 --> 00:32:34.139
I was like, oh, look at this.
00:32:34.139 --> 00:32:40.460
SEO's gone, and it's and it's it's not the death of SEO, but it's it's the evolution, it's evolved again, hasn't it?
00:32:41.659 --> 00:32:47.420
As a marketer, I love this because link building was kind of a hack.
00:32:47.420 --> 00:32:53.339
There was a lot of grey market there of people just going and buying links.
00:32:53.339 --> 00:32:59.819
There was a lot of garbage links on fake websites that were spun up just for links.
00:32:59.819 --> 00:33:08.940
And now what matters is more the not even the quality of the link, the quality of the content that is getting linked back to you.
00:33:08.940 --> 00:33:13.819
So we're we're back to an age where PR really matters and real PR.
00:33:13.819 --> 00:33:23.740
You're not just going and buying a placement, you're actually getting a real story that's talking about your business because what's getting written about you now actually matters.
00:33:23.740 --> 00:33:25.180
So I think this is great.
00:33:25.180 --> 00:33:33.819
We're back to good marketing or back to if you're working with a great PR agency that's getting you real placements, that's gonna matter in the algorithms.
00:33:33.819 --> 00:33:43.339
If you are creating content that's educating your buyer who's actually on the journey of doing research in AI search, then you're going to get rewarded.
00:33:43.339 --> 00:33:45.339
So for me, I think it's great.
00:33:45.339 --> 00:33:50.539
The other thing I notice in these algorithms that they are very good at detecting is differentiation.
00:33:50.539 --> 00:33:53.099
So I come from a product marketing background.
00:33:53.099 --> 00:34:07.099
Uh, so this is the the core of what's now organic that I love, uh, which is that the algorithms, if if it's just same same amongst you and your competitors, they don't know who to recommend because they're all saying exactly the same thing.
00:34:07.099 --> 00:34:15.740
Versus if you we've seen businesses where they we saw a construction business that had a longer warranty than all of their competition.
00:34:15.740 --> 00:34:18.940
And so we built a bunch of content just around this warranty.
00:34:18.940 --> 00:34:29.420
And so AI started recommending them because they had this very strong point of differentiation that could make the that basically showed up in algorithms where like everything looks the same.
00:34:29.420 --> 00:34:30.940
Oh, there's an interesting point here.
00:34:30.940 --> 00:34:33.180
Let's bring it into a summary.
00:34:33.180 --> 00:34:40.220
So the game has, I think the game has changed to quality, even though there's more automation, which is hard to conceptualise.
00:34:40.860 --> 00:34:46.860
Yeah, it's basically it's it's a much more, what it's a more considered it's a more consider like you're right.
00:34:46.860 --> 00:34:50.460
The hacking SEOs will figure some sort of hack out though.
00:34:50.460 --> 00:34:52.700
I can't believe they're not gonna figure it out somewhere.
00:34:52.940 --> 00:34:58.139
I mean, but when this first started, especially I was doing podcasts, and the number the questions they were getting were like, what's the hack?
00:34:58.219 --> 00:34:59.099
What's the hack?
00:34:59.340 --> 00:35:04.619
Uh and people on LinkedIn, this drives me nuts because people on LinkedIn are like, this is the hack, this is the silver bullet.
00:35:04.940 --> 00:35:08.940
You're gonna you're gonna do Reddit for AEO as the silver bullet.
00:35:09.099 --> 00:35:10.860
We just see none of that in the data.
00:35:11.019 --> 00:35:13.179
We don't see any silver bullets.
00:35:13.420 --> 00:35:19.179
We basically see, oh, there's this consistency of training the algorithms, and that's what's working.
00:35:19.259 --> 00:35:27.900
And getting good press is is really moving the needle, but there's no like one hack that's gonna get you ahead of the competition, and people hate that.
00:35:28.219 --> 00:35:29.980
Yeah, they will do, they will hate that.
00:35:30.300 --> 00:35:37.500
I mean, I I like it being in the PR industry because people hopefully more people will be buying PR and come come to us for to get coverage.
