This week I sat down with Tim Soulo, CMO at Ahrefs, for a conversation that bounced between product strategy, marketing craft, data moats, and the new chaos of AI-driven discovery. If you have felt that the ground is moving under search, you are not imagining it. We talked about what is changing, what is hype, and what still seems stubbornly true.
Tim has a front-row seat to all of it. Ahrefs is one of the rare companies that crawls the web at meaningful scale, ships tools used across the industry, and runs its own marketing like a living lab. That combination makes Tim interesting to talk to because he is watching the changes coming through data, customer behavior, and his team's experiments in real time.
Key takeaways
- Ahrefs' durability comes from a simple loop: high-quality data plus a marketing team that uses the product daily, so features are validated internally before they are marketed externally.
- Humor is not a personality quirk, it is a distribution strategy. The Semrush/Adobe campaign worked because Ahrefs understood audience sentiment first, then shipped something shareable.
- In the shift from SEO to GEO, Tim believes the fundamentals still win: publish exhaustive, clear information on your own site and earn credible third-party coverage. Today's listicle hacks will get patched.
- Tim believes that many AI monitoring tools are easy to launch and hard to defend. Data moats and infrastructure still matter, especially if you are competing with companies that crawl the web.
- Vibe coding is lowering the bar for 'a tool exists', which makes differentiation move toward unique data, trust, integration, and the ability to keep shipping.
Ahrefs 'still feels young' and why that matters
Tim started with a point I have heard from a few founders and long-tenured leaders: time compresses when you are shipping. He has been at Ahrefs for a decade, and the company is well past the 'new kid' stage, yet he still feels that sense of momentum.
What has kept Ahrefs relevant, in his view, is not a single feature or a single marketing channel. It is the combination of high-quality data and a team that uses the product to do its own work. That 'we need it ourselves' feedback loop is a shortcut to product direction. You do not have to guess what might be useful if your own marketers are pushing on the same problems customers have.
That internal validation also creates a certain confidence. You are not relying on theoretical personas. You are building tools your own team would complain about if they were missing.
Marketing to SEOs without sanding off the geekiness
One of the things I always notice about Ahrefs is that it wears its 'SEO tool company' identity without apology. Tim and I joked about the brand name itself. When your company name is literally rooted in a technical HTML attribute, you are making a choice. You are leaning into the fact that your audience includes people who enjoy the craft.
Tim's take was refreshingly simple: he does not get much pushback internally. Their founder does not micromanage marketing, which gives the team freedom to find a voice that resonates. That is not a small point. Many marketing teams end up defaulting to the safest, blandest version of themselves because someone senior is nervous about standing out.
There is still a practical consideration, though. If your marketing is too inside-baseball, you risk only speaking to the top slice of experts. Tim described Ahrefs' content approach as intentionally broad: publish for beginners, juniors, and mid-level marketers, because that is where scale is. The most advanced practitioners can piece things together themselves. They might use the API, build their own workflows, and go deeper in their own way.
I liked his framing here because it is a reminder that 'speaking to experts' and 'being useful at scale' are different jobs. Great marketing teams learn to do both without confusing them.
'We solve problems' beats 'ideal customer profiles'
Another moment that stuck with me: Tim said they do not try to define strict ICPs. Instead, they think in terms of problems. If someone has the problem Ahrefs solves, they should be a customer, whether they are a solo affiliate marketer or a giant enterprise team.
That sounds obvious, but it runs against a lot of modern SaaS dogma (and, to be transparent, how we are operating at SearchPilot!). The tension is real: ICP thinking can keep you focused, but it can also make you blind to demand that does not fit the template. Ahrefs' pricing tiers let them serve a broad market without pretending every customer is the same.
The interesting implication is that their marketing can stay curious. It can follow the problem wherever it shows up, rather than forcing every story into one narrow buyer persona.
Humor as a distribution strategy
We spent a chunk of time on the Semrush/Adobe moment, because it is a great example of how brand voice and growth tactics collide. Tim described their reactive campaign as something that did not happen on impulse. Before they went hard, he tested sentiment publicly. He essentially asked the market, 'Do you want us to play this game?' and watched the response.
That move matters because reactive marketing is easy to get wrong. The internet is quick to punish tone-deaf opportunism. Tim's approach was closer to reading the room, then writing the joke.
Once they saw demand, they leaned into what Ahrefs does best: be clever, be self-aware, and be willing to poke fun at themselves too. The 'hard to pronounce' checkbox on the comparison page is a tiny detail, but it signals something important: 'We are not pretending we are above this.'
Under the humor, there is a serious point Tim made: a boring landing page does not travel. Humor is what compels people to share. If you can create something that makes someone smile, you can earn distribution you would otherwise have to buy.
From SEO to GEO: the fundamentals Tim thinks still hold
When we moved into AI, I asked Tim about the extreme long tail of prompts. Traditional search has head terms and enormous query volumes. LLM conversations can feel more like one-off dialogues, where every prompt is unique. How do you track that?
Tim's answer cut through the tooling noise. He sees the solution to AI visibility as twofold:
First, have exhaustive information on your own site. If your product, positioning, and features are not clearly explained where a model can find them, you are already behind.
Second, pay attention to what other websites say about you. Models learn from the wider web, and right now that creates obvious opportunities for manipulation. Tim called out the listicle wave: agencies pumping out 'best X tools' posts to manufacture brand mentions and rankings in AI answers.
His bigger point was that these hacks will not last. The loopholes will be patched, and they will be patched faster than Google patched past loopholes. The end state, in Tim's view, is the same cat-and-mouse dynamic SEOs already know. Different surface, similar game.
Why there are so many AI monitoring tools right now
I wanted to hear Tim’s view on why there are so many AI monitoring tools that have launched so quickly, with so much funding and noise?
Tim did not sugarcoat it. He called it a gold rush, and in many cases, gambling. The upside is big enough that people will try, even if most attempts fail.
He also pointed out the uncomfortable truth: many of these tools are easy to launch. If your product is essentially 'run a set of prompts, collect outputs, display dashboards', you can ship a version quickly. Add modern AI-assisted coding on top and the barrier drops even more.
This is where Tim drew a sharp line between 'easy to ship' and 'hard to defend'. Ahrefs has a moat because it crawls the web at scale and has accumulated years of data and quality signals that people rely on in real workflows. You cannot vibe-code a full web crawler that stores the internet, maintains freshness, and builds reliable link signals on your laptop.
If the only thing you have is a UI on top of commodity data, you should be nervous.
Vibe coding and why moats are getting tested
Tim shared a story that their CEO gave everyone in the company access to a vibe-coding tool, and Tim used it to build mini apps he had wanted for years. He started as a skeptic, got frustrated when it threw errors, then watched it fix those errors and eventually deliver working software.
From there he went a step further: he does not see how a lot of 'mini tools' survive when a motivated user can build something tailored in an afternoon. That does not kill software, but it changes where value lives. The value shifts away from 'a simple tool exists' and toward things like unique data, distribution, trust, integration, and continuous maintenance.
He connected this back to Ahrefs directly: their defensibility is not a dashboard. It is the data and the infrastructure.
Yep as R&D, and the pivot toward search APIs
We also talked