전사 🔗
WEBVTT
00:00:00.510 --> 00:00:07.435
I'm saying that, vibe coding and the AI did to marketers, what a spider bite did to Peter Parker.
00:00:07.935 --> 00:00:11.769
Hello everyone, and welcome to the Web3 CMO Stories podcast.
00:00:11.769 --> 00:00:12.951
My name is Joeri Billast.
00:00:12.951 --> 00:00:17.734
I'm your podcast host, and today I'm really excited to be joined by Tim.
00:00:17.734 --> 00:00:18.271
Hey, Tim.
00:00:18.271 --> 00:00:19.038
How are you?
00:00:19.993 --> 00:00:20.574
Hey, Joeri.
00:00:20.574 --> 00:00:21.682
Thanks for having me
00:00:22.111 --> 00:00:25.452
really excited for this conversation, I'm speaking with Tim Soulo.
00:00:25.452 --> 00:00:34.304
He's the CMO at Ahrefs and the author of Discoverable, which is planned to be released in November, if I'm not mistaken.
00:00:34.304 --> 00:00:51.222
Tim, as my listeners know, I always like to dive straight in, I saw on your LinkedIn, you joined Ahrefs when it was, a relatively small company, And you helped it grow into a 150 million plus, ARR business.
00:00:51.222 --> 00:00:57.899
So looking back, what did you believe about marketing back then that you don't believe anymore today?
00:00:58.899 --> 00:01:05.585
Oh, There's a lot of stuff because, when I just joined the company, I didn't have much experience in marketing to start with.
00:01:05.585 --> 00:01:14.543
I didn't have, any knowledge of a CMO or, like, how to help SaaS business grow, so I figured it out as I went.
00:01:14.543 --> 00:01:23.378
And in the process, I had to test a lot of conventional knowledge, a lot of, marketing strategies that were popular at the time.
00:01:23.378 --> 00:01:25.730
some of them worked, some of them didn't.
00:01:25.730 --> 00:01:32.822
eventually I developed, a set of my marketing principles that, I think will stand the test of time.
00:01:32.822 --> 00:01:41.757
I think they didn't change in the past, 10 years that I've been doing marketing, and I'm already seeing how they are surviving the AI apocalypse.
00:01:41.757 --> 00:01:44.039
as you mentioned, I wrote a book about this.
00:01:44.039 --> 00:01:48.406
I actually just received the first author copy, from the printers.
00:01:48.406 --> 00:01:50.099
it's called Discoverable.
00:01:50.099 --> 00:01:52.082
It would be out, in, November.
00:01:52.082 --> 00:02:09.693
And yeah, I'm super excited because I originally was telling everyone that it took me two years to write this book, but then I opened the Google document, which I started when I was putting my ideas and brainstorming the angles and titles for the book, and the document is actually four years old.
00:02:09.693 --> 00:02:24.200
So, like, from coming up with an idea to write a book and starting collecting, ideas, stories, information, and up to receiving the hard copy in the mail, it took me four years, which is quite crazy
00:02:25.069 --> 00:02:26.836
my first book took three years.
00:02:26.836 --> 00:02:29.383
My second book actually took, two months.
00:02:29.383 --> 00:02:30.673
Now, Ahrefs
00:02:31.891 --> 00:02:32.740
ahrefs
00:02:33.080 --> 00:02:34.647
Okay, Ahrefs
00:02:34.661 --> 00:02:35.580
That's a common thing.
00:02:35.580 --> 00:02:36.611
You shouldn't cut this out.
00:02:36.611 --> 00:02:37.615
that's a common thing.
00:02:37.615 --> 00:02:40.265
People always ma- make fun of our pronunciation
00:02:40.546 --> 00:02:40.856
Ahrefs.
00:02:40.856 --> 00:02:43.403
Yeah, I never heard it like that actually.
00:02:43.403 --> 00:02:50.961
So, Ahrefs grew without VC funding for years without outbound sales.
00:02:50.961 --> 00:03:01.943
So what did that force you to understand about demand and brand building that companies with large sales and advertising budgets sometimes never learn?
00:03:02.943 --> 00:03:06.763
I wouldn't say that I was forced to understand something.
00:03:06.763 --> 00:03:11.525
I just didn't know that companies can grow with outbound.
00:03:11.525 --> 00:03:33.683
I didn't realize that you can pinpoint, your ideal customers, you can stalk them on LinkedIn, you can send them emails, you can invite them to dinners I didn't know that this is a possibility because when I joined the company, there were 16 people, like the entire team was 16 people, I was the only marketer.
00:03:33.683 --> 00:03:38.223
We didn't have, any salespeople, just engineers and product people.
00:03:38.223 --> 00:03:48.197
And, I just couldn't spend my time going on dinners or stalking people on LinkedIn and inviting them to sales calls.
00:03:48.197 --> 00:03:49.947
And then again, sales calls with who?
00:03:49.947 --> 00:03:50.709
Like, with me.
00:03:50.709 --> 00:03:55.587
I would be a marketer, I would be a salesperson, SDR, BDR, everything.
00:03:55.587 --> 00:03:58.479
So I needed to do things that scale.
00:03:58.479 --> 00:04:09.657
I was the only person, and I needed to do things that would bring me not dozens of customers per month, but hundreds, better yet, thousands.