00:35:37.739 --> 00:35:40.860
Um I mean the whole thing's fascinating.
00:35:41.099 --> 00:35:44.219
Can I just go back to we when you your previous role then?
00:35:44.300 --> 00:35:45.820
So you worked at Uber Eats.
00:35:46.059 --> 00:35:53.659
Um now you were you were in Uber when it you you you were part of the launch of Uber Eats, is that correct?
00:35:54.059 --> 00:35:57.179
Yeah, so I was the first marketing hire on Uber Eats.
00:35:57.179 --> 00:36:04.139
It was actually kind of fun when I was interviewing because I was basically interviewing to join a team of experiments.
00:36:04.139 --> 00:36:07.259
And there was many experiments uh on that team being run.
00:36:07.259 --> 00:36:11.179
Uber Eats was was one of them, and nobody wanted this role.
00:36:11.179 --> 00:36:29.900
So basically, when I was being during the interview process, they were basically saying no one internally wanted to work on this because the the core rides business at Uber was so strong and growing so quickly that no one that was working on something that was going gangbusters and working wanted to go work on the side experiment.
00:36:29.900 --> 00:36:32.699
And so they were looking for someone that came from a startup background.
00:36:32.699 --> 00:36:35.259
So I came from a startup background.
00:36:35.259 --> 00:36:41.179
So at the time when I joined, Eats was one of the experiments that I was working on.
00:36:41.179 --> 00:36:52.699
And the hypothesis was basically we're we're getting these complaints from drivers that they're not making money during lunch because people are eating during lunch, and so they're not requesting rides.
00:36:52.699 --> 00:36:58.539
And so drivers wanted to figure out they were basically asking us, how can you help us make money during lunchtime?
00:36:58.539 --> 00:37:02.699
So the idea was oh, all these drivers have trunks.
00:37:02.699 --> 00:37:08.619
And so we're gonna fill their trunks with sandwiches or burritos or basically one lunch item.
00:37:08.619 --> 00:37:15.019
And then during lunch in the Uber Rides app, we're gonna surface this for this two-hour period from noon to 2 p.
00:37:15.019 --> 00:37:15.340
m.
00:37:15.340 --> 00:37:19.340
, you can order the item of the day and a driver will deliver it to you.
00:37:19.340 --> 00:37:23.500
So that was our idea to get drivers driving during lunch.
00:37:23.500 --> 00:37:26.460
And that experiment kind of worked okay.
00:37:26.460 --> 00:37:28.380
People were ordering during lunch.
00:37:28.380 --> 00:37:33.911
We started it in San Francisco, which is a very dense downtown, so people at work could pick the item of the day.
00:37:33.911 --> 00:37:37.340
Uh, and a team in Toronto was also running the same experiment.
00:37:37.340 --> 00:37:46.539
The Toronto team came back to us and they said, People they like it okay, but the number one request that we get is they want to order from a restaurant they choose.
00:37:46.539 --> 00:37:48.460
So we are gonna try that.
00:37:48.460 --> 00:37:50.779
And the SF team was kind of like pooh-poohing that.
00:37:50.779 --> 00:37:54.219
They're like, oh, that sounds like a lot of logistics is gonna be difficult.
00:37:54.219 --> 00:37:59.980
And then the Toronto team launched that, and that just went gangbusters.
00:37:59.980 --> 00:38:07.500
It was basically it was so busy they could not find enough drivers during lunch to deliver the food.
00:38:07.500 --> 00:38:10.940
And that's where the team perked up and said, Okay, there's something here.
00:38:10.940 --> 00:38:18.619
They presented it at the time it was Travis Kalanick running the company, and Travis was very supportive of experiments.
00:38:18.619 --> 00:38:25.340
So he was kind of like, I hire smart people and I let them try stuff, and then I see what's working.
00:38:25.340 --> 00:38:34.139
And so the team presented to him, and he was like, team, like, go hard, go grow this, figure this out.
00:38:34.139 --> 00:38:39.179
So then we launched a city per week for probably a year and a half.
00:38:39.179 --> 00:38:40.380
It was wild.