00:04:09.657 --> 00:04:17.533
So for example, as of today, over 20,000 people sign up for Ahrefs every single week.
00:04:17.533 --> 00:04:21.939
a lot of these signups are kind of people canceling and then returning.
00:04:21.939 --> 00:04:26.865
Some of the signups are people, getting more users within their plan and inviting.
00:04:26.865 --> 00:04:33.901
But still, the total number of new signups per week is 20,000, and you don't get there with outbound.
00:04:33.901 --> 00:04:36.055
You need to do marketing.
00:04:36.055 --> 00:04:41.009
You need to make people discover you rather than you going after your clients.
00:04:41.009 --> 00:04:48.369
You need to make sure that anyone who has the kind of problem that you're solving, that they can find you.
00:04:48.369 --> 00:04:50.651
this is the central idea of my book.
00:04:50.651 --> 00:04:55.055
There are only three ways how people discover products and businesses.
00:04:55.055 --> 00:05:04.571
One of them is search, which has been our bread and butter at Ahrefs
Load full transcript
for the entire time, because Ahrefs started as an SEO tool for search engine optimization.
00:05:04.571 --> 00:05:07.093
But today, search is a lot more than just Google.
00:05:07.093 --> 00:05:10.281
It is ChatGPT, it is TikTok, it is YouTube.
00:05:10.281 --> 00:05:17.932
People use all sorts of tools and platforms and AI assistants to do their searches, and you need to be there.
00:05:17.932 --> 00:05:26.249
You need to show up in those searches, because if they have a problem and they search for it, if you're not showing up, your competitors are, and they're getting all those clients.
00:05:26.249 --> 00:05:32.204
The second way how people discover products and businesses is they simply notice them.
00:05:32.204 --> 00:05:38.963
You're minding your own business, you're scrolling your Instagram or Facebook or LinkedIn, and something pops into your attention.
00:05:38.963 --> 00:05:41.086
How can it pop into your attention?
00:05:41.086 --> 00:05:53.165
It's either advertising, which is one of the very fundamental things that marketers do, you just advertise, or you create the kind of content that gets spread organically.
00:05:53.165 --> 00:06:00.733
So you publish a YouTube video, it performs very well, and in that video, you talk about your business, about your product, and people learn about it this way.
00:06:00.733 --> 00:06:09.767
So they notice your video because it was good or because you advertise it And the third way how people discover products and businesses is through recommendations.
00:06:09.767 --> 00:06:11.208
So this is word of mouth.
00:06:11.208 --> 00:06:19.846
This is when your happy customers or people who know you and respect you or they have learned something from you, they pass on the knowledge to someone else.
00:06:19.846 --> 00:06:25.156
And again, simple, simple framework of, how to make people talk about you.
00:06:25.156 --> 00:06:29.658
There are kind of two psychological reasons why people would talk about you.
00:06:29.658 --> 00:06:31.595
First of all, they want to brag.
00:06:31.595 --> 00:06:35.146
Like, "Oh, like I'm using this product, I'm getting these good results.
00:06:35.146 --> 00:06:38.374
Look at me, like my website's getting a lot of traffic.
00:06:38.374 --> 00:06:39.653
I'm getting a lot of clients.
00:06:39.653 --> 00:06:44.769
You should also use this service." So they just want to brag about how successful they are.
00:06:44.769 --> 00:06:46.795
Or they want to be useful.
00:06:46.795 --> 00:06:51.845
So they have a friend who is not doing well with something, and they want to help him.
00:06:51.845 --> 00:06:57.173
So these are kind of the two primary motivators for people to talk about things.
00:06:57.173 --> 00:07:08.867
Either they want to brag, "Look at my shoes, look at my motorcycle," whatever, like, or, "Look at how, like I'm solving business problems with this tool." Or, "You have a problem?
00:07:08.867 --> 00:07:10.704
I actually know how to solve it.
00:07:10.704 --> 00:07:21.896
Like my friend tried this software." I might not even use the software or product myself, but if I know from somewhere else that the product is useful, if I have like validation for it, I want to pass on this knowledge.
00:07:21.896 --> 00:07:28.637
So these are some of the things that I'm discussing with the book, like the fundamental principles of how to make your business discoverable
00:07:29.000 --> 00:07:29.529
Yeah.
00:07:29.529 --> 00:07:30.901
But I love that actually.
00:07:30.901 --> 00:07:41.483
So when I wrote my first book, "The 5K Challenge," I also mentioned why do people share your post, It's because it makes them look good or indeed, you know, there is a solution to a problem that people might have.
00:07:41.483 --> 00:07:43.862
Now, I love that you mentioned the fundamentals.
00:07:43.862 --> 00:07:45.968
we have a common friend, Mark Schaefer,
00:07:46.857 --> 00:07:47.440
Hmm.
00:07:47.980 --> 00:07:48.459
and
00:07:48.600 --> 00:07:50.211
book on your shelf, Audacious.
00:07:50.396 --> 00:07:50.608
see
00:07:50.661 --> 00:07:51.235
a good book
00:07:51.334 --> 00:07:55.678
I was also one of the authors of the, the most amazing marketing book ever, actually.
00:07:55.678 --> 00:08:03.059
I was, I am one of the co-authors, and actually, he also was the person that give me, the boost to write my latest book.
00:08:03.059 --> 00:08:17.875
I know that Mark, spoke at, Ahrefs Evolve, on your event, And you mentioned it, I think, in one of your LinkedIn posts precisely because it was not about SEO or prompts or citations, but it was about fundamental marketing principles.
00:08:17.875 --> 00:08:26.579
why do marketers keep gravitating towards tactics when the fundamentals are often where the real advantage lies, you think?
00:08:27.579 --> 00:08:29.351
there are multiple reasons to that.
00:08:29.351 --> 00:08:32.134
one of the reasons is specialization.
00:08:32.134 --> 00:08:41.706
So as marketing organization starts growing, you need people who don't just look at marketing holistically, but who need to specialize.
00:08:41.706 --> 00:08:51.668
For example, I don't have a lot of knowledge in how to run Facebook ads or how to run Google ads, how to configure campaigns, I don't have this knowledge.
00:08:51.668 --> 00:08:56.664
I know just general information of how it works, but I don't have specialized knowledge.
00:08:56.664 --> 00:08:59.610
But I have people on my team who take care of that.
00:08:59.610 --> 00:09:11.234
So basically, my book with fundamentals is for people who are kind of aspiring to be leaders, who need to understand the holistic view of everything rather than going deep on any discipline.
00:09:11.234 --> 00:09:16.480
But, in the early days, you still, cannot start your job as a leader, typically.
00:09:16.480 --> 00:09:20.544
you need to grow in your ranks and understand how, all the bits and pieces work together.
00:09:20.544 --> 00:09:26.063
And this is when you need to be a specialist, and
this is when you need to have some specialist knowledge.
00:09:26.063 --> 00:09:32.721
And this is why people kind of get carried away with, very specific tactics, because this is just their role.
00:09:32.721 --> 00:09:51.287
And there is nothing wrong about it, because when I was also starting out at Ahrefs, and we needed to do a lot of SEO, I was also very deep in the very specific tactics of SEO, how to pick keywords in the right way, how to build backlinks, how to create content to address those keywords, et cetera, So it's quite natural.
00:09:51.287 --> 00:09:57.091
But then as my team grew, I needed a more holistic understanding of marketing to be able to lead this team
00:09:57.875 --> 00:09:58.346
Yeah.
00:09:58.346 --> 00:10:06.408
you mentioned the content and, and le- let's talk about what's, to something you said about AI, use cases.
00:10:06.408 --> 00:10:11.721
I think you said that AI use cases in marketing are not particularly good, I think most of them.
00:10:11.721 --> 00:10:17.476
What are marketers fundamentally misunderstanding about where AI creates the value?
00:10:18.476 --> 00:10:35.953
Yeah, I was skeptical about AI for a very, very long time, because, while a lot of people were kind of seeing the potential and looking into the future, I had a more conservative look and I was like, "No, guys.
00:10:35.953 --> 00:10:46.227
First I need to see it actually do something cool, and then I would be a believer." And, in this last year or maybe two years, AI has advanced a lot.
00:10:46.227 --> 00:10:52.418
So I transformed from AI skeptic to AI believer, AI advocate essentially.
00:10:52.418 --> 00:10:55.426
I use AI all the time in my work.
00:10:55.426 --> 00:10:56.976
Like, there's a lot of use cases.
00:10:56.976 --> 00:11:02.822
one of the most common use cases of AI for anyone is actually coding or as they say vibe coding.
00:11:02.822 --> 00:11:06.482
Because as marketers, we typically have a lot of ideas.
00:11:06.482 --> 00:11:13.057
We have ideas for landing pages, we have ideas for tools, we have ideas for products, for campaigns, Like, what to do with data.
00:11:13.057 --> 00:11:15.277
A lot of ideas are coming from the marketing department.
00:11:15.277 --> 00:11:24.052
But you don't necessarily have the skill, so if you have an interesting tool that you want to build, you don't have the coding skill to build it, and you don't have the time to do it.
00:11:24.052 --> 00:11:27.644
But now with AI You can just do it.
00:11:27.644 --> 00:11:40.649
Like, the first tool that I built with AI, and at first I thought it wouldn't be possible, and it turned out it would be, it was, I built a simple tracker of LinkedIn audience, of different people.
00:11:40.649 --> 00:11:45.875
So I, like, I saw that my audience on LinkedIn is growing, but I didn't know how fast.
00:11:45.875 --> 00:11:56.650
So I didn't want to track in a spreadsheet, like, how many new followers I'm adding in a week, and especially I didn't want to do it
for some other people in the industry because I wanted some benchmarks.
00:11:56.650 --> 00:12:06.541
I wanted to understand, are some other people in my industry who are also, like, posting on LinkedIn, are they growing their following faster than me or slower than me?
00:12:06.541 --> 00:12:10.018
And are they posting more than me, or are they posting less?
00:12:10.018 --> 00:12:12.606
Are they getting more engagements or less engagements?
00:12:12.606 --> 00:12:19.683
I need a benchmark to understand how well I am doing because I don't get this in isolation when I just look at my performance.
00:12:19.683 --> 00:12:28.056
So originally, I actually I hired someone on Upwork to start building this tool for me just for my personal use.
00:12:28.056 --> 00:12:30.878
I wasted maybe $5,000.
00:12:30.878 --> 00:12:42.404
And then while I was in the process of working on this very simple tool with this person, AI got so good that you could just build this thing with a few prompts.