00:38:40.699 --> 00:38:41.579
Crazy.
00:38:42.539 --> 00:38:43.259
Fun experience.
00:38:43.420 --> 00:38:51.579
So with launching a city for that business too, and Eats was even more complicated because on the rides business, it was step one was we need drivers.
00:38:51.820 --> 00:38:53.259
Step two was okay, cool.
00:38:53.340 --> 00:38:56.860
Now we have drivers on the road, people can start requesting a ride.
00:38:57.099 --> 00:38:59.500
In the Eats business, you had one more variable.
00:38:59.659 --> 00:39:01.500
It was step one, we need restaurants.
00:39:01.659 --> 00:39:03.900
And so we'd send someone to that city.
00:39:04.059 --> 00:39:07.980
That person would go door to door and sign up restaurants physically.
00:39:08.059 --> 00:39:08.539
Yep.
00:39:09.019 --> 00:39:14.460
They would literally walk around with their clipboard and sign up restaurants to this delivery service.
00:39:14.699 --> 00:39:20.059
Then they would they would ping me and they'd say, okay, we have the restaurants.
00:39:20.300 --> 00:39:23.659
Now we need more drivers during lunch hour.
00:39:23.820 --> 00:39:31.659
So in a lot of markets, we were getting specific couriers due to licensing, uh, because regular drivers at the time weren't able to deliver food.
00:39:31.820 --> 00:39:34.380
Uh so I would run campaigns to get couriers.
00:39:34.460 --> 00:39:42.940
So this was a lot of job board campaigns, a lot of Craigslist, posting to places, who's looking for a job to deliver food.
00:39:43.179 --> 00:39:46.219
And then once we had enough couriers, we'd say, okay, awesome.
00:39:46.380 --> 00:39:48.860
Turn on the eater marketing.
00:39:49.019 --> 00:39:55.820
Let's go tell people in exactly that location that they can order food in their city, essentially.
00:39:55.980 --> 00:39:58.219
And then we had to figure that out for every single market.
00:39:58.380 --> 00:40:01.019
So one week I'd be working with a team in London.
00:40:01.099 --> 00:40:11.659
The next week it was a team in Tokyo, uh, which was wild because every every city has a slightly different dynamic, different job boards, different places that you market.
00:40:12.300 --> 00:40:17.980
Uh but now it's fun because a lot of people who worked at Uber at that time know my name.
00:40:18.219 --> 00:40:29.739
They not not because they met me, but because they're like, oh, I saw your you were the person with the spreadsheet where I had to go to to fill out to get couriers, or that I was ready to turn on my eater marketing.
00:40:29.900 --> 00:40:36.860
So now even today I'll meet people who were at Uber at that time and they'll be like, oh, Jenna, the girl from the spreadsheet.
00:40:37.659 --> 00:40:38.380
No why?
00:40:38.699 --> 00:40:39.099
Crazy.
00:40:39.259 --> 00:40:47.019
And then then you got the So you were you there when they started with the, I suppose, and then the driverless cars, or is that and they're not doing that there yeah.
00:40:47.259 --> 00:40:48.300
Because there's a lot in Canada.
00:40:48.380 --> 00:40:50.619
That was a lot out that way, isn't there?
00:40:50.940 --> 00:40:52.699
Yeah, a little a little bit later.
00:40:52.860 --> 00:40:57.019
So I left in 2020, so I was there for for four years.
00:40:57.259 --> 00:40:59.420
Uh candidly a little burnt out.
00:40:59.739 --> 00:41:00.300
I did.
00:41:00.539 --> 00:41:01.420
Oh my gosh.
00:41:01.739 --> 00:41:06.139
So I at in 2020, this is like classic Silicon Valley.
00:41:06.300 --> 00:41:09.340
2020, I was like fully vested, my four-year vest.
00:41:09.420 --> 00:41:11.420
It was like, okay, this has been a ton of work.
00:41:11.579 --> 00:41:13.659
I might want to think about what to do next.
00:41:13.980 --> 00:41:19.500
I left Uber and bought a one-way ticket to Asia.