00:12:42.404 --> 00:12:53.828
And right now, AI at the stage where you can just take a link URL of my tool, you can feed it to AI and say, like, "Build it for me." And it would one shot.
00:12:53.828 --> 00:12:55.725
You don't have to provide any instructions.
00:12:55.725 --> 00:13:00.044
AI would essentially figure out everything it needs to build this kind of tool.
00:13:00.044 --> 00:13:01.870
So it's crazy how good it is.
00:13:01.870 --> 00:13:04.476
You just need to have, good ideas of what to build.
00:13:04.476 --> 00:13:11.505
I'm saying that, vibe coding and the AI did to marketers, what a spider bite did to Peter Parker.
00:13:11.505 --> 00:13:13.365
It turned us into,
00:13:14.365 --> 00:13:14.773
Yeah.
00:13:14.773 --> 00:13:20.085
Indeed, I think you, you've written that AI can increase productivity enormously, and you mentioned ideas.
00:13:20.085 --> 00:13:27.475
You need ideas, taste, personality, and effort, of course, to determine, whether the marketing is actually good.
00:13:27.475 --> 00:13:36.227
do you think is-- which one of those is the most valuable, as AI capa-capabilities become commoditized, everything becomes better and better?
00:13:37.227 --> 00:13:48.597
Yeah, I recently posted that, AI multiplied by zero effort equals slop, and AI multiplied by maximum effort equals gold.
00:13:48.597 --> 00:14:07.243
So one of the most common ways that people misuse AI is they think that they should avoid effort, which is actually true for vibe coding, because when I ask AI to build something for me, it avoids all the effort of writing all the code.
00:14:07.243 --> 00:14:09.534
But the thing is, I just need an application.
00:14:09.534 --> 00:14:33.589
but when you apply the same thing for creating content that should get you noticed, one of the way to get noticed is to create content that would spread naturally, and people say, write me a good article about this," or, "Turn my article into a LinkedIn post," and they publish it as is, avoiding all the effort of actually making it great, and this just doesn't work.
00:14:33.589 --> 00:14:43.245
Because what people notice, and to get noticed is one of the fundamental principles of being discoverable, what people notice is effort.
00:14:43.245 --> 00:14:50.176
If I see that you didn't put in any effort into what you're doing, I'm not interested in that.
00:14:50.176 --> 00:15:00.941
So this is why I'm saying that with my book Discoverable, it took me four years to write it, I started collecting ideas four years ago.
00:15:00.941 --> 00:15:03.034
It wasn't a simple, fast process.
00:15:03.034 --> 00:15:35.690
And even more than that, because I was tracking all my progress with AI, and it has like all the revisions and everything, I asked AI, "Okay, look at all the notes, all the stories and the ideas that didn't make it to the book, but are still in our kind of, separate folders." And it said, "Oh, like you actually wrote two books, but shipped one." Because the total word count of everything I wrote was over 100,000 words, and I shipped right around 55,000 words or something.
00:15:35.690 --> 00:15:39.355
So I cut all the content that I've produced in half.
00:15:39.355 --> 00:15:40.265
this is effort.
00:15:40.265 --> 00:15:46.371
And I would imagine that this is why people would want to read my book, that there's a lot of effort into it.
00:15:46.371 --> 00:15:51.436
if I would say that, I've been doing my podcast for a year.
00:15:51.436 --> 00:15:53.375
I interviewed 60 people.
00:15:53.375 --> 00:15:56.279
I used AI to turn it into a book.
00:15:56.279 --> 00:15:57.043
Go read it.
00:15:57.043 --> 00:16:01.820
I don't think anyone would be interested to read it, because they would understand that it's slop.
00:16:01.820 --> 00:16:17.366
Even though there's work in interviewing 60 people, the fact that I simply asked AI, "Turn those 60 interviews into a book," and it did it like in two hours and they shipped the book, that's not effort, and that's not noticeable.
00:16:17.366 --> 00:16:19.370
That's not something people would pay attention to.
00:16:19.759 --> 00:16:22.173
But that's interesting because, I did something in between.
00:16:22.173 --> 00:16:28.130
I would say my first book took, so many years to write it, to rewrite it, to change the language, to restructure that.
00:16:28.130 --> 00:16:41.974
And then my last book, because building the podcast was effort, more than 300 episodes, I have a lot of content in there, and I have my keynote, and then I asked AI to use, the podcast episodes as a, I would say,
00:16:42.434 --> 00:16:42.774
you did.
00:16:42.774 --> 00:16:43.437
Interesting.
00:16:44.312 --> 00:16:48.106
as examples to support the keynote that I gave.
00:16:48.106 --> 00:16:50.020
So I said, "These are the ideas that I see.
00:16:50.020 --> 00:16:59.552
AI, do you see patterns or do you agree with that?" So this was not AI writing the book, but it supported me to add content where it would make sense.
00:16:59.552 -->
00:17:08.732
and the interesting thing was, that I sent it to Philip Kotler, you know, the guy who popularized, the four Ps in marketing, and he said it was the best thing he ever read about CMOs.
00:17:08.732 --> 00:17:20.066
So I said, "Okay," so this, there is a way to make it work between, doing, having AI doing everything with zero effort and putting a lot of work, and then going there in between.
00:17:20.066 --> 00:17:27.358
I could not do this three years ago, but now, with the possibilities of AI, and if you are, working together with them, it could work.
00:17:27.358 --> 00:17:30.715
But that's just, to have like, an contrarian opinion.
00:17:31.715 --> 00:17:36.750
I don't think it's contrarian because I didn't say that I didn't use AI to write my book.
00:17:36.750 --> 00:17:40.444
I used AI profusely, and I kind of did the same thing as you did.
00:17:40.444 --> 00:17:45.429
For example, I interviewed someone and I don't really remember their story exactly.
00:17:45.429 --> 00:17:54.930
So I would ask AI, "Go search the transcript, find this story, write it in a few paragraphs for me." So yeah, I don't want to re-listen to the entire interview.
00:17:54.930 --> 00:17:57.934
I don't want to change the words into a story.
00:17:57.934 --> 00:17:59.739
I can have AI do this for me.
00:17:59.739 --> 00:18:04.211
But still, how much effort is in the book?
00:18:04.211 --> 00:18:06.981
How many hours you spent polishing your ideas?
00:18:06.981 --> 00:18:09.378
how much time you spent collecting your ideas?
00:18:09.378 --> 00:18:10.586
these are different things.
00:18:10.586 --> 00:18:24.634
Because like I said, if what I did, if I'm true with people who are buying my book and I tell them, "Listen, this book is simply 50 podcast episodes rewritten by AI into a book.
00:18:24.634 --> 00:18:25.636
I didn't touch it.
00:18:25.636 --> 00:18:28.750
I didn't even read it myself," No one would buy it.
00:18:28.750 --> 00:18:32.928
But if I tell them, "Guys, I was working on this for four years.
00:18:32.928 --> 00:18:48.575
Yeah, I used AI to pull things here and there, but like I read my own book maybe 15 times, maybe more, to polish like every freaking sentence, to obsess about like everything," people would want to read it because they understand, like I've put in effort.
00:18:48.575 --> 00:18:51.362
It's also like it's one of the concepts that I'm discussing in the book.
00:18:51.362 --> 00:18:53.130
it's a form of signaling.
00:18:53.130 --> 00:18:57.074
because why do peacocks have this huge tail?
00:18:57.688 --> 00:18:57.815
Mm-hmm.
00:18:57.888 --> 00:19:10.474
This is signaling, "I am so fit, I am so well fed that I can get around with this unwieldy, completely unnecessary thing under my back." This is signaling.
00:19:10.474 --> 00:19:14.186
Like, it's the same like my effort into the book.
00:19:14.186 --> 00:19:17.589
I'm putting it because I'm signaling that this is good work.
00:19:17.589 --> 00:19:18.762
I know that it is good.
00:19:18.762 --> 00:19:20.805
I'm willing to invest effort into it.
00:19:20.805 --> 00:19:25.633
So effort, I still think that effort is what sells.
00:19:25.633 --> 00:19:26.737
It's what gets you noticed.
00:19:26.737 --> 00:19:28.182
It what gets you respect
00:19:28.654 --> 00:19:29.108
Yeah.
00:19:29.108 --> 00:19:34.433
I think it's good to put effort indeed in the right place, you know, where, you need the human, effort,
00:19:34.463 --> 00:19:35.096
Of course.
00:19:35.096 --> 00:19:35.435
Yeah.
00:19:35.435 --> 00:19:35.933
Yeah
00:19:35.974 --> 00:19:40.375
the value and the things that is actually, just spending time having AI do.
00:19:40.375 --> 00:19:43.395
So I think we agree upon that, but it's interesting writing a book.
00:19:43.395 --> 00:19:48.598
It's an adventure and you find out, things along as you go, now I have a few other interesting questions for you.
00:19:48.598 --> 00:19:58.032
a lot of marketers are, currently measuring AI visibility, citations, mentions, but being cited is not necessarily the same as being recommended.
00:19:58.032 --> 00:20:04.913
What would make you believe a brand that is generally winning in AI, driven discovery?
00:20:05.913 --> 00:20:20.381
so yeah, there is a lot of holy wars happening on LinkedIn on the topic of AI visibility, which prompts you have to track, what kind of prompts you have to track, how many, how often should you check them?
00:20:20.381 --> 00:20:24.083
Should you check them across all models or just a few models?
00:20:24.083 --> 00:20:32.820
Should you track them via API or, like, logged, There, there's so many options, it's kind of crazy, and people keep bickering about it.
00:20:32.820 --> 00:20:39.654
My perspective is that you should track your success in AI search more holistically.
00:20:39.654 --> 00:20:51.229
And when I'm saying more holistically, there are more indicators that you're performing well in AI search rather than whatever AI visibility tools are telling you.
00:20:51.229 --> 00:20:56.414
And I'm one of the people behind one of these AI visibility tools, actually.
00:20:56.414 --> 00:20:59.328
So I'm kind of, in a way, shooting myself in the foot.
00:20:59.328 --> 00:21:01.557
But so what are those things?
00:21:01.557 --> 00:21:05.622
First of all, referral traffic from AI chatbots.
00:21:05.622 --> 00:21:09.405
You know if ChatGPT is sending you traffic or not.
00:21:09.405 --> 00:21:10.219
that's objective.
00:21:10.219 --> 00:21:16.150
And you know if this traffic is going up month to month or if it is going down or if it's not changing.
00:21:16.150 --> 00:21:17.877
It is very easy to measure.