00:41:19.659 --> 00:41:19.900
Wow.
00:41:20.139 --> 00:41:27.259
And I was like, I'm just gonna go to Asia and I'm gonna go travel and I'm gonna figure out what I want to work on next.
00:41:27.500 --> 00:41:35.259
And then I got to Asia in February 2020, and it very quickly dawned on me that it was empty, empty.
00:41:35.420 --> 00:41:41.820
I took a commercial flight from Indonesia to Singapore uh with two of us.
00:41:42.139 --> 00:41:44.380
So so whoa, whoa, whoa, whoa, whoa.
00:41:45.019 --> 00:41:45.980
Hang on a minute.
00:41:46.059 --> 00:41:48.139
I'm just thinking about the media at this point.
00:41:48.380 --> 00:41:55.019
The media at February 2020 in the UK, where when I was in the UK here in the UK, it was all about Spain.
00:41:55.179 --> 00:42:03.019
Was it well, there was a virus in China, and then there was a virus in and it was moving over to Spain, and it was across Europe, and it was in Italy, Spain.
00:42:03.179 --> 00:42:07.179
This was about February time, and I remember us all going, Oh, is this gonna come into UK?
00:42:07.340 --> 00:42:10.059
Or are they gonna lock down this now?
00:42:10.300 --> 00:42:16.940
That must have been happening in America, and you've gone, I know where I need to be, I need to be over there where it's all happening.
00:42:18.139 --> 00:42:18.940
So it was interesting.
00:42:19.099 --> 00:42:24.219
America was like, America was like, this is not happening, and it's not a thing, essentially.
00:42:24.380 --> 00:42:30.380
And so when I bought this ticket, America was like, it's only in China, like rest of the world is fine.
00:42:30.539 --> 00:42:33.900
And so I had a I had a close friend who was working in Japan.
00:42:34.059 --> 00:42:39.820
So I was like, I'm gonna go to Japan and then I'm gonna go to Indonesia and surf and and figure out what to do.
00:42:40.059 --> 00:42:44.059
When I got there, I was like, oh, I've seen the future.
00:42:44.300 --> 00:42:47.579
Uh this is a thing and it is coming.
00:42:47.820 --> 00:42:51.739
Uh and so I was I was gone for about a month.
00:42:51.980 --> 00:42:57.340
I bought some Zoom stock during that time because I was like, I've seen the future.
00:42:57.900 --> 00:42:58.300
Yeah, yeah.
00:42:58.380 --> 00:43:00.699
I was like, I've seen the future and this is coming.
00:43:00.860 --> 00:43:02.619
And I was like, what should we do?
00:43:03.980 --> 00:43:10.460
I was sitting in a sitting in Indonesia at a surf camp, and I was like, everyone's gonna be on Zoom, like this is the future.
00:43:10.619 --> 00:43:13.820
But that's because the surf camp was emptying out, like everyone is going home.
00:43:13.980 --> 00:43:18.059
By the time the Brits were going home, I was like, okay, this is it's time to leave.
00:43:18.219 --> 00:43:27.820
And so I flew back, and I think it was like the fur first of March or just before, and that was right before the the people got stuck at airports.
00:43:27.980 --> 00:43:31.259
So I I was about a week before that happened.
00:43:32.300 --> 00:43:40.059
And then I was kind of re that was probably I think the airport crowds was like the second week of March or first week of March.
00:43:40.380 --> 00:43:42.860
So you were away for three weeks or something then?
00:43:43.179 --> 00:43:45.500
Yeah, I was gone for about a month.
00:43:45.900 --> 00:43:46.219
Okay.
00:43:46.460 --> 00:43:47.179
I mean, it was great.
00:43:47.259 --> 00:43:49.980
Yeah, I was in I was in Asia when it was empty.
00:43:50.219 --> 00:43:53.659
Uh it was there's just never gonna be a time like that.
00:43:53.820 --> 00:44:03.179
But it got to the point where uh my friend in Japan who had lived in Asia for and lived and worked in Asia for a long time, was like, this is gonna be a thing, it's gonna be years.
00:44:03.420 --> 00:44:07.099
And if you're gonna stay here, you're gonna you're gonna wait here.
00:44:07.340 --> 00:44:11.980
So if you decide to stay in Asia, you might be here for three years because you can't fly back.
00:44:12.219 --> 00:44:14.940
Because we've had pandemics before, so we know what this looks like.
00:44:15.099 --> 00:44:17.420
So that's when I was like, okay, playing ticket home.
00:44:17.579 --> 00:44:23.980
And even my friends who were trying to get back to the UK were having trouble getting back to the UK, and I was like, okay, get out here immediately.
00:44:24.380 --> 00:44:25.579
Yeah, blinding.
00:44:25.739 --> 00:44:30.940
The fact that you bought stocks in Zoom, I remember that period was bizarre.
00:44:31.179 --> 00:44:36.539
Um, we we got and um ev luckily all my staff had laptops at that period.
00:44:36.699 --> 00:44:48.300
I'd literally just given the fact because we had desktops before, a few of them had desktops, but I'd rolled them all out a couple of before the pandemic had started, and literally I had the last one the week before the pandemic.
00:44:48.699 --> 00:44:50.219
Pandemic started, it was unusual.
00:44:50.300 --> 00:44:52.860
I mean, because then laptop you couldn't get a laptop anywhere.
00:44:53.019 --> 00:44:55.099
As soon as the pandemic started, they sold out everywhere.
00:44:55.179 --> 00:44:56.300
You couldn't get any laptop.
00:44:56.460 --> 00:44:57.019
Do you remember that?
00:44:57.500 --> 00:45:02.619
And then everybody was jumping on Teams and Zoom, and nobody knew what Teams was because Teams was Skype.
00:45:02.779 --> 00:45:04.300
But I remember Skype fell over.
00:45:04.380 --> 00:45:09.099
The UK's um broadband fell over it on the first day.
00:45:09.340 --> 00:45:11.019
Like it didn't work, it went down.
00:45:11.099 --> 00:45:11.739
It was really bad.
00:45:12.699 --> 00:45:13.259
Unusual.
00:45:13.900 --> 00:45:15.659
And we've just gone, oh yeah, that that happened.
00:45:15.739 --> 00:45:25.019
And and but the weirdest thing about that story is you've gone to Asia where it was coming from and got on planes when you the worst place you can play is on a plane.
00:45:25.420 --> 00:45:30.940
But remember, at the time that I was at the time I was going, the US was like, this is not a thing.
00:45:31.179 --> 00:45:44.860
And even even Europe was like, oh, it's in China, it's starting to appear in some places in Spain and Italy, but it was not a it was not like uh a thing yet.
00:45:45.179 --> 00:45:55.019
So in in fairness, when I flew to Asia, Asia was going through it, but they were like, oh, it's contained, we've had these things before, we'll figure it out.
00:45:55.259 --> 00:45:58.380
It took a while before it started to unfold.
00:45:58.539 --> 00:46:04.139
And before this started to become like when I when I left, I was like, okay, this this is something.
00:46:04.300 --> 00:46:09.579
But when I was telling my friends in the US, as I was in Asia that this was going on, they were like, no, no, no.
00:46:09.980 --> 00:46:13.340
It's fine, it's not coming to the US, nothing's happening, it's no big deal.
00:46:13.659 --> 00:46:23.420
Well, I was much in the US, because compared to the UK, I I had a friend in the I went to see him in New York, and he's because he we did a he did a job swap into London, so he stayed in my house with me.
00:46:23.500 --> 00:46:26.300
That's how we became friends with one of my housemates.
00:46:26.460 --> 00:46:33.019
And uh, I could never get my head round that the England in the UK is smaller than the state of New York and America.
00:46:33.179 --> 00:46:49.420
And so when I think of America, I suppose it is so big and it's so many states that you just think, oh, there's it's so spread out compared to the UK, where we're all densely populated, you know, densely populated in such a small and and so yeah, it was quite a scary, weird, weird time, wasn't it?