00:21:17.877 --> 00:21:23.140
Every analytics tool, our own free web analytics tool, every analytics tool will show you.
00:21:23.140 --> 00:21:27.431
Second thing, visits from AI bots.
00:21:27.431 --> 00:21:33.913
So whenever AI needs to pull information from your page, AI bot has to visit your page.
00:21:33.913 --> 00:21:49.397
So again, at Ahrefs we have bot analytics, but there are many other bot analytics systems that you can install, and you can see if AI bots from, like, OpenAI, Claude, et cetera, if they're visiting the pages of your website and which pages they're visiting and how often.
00:21:49.397 --> 00:21:56.982
Because this is a good indication that your pages are getting considered Next, self-attribution.
00:21:56.982 --> 00:22:04.522
If people are saying that, "I found you in ChatGPT, I found you in Claude," you can have a simple form, how did you hear about us?
00:22:04.522 --> 00:22:05.821
People will leave feedbacks.
00:22:05.821 --> 00:22:07.213
You can cross-reference it.
00:22:07.213 --> 00:22:08.461
Is it exact science?
00:22:08.461 --> 00:22:09.155
No.
00:22:09.155 --> 00:22:12.933
Like, people are prone to a lot of biases when they're, leaving those feedbacks.
00:22:12.933 --> 00:22:17.243
They might just not care and just write ChatGPT while they actually found you in ads.
00:22:17.243 --> 00:22:18.660
That doesn't matter.
00:22:18.660 --> 00:22:20.472
Still, you would notice trends.
00:22:20.472 --> 00:22:30.726
if your traffic from AI chatbots is going up, if the bot visits are going up, and you also see the same thing in your self-attribution, then it has to be correct.
00:22:30.726 --> 00:22:31.981
And again, sales calls.
00:22:31.981 --> 00:22:39.912
if you are in a company that does a lot of sales calls, and your salespeople are asking people, "How did you heard about us?" And they're like, "Oh, you're all over ChatGPT.
00:22:39.912 --> 00:22:43.585
Whenever I ask about my problem, you always show up," you can easily track that.
00:22:43.585 --> 00:22:58.432
And finally, on top of this, you have AI visibility, where you work on figuring out what would be a great, objective, unbiased list of prompts that you want to track, and there are, like, a lot of methodologies on, how to do it.
00:22:58.432 --> 00:23:00.304
We teach some of that on our blog.
00:23:00.304 --> 00:23:04.392
You set it up, and again, it's not about the actual number.
00:23:04.392 --> 00:23:11.355
Like, what is it today, your AI visibility is 7.5% or 15.3%?
00:23:11.355 --> 00:23:12.065
I don't care.
00:23:12.065 --> 00:23:14.950
Tell me how it progresses month to month.
00:23:14.950 --> 00:23:18.616
If it was 5% yesterday, will it be 6% tomorrow?
00:23:18.616 --> 00:23:20.743
And what is it for my competitors?
00:23:20.743 --> 00:23:22.567
Do they have more visibility than...
00:23:22.567 --> 00:23:28.602
So the actual numbers are not as relevant as the trends and comparisons to your competitors.
00:23:28.602 --> 00:23:30.869
So this is how I'm thinking of AI search.
00:23:30.869 --> 00:23:39.672
And finally, the most important thing is that AI search is essentially a reflection of the web.
00:23:39.672 --> 00:23:45.593
Because, like, today, AI is pulling their citations from these websites.
00:23:45.593 --> 00:23:47.438
Tomorrow, they're pulling from somewhere else.
00:23:47.438 -->
00:23:52.709
At some point, Reddit was, the source where AI pulled a lot of information.
00:23:52.709 --> 00:23:55.134
Then it dropped, in its visibility.
00:23:55.134 --> 00:23:56.843
Then YouTube, then LinkedIn.
00:23:56.843 --> 00:24:00.221
It changes all the time because those AI systems are adapting.
00:24:00.221 --> 00:24:06.364
What doesn't change is that they get information from reputable sources around the web.
00:24:06.364 --> 00:24:12.611
So what you need to do, you need to track your mentions on the reputable sources around the web.
00:24:12.611 --> 00:24:18.777
Again, Ahrefs has tools for that, which is why I think we're super well-positioned, in this new, AI world.
00:24:18.777 --> 00:24:28.118
But in addition to everything that I just discussed, the actual source of what AI says about you is the internet.
00:24:28.118 --> 00:24:32.832
So you need to map what the internet says about you and control that.
00:24:32.832 --> 00:24:41.547
So basically, whenever someone publishes an article on a reputable website and they make a mistake about, for example, your pricing, which literally happened to us.
00:24:41.547 --> 00:24:44.201
We made a mistake when we released our new product.
00:24:44.201 --> 00:24:46.311
Our pricing was kind of convoluted.
00:24:46.311 --> 00:24:48.017
It was a little bit hard to figure out.
00:24:48.017 --> 00:24:54.965
And some people, when they published articles, they made false statements, not because they wanted to harm us, but because they misunderstood.
00:24:54.965 --> 00:25:01.546
And then we had to go around, reach out to these people, and ask them to fix because they misunderstood our pricing
00:25:48.706 --> 00:25:49.792
Yeah, it's interesting.
00:25:49.792 --> 00:25:53.473
we have another person that we know in common, Andy Crestodina.