00:46:49.579 --> 00:46:51.659
I brought us all together though, I think, globally.
00:46:51.900 --> 00:46:56.460
I think the the Zoom, the gl like who's who's had a conference call since 2020?
00:46:56.619 --> 00:46:59.340
Do you know like an actual do you remember the spiderphones?
00:46:59.659 --> 00:47:00.860
That's true, yeah, yeah.
00:47:01.340 --> 00:47:03.980
They they got they got women killed off, didn't they?
00:47:04.219 --> 00:47:05.179
Zoom killed them off.
00:47:05.420 --> 00:47:10.619
So so I thought I had made the best decision because I was like, oh, I left Uber and Uber's like probably toast.
00:47:10.779 --> 00:47:13.179
And my Uber colleagues were freaking out.
00:47:13.259 --> 00:47:17.900
They're like, our business is gonna be gone because people can't get in a car and ride.
00:47:18.059 --> 00:47:22.300
Like, if people aren't out in the world doing things, this business is like gonna dry up overnight.
00:47:22.460 --> 00:47:24.219
And so I was like, oh, thank goodness I left.
00:47:24.300 --> 00:47:26.139
Like, this business is gonna go down in flames.
00:47:26.380 --> 00:47:28.539
And Uber Eats is what saved it.
00:47:28.779 --> 00:47:44.860
So during the pandemic, people started ordering delivery, and the Uber Eats business just went through the roof and saved the whole business because the rides business was making almost nothing uh for for almost a year, and the Eats business completely saved that company.
00:47:45.179 --> 00:47:47.500
So, what you're saying is you saved Uber.
00:47:49.099 --> 00:47:49.900
Absolutely.
00:47:50.139 --> 00:47:50.380
Yes.
00:47:53.099 --> 00:47:54.779
You know, this brings me to another story.
00:47:55.340 --> 00:48:00.699
So in Silicon Valley, everyone has a story of something they they made and something they missed.
00:48:00.940 --> 00:48:03.420
And this the misses are wild.
00:48:03.820 --> 00:48:14.059
Uh like I have a close friend who invested early, he's an investor, and he invested early in Tesla, but he passed on LinkedIn, he passed on GoPro.
00:48:14.300 --> 00:48:23.980
I have I have another friend who was uh interviewed to be the CTO of Uber very early and didn't get the role, and someone else got the role.
00:48:24.059 --> 00:48:28.940
And he like many years later, he was like, that was a hundred million dollar mistake.
00:48:29.500 --> 00:48:31.820
And so the misses in Silicon Valley are wild.
00:48:31.900 --> 00:48:36.699
So my miss, so may many people say my hit was Uber Eats.
00:48:36.860 --> 00:48:44.460
My my big miss was this is probably six years, maybe seven years ago now.
00:48:44.779 --> 00:48:47.579
I went to a wedding of a a good friend of mine.
00:48:47.659 --> 00:48:50.380
Uh, and he they had their wedding in Morocco.
00:48:50.619 --> 00:49:00.619
And I get to Morocco and I meet this person who is working on AI, which is at the time is seven years ago not really a thing.
00:49:00.860 --> 00:49:03.420
Uh, and he's like, it's gonna be a thing.
00:49:03.739 --> 00:49:05.259
AGI is coming.
00:49:05.500 --> 00:49:12.300
I need to be the one, it's gonna be so powerful that good people need to be working on this.
00:49:12.619 --> 00:49:15.500
This person ends up being Greg Brockman of OpenAI.
00:49:17.739 --> 00:49:28.940
I say to Greg, after spending time with him in Morocco, in the so at one part of the wedding, we decided the most of the wedding guests decided to go to the Sahara Desert.
00:49:29.019 --> 00:49:34.139
Most of us had not seen it, and many people decided, oh, we're gonna go with the bride and groom to the Sahara Desert.
00:49:34.539 --> 00:49:37.500
The bride and groom were like, we're gonna drive.
00:49:37.659 --> 00:49:41.019
Uh, and the the rest of the people had a commercial flight.
00:49:41.179 --> 00:49:45.820
A few of us did not buy a ticket on that commercial flight and decided last minute.