00:25:53.473 --> 00:26:05.060
He was on my podcast not so long ago, and he recently mentioned that your AI traffic research, he's also a numbers guy, when we discussed how AI is changing buyer intent.
00:26:05.060 --> 00:26:17.396
So you already explained a bit, but do you think AI is simply becoming another discovery channel, or is it really fundamentally changing how people evaluate and choose brands?
00:26:18.396 --> 00:26:27.723
To be honest, I struggle to answer this because I think people were using Google in a very similar way.
00:26:27.723 --> 00:26:29.988
People were using YouTube in a very similar way.
00:26:29.988 --> 00:26:35.451
So you are choosing a backpack for hiking.
00:26:35.451 --> 00:26:39.144
You go to Google, you read a bunch of reviews.
00:26:39.144 --> 00:26:44.399
Those reviews help you shape your perspective of what a good backpack for hiking looks like.
00:26:44.399 --> 00:26:48.250
Or you go to YouTube, someone has like 10 backpacks.
00:26:48.250 --> 00:26:53.413
They show them to you, they review, they like put stuff in it to show you how convenient it is.
00:26:53.413 --> 00:26:55.594
You make up your mind.
00:26:55.594 --> 00:27:06.172
Maybe what is happening is that with AI, people get a little lazier and they want to skip some of that research work.
00:27:06.172 --> 00:27:27.322
But I would argue that they only do it when the decision is not high stakes for them So if you don't really care that you're buying the best product or you don't really care that you're getting the best deal then you might just go for whatever AI is offering you.
00:27:27.322 --> 00:27:37.509
But when you have some high stakes product that, for example, your entire company would depend on, then you would still have to do a lot of research even with AI.
00:27:37.509 --> 00:27:40.196
You will be asking a lot of additional questions.
00:27:40.196 --> 00:27:45.682
If it would say that people on Reddit are saying this, you would actually go and review those Reddit threads yourself.
00:27:45.682 --> 00:27:51.750
So you would still do a lot of work, and I'm not seeing how it is different to what people were doing before.
00:27:51.750 --> 00:28:00.974
And finally, I also have a very interesting take in regards to shopping with AI, because a lot of people are saying that AI would do all the shopping for us.
00:28:00.974 --> 00:28:03.132
I again disagree with this.
00:28:03.132 --> 00:28:06.925
So first of all, there are like low stakes decision and there are high stakes decision.
00:28:06.925 --> 00:28:17.890
And for high stakes decision, you cannot just delegate it to AI because, you are responsible But then there are things, like a lot of people like shopping.
00:28:17.890 --> 00:28:27.575
A lot of people like the whole process of studying different products, how they're different, making this decision.
00:28:27.575 --> 00:28:32.180
So shopping is in many ways a pleasurable activity.
00:28:32.180 --> 00:28:44.441
And for example, if, my wife wants to get herself a new dress, and she wants to go to the shops, she wants to, try them on, she wants to take pictures, send to me, send to her girlfriends.
00:28:44.441 --> 00:28:47.475
It's a whole kind of activity that she enjoys doing.
00:28:47.475 --> 00:28:55.098
Why would she say to AI, "I have, someone's birthday coming up and they want to do, like, I don't know, white party.
00:28:55.098 --> 00:28:56.865
Get me some trendy white dress.
00:28:56.865 --> 00:29:05.307
I don't care what it is." So there, there's a lot of nuance to how AI is going to reshape, our habits and our activities.
00:29:05.753 --> 00:29:06.167
Yeah.
00:29:06.167 --> 00:29:12.056
We are coming towards the end of the podcast episode, but I've, another question about your book, Tim.
00:29:12.056 --> 00:29:20.711
One idea in Discoverable that I found particularly interesting is getting found, getting noticed, and getting talked about.
00:29:20.711 --> 00:29:28.724
for an entrepreneur with limited resources, how do you decide which of those, deserves attention first?
00:29:29.724 --> 00:29:51.613
Yeah, I actually have a framework in my book for making marketing decisions, the framework is
called Brew, B-R-E-W, where B stands for business potential, which is essentially how likely is the person to become your customer, to purchase your product after they consume a piece of content that you're giving them.
00:29:51.613 --> 00:29:56.908
So if you published an article, how likely is a person to purchase your product after reading an article?
00:29:56.908 --> 00:29:58.254
So that's business potential.
00:29:58.254 --> 00:30:00.784
The second in Brew is reach.
00:30:00.784 --> 00:30:02.118
R stands for reach.
00:30:02.118 --> 00:30:06.137
So how many people are you going to reach with that piece of content?
00:30:06.137 --> 00:30:12.071
in my book I made an example of, case studies, which is, my common argument with my marketing team.
00:30:12.071 --> 00:30:19.807
They want to do more case studies, and I'm like, "But how are we going to get people to see those case studies?" Case studies seems like a sales tool.
00:30:19.807 --> 00:30:30.152
So when a salesperson is on a call with someone, they would show them, "Oh, we have these clients, they're reaching these results," or maybe you just feature like one or two case studies on your homepage, but why do you need more?
00:30:30.152 --> 00:30:30.980
What is the reach?
00:30:30.980 --> 00:30:36.383
How are people going It's not a YouTube video that you publish, and it has the potential to get thousands of views.
00:30:36.383 --> 00:30:37.685
No one cares about case studies.
00:30:37.685 --> 00:30:39.052
So the second is reach.
00:30:39.052 --> 00:30:42.974
Business potential, reach, and then E stands for effort.
00:30:42.974 --> 00:30:49.733
How hard it would be, because if something has like high business potential, it can sell your product well.
00:30:49.733 --> 00:30:58.236
And then like you can publish it on YouTube and create a great video, and that great video would go viral and get a lot of eyes on your product and people would want to buy it.
00:30:58.236 --> 00:31:00.181
What's the effort?
00:31:00.181 --> 00:31:12.183
If the effort is doing what MrBeast is doing, buying some Ferraris, crashing them to get millions of views, and at the same time show your product, for example, you have Your business is fixing Ferraris.
00:31:12.183 --> 00:31:19.383
So you want to smash a bunch of Ferraris and then restore them, and like promote your business, like how good you are at restoring cars.
00:31:19.383 --> 00:31:23.692
It's a lot of effort to do a video like that and to make it go viral.
00:31:23.692 --> 00:31:24.839
It's, it's crazy hard.
00:31:24.839 --> 00:31:28.478
And finally, the, the final letter in the Brew is W.
00:31:28.478 --> 00:31:30.182
Who is going to do it?
00:31:30.182 --> 00:31:33.432
Because you might have an idea that has high business potential.
00:31:33.432 --> 00:31:37.627
You might have a distribution plan of how to get reach.
00:31:37.627 --> 00:31:43.169
You might even be okay with the effort and resources required to pull this off.
00:31:43.169 --> 00:31:45.806
But then like who is going to do it?
00:31:45.806 --> 00:31:51.022
Do you have bandwidth on your, on your team, or do you have person on your team who is good with YouTube?
00:31:51.022 --> 00:31:52.062
As simple as that.
00:31:52.062 --> 00:31:54.923
Do, does someone on your team have, experience with YouTube?
00:31:54.923 --> 00:32:01.293
you can kind of partly solve it with hiring someone like an agency, and it would solve a whole problem for you.
00:32:01.293 --> 00:32:02.940
But basically, that's the framework.
00:32:02.940 --> 00:32:15.388
Whenever you're making a marketing decision, you need to understand how well it would sell your product, how many people you're going to reach, how much effort that would require, and who exactly is going to do this.
00:32:15.388 --> 00:32:18.830
So this is the framework I use, in my marketing to make decisions.
00:32:19.483 --> 00:32:19.815
Right.
00:32:19.815 --> 00:32:21.413
Sounds really interesting.
00:32:21.413 --> 00:32:27.365
and people listening to this episode, if they are triggered, by your framework Tim, where would you like me to send them?
00:32:27.365 --> 00:32:28.856
Where can they find the book?
00:32:28.856 --> 00:32:30.604
Where can they find Ahrefs
00:32:30.657 --> 00:32:45.700
the book should be already on Amazon, so people can pre-order it on Amazon or they can come to one of the conferences that we are sponsoring or to our conference, Ahrefs, because we'll be giving out a lot of these books for free.
00:32:45.700 --> 00:32:48.298
Because I didn't write this book to make money.
00:32:48.298 --> 00:32:54.822
I wrote this book to share my 15 years of marketing knowledge and the principles that I developed along the way.
00:32:54.822 --> 00:32:58.696
And yeah, basically the book is relatively short.
00:32:58.696 --> 00:33:02.643
It's just, like, 220 pages, so it's about three hour read.
00:33:02.643 --> 00:33:10.476
And if you want to spend three hours with my best thinking about marketing, with my best ideas, get the book.
00:33:10.476 --> 00:33:18.579
I honestly cannot see how you would read it and not get at least 10 cool ideas for your business
00:33:19.519 --> 00:33:24.307
That four years actually summarized in 10 good ideas for the business.
00:33:24.307 --> 00:33:25.826
Who wouldn't want that, Tim?
00:33:25.826 --> 00:33:29.820
So Tim, thanks, thanks so much for joining me today.
00:33:29.820 --> 00:33:31.706
It was a really insight conversation.
00:33:31.706 --> 00:33:32.327
Thank you
00:33:33.179 --> 00:33:34.274
Thanks for having me, Joeri
00:33:35.215 --> 00:33:37.255
Guys, what a brilliant episode.
00:33:37.255 --> 00:33:39.178
it sparked some ideas.
00:33:39.178 --> 00:33:41.663
obviously already had a look at Tim's book.
00:33:41.663 --> 00:33:42.745
It's really good.
00:33:42.745 --> 00:33:46.334
So go ahead and find it on Amazon.
00:33:46.334 --> 00:33:48.132
Let me know what you think about it.
00:33:48.132 --> 00:33:51.418
I will put all links in the show notes as always.
00:33:51.418 --> 00:33:59.448
if you think that it's also useful for people around you, this podcast episode, Tim's book, be sure to share it with them.
00:33:59.448 --> 00:34:06.484
If you are not yet following this show, hit the subscribe button because this is a really good moment to do this.
00:34:06.484 --> 00:34:14.730
If you haven't given me a review yet, if you give me these five stars, this really goes a long way, and of course, would love to see you back next time.
00:34:14.730 --> 00:34:26.637
Take care a special thanks as well to RYO, uh, the main sponsor of Web3 CMO Stories, supporting conversations at the intersection of Web3 AI and the future of digital leadership.