00:49:45.900 --> 00:49:48.139
So we decided to drive with the bride and groom.
00:49:48.219 --> 00:49:50.219
There were there was probably like six of us.
00:49:50.380 --> 00:49:52.539
One of those people was Greg Brockman.
00:49:52.779 --> 00:49:55.659
So I spent hours in a car driving with him.
00:49:55.739 --> 00:49:58.619
And he's telling me about here's the future, here's what we're gonna do.
00:49:58.779 --> 00:50:01.739
And at the end of that, I'm like, you know what?
00:50:02.059 --> 00:50:06.059
I I don't invest much, but I want to give you some money.
00:50:06.219 --> 00:50:07.820
I would like to angel invest.
00:50:07.980 --> 00:50:09.420
I just believe in what you're doing.
00:50:09.579 --> 00:50:10.940
This is super interesting.
00:50:11.179 --> 00:50:12.219
I love the mission.
00:50:12.380 --> 00:50:13.579
And he's like, no, no, no.
00:50:13.900 --> 00:50:20.699
Like, I'm only accepting for money from investors that wanna basically are gonna lose their money.
00:50:20.860 --> 00:50:23.019
And I don't want to take anyone's personal money.
00:50:23.099 --> 00:50:23.259
Yeah.
00:50:23.420 --> 00:50:27.900
And I followed up and I followed up and I sent him emails and I was like, no, I don't care if I lose my money.
00:50:27.900 --> 00:50:30.300
I believe in your mission, like I am inspired.
00:50:30.300 --> 00:50:31.980
And he was like, no, no, no.
00:50:31.980 --> 00:50:35.019
Now I'm like, oh my God, that is the world's biggest miss.
00:50:35.019 --> 00:50:41.929
Like, had I been the first investor in OpenAI, that would have been, I would have a wildly different life today.
00:50:41.929 --> 00:50:44.940
Well, but everyone in Silicon Valley has this story.
00:50:45.259 --> 00:50:48.699
I think the first investor in OpenAI was Elon Musk, wasn't it?
00:50:48.699 --> 00:50:52.860
And um, he even he was, I mean, he's not been very happy about the whole situation.
00:50:52.860 --> 00:50:56.539
The whole thing, that whole story is fascinating.
00:50:56.539 --> 00:51:09.659
How what is happening with some four to six people in in that part of the world where you live is driving the American economy, it's driving the global economy.
00:51:09.659 --> 00:51:21.420
There's clearly a bubble, because everybody know everybody, you know, you know when the dot-com crash, the first one, I'm old enough to remember that, and you could see it, you were like, There's a bubble here, and it's gonna go apart.
00:51:21.420 --> 00:51:38.059
There's definitely a bubble here because there's these businesses are valued at such high rates, they're not making as much, but the s the the technology is so powerful, and we are only seeing you know version six, and they're probably using version nine at the moment, and they're the ones going.
00:51:38.059 --> 00:51:47.099
In fact, I think the head of safety resigned again, another head of safety at OpenAI resigned two weeks ago, or yes, uh very, very recently.
00:51:47.099 --> 00:51:53.420
It really is it's a really exciting time, but a really um worrying time as well.
00:51:53.420 --> 00:51:56.139
I think there's I don't I don't know where it's gonna go.
00:51:56.139 --> 00:52:09.019
It's fascinating because we're all trying to it's it is I I heard this analogy where they said it's like if you're a surfer, you can see a tsunami heading towards you, you're on your surfboard, and you're going, Well, I've got two choices here.
00:52:09.019 --> 00:52:17.179
I can just get totally taken out by the tsunami, or I can turn my surfboard around and try and catch it for as long as I can until it eats me up.
00:52:17.179 --> 00:52:20.539
And I think that's basically where I am with AI right now.
00:52:21.019 --> 00:52:23.340
So Greg was there seven years ago.
00:52:23.659 --> 00:52:31.099
Seven years ago, Greg was this is coming, it's happening, this is a powerful technology, it will be built.
00:52:31.500 --> 00:52:35.980
And why I'm working on it is because if it's not me, someone else will build it.
00:52:36.139 --> 00:52:46.380
And it will be China, it will be Russia, someone's gonna build this because the all of the groundwork is there for this technology to finally accelerate.
00:52:46.539 --> 00:52:53.340
We're finally at the the level of computing, electricity, all of the elements we need for this to happen.
00:52:53.500 --> 00:52:58.300
And he was like, the tsunami's coming, and I'm just gonna start paddling now.
00:52:58.619 --> 00:53:00.300
So he he was there.
00:53:01.500 --> 00:53:04.059
Uh as my dog's just gone crazy.
00:53:04.300 --> 00:53:06.300
Um you've been on the show now.
00:53:06.460 --> 00:53:11.820
Um if you were me, who's the next person you'd invite on this show and why?
00:53:14.059 --> 00:53:17.500
Ooh, this is a great question.
00:53:17.820 --> 00:53:29.019
Um the things the thing that I think is exciting in marketing right now um are the more traditional businesses.
00:53:30.139 --> 00:53:41.900
So I think the kind of how technology is getting applied to small business and traditional businesses that haven't changed in 50 years.
00:53:42.139 --> 00:53:44.539
I I think that stuff's interesting in marketing.
00:53:44.699 --> 00:53:56.860
So I have I have a few Uber colleagues that are working in this space trying to bring traditional businesses into basically how technology businesses run, which is kind of like the frontier.
00:53:57.019 --> 00:53:59.739
Um, I think that's an interesting thing in marketing.
00:53:59.980 --> 00:54:06.139
Um, I would love to see an episode on someone talking about advertising in AI.
00:54:06.779 --> 00:54:08.619
Um, so I'm on the organic side.
00:54:08.699 --> 00:54:13.659
I think seeing what's working in ads, because that's kind of the new frontier of ads.
00:54:13.820 --> 00:54:20.860
Uh I've chatted with a few CMOs in my CMO groups about um running ads in ChatGPT, but it's very new.
00:54:21.019 --> 00:54:21.179
Yeah.
00:54:21.259 --> 00:54:23.659
Uh there's very little tracking, there's little targeting built.
00:54:23.739 --> 00:54:25.179
So I'd be very curious about the code.
00:54:25.340 --> 00:54:38.219
We've seen a lot of people complaining about the fact that ChatGPT ads, you run them and you don't get a lot of conversions out of them, but you get loads of trap organic traffic out of them, and OpenAI is denied that's the thing.
00:54:38.219 --> 00:54:38.940
No, no, no.
00:54:38.940 --> 00:54:41.500
But I've seen multiple people talking about it.
00:54:41.500 --> 00:54:44.059
As soon as you turn the ads off, loads of traffic dries off.
00:54:44.059 --> 00:54:46.539
So it must be there's some sort of correlation between it.
00:54:46.539 --> 00:54:49.204
But it's obviously trying to figure out how they work.
00:54:49.204 --> 00:54:49.710
Okay.
00:54:49.710 --> 00:54:57.340
Um, Jenna, how can people get in touch with you if they want to, you know, they've loved what they've heard um and they want to get in touch with you and use your services?
00:54:57.340 --> 00:54:58.940
What's the best way to get hold of you?
00:54:59.340 --> 00:55:04.219
Yeah, so the best way to find us is at our on our website, which is gethatter.ai.
00:55:04.860 --> 00:55:08.539
They can also find me on LinkedIn, so I'm Jenna Hannon.
00:55:08.699 --> 00:55:14.219
I love talking about AI search, so anyone can can ping or or DM me.
00:55:14.460 --> 00:55:18.539
Um, or they could email us at team at gethatter.ai.
00:55:19.259 --> 00:55:19.659
Excellent.
00:55:19.820 --> 00:55:20.059
Cool.
00:55:20.219 --> 00:55:21.340
Right, great, great, great.
00:55:21.420 --> 00:55:22.460
Thanks for coming on the show.
00:55:23.179 --> 00:55:24.940
Thanks so much for having me, Chris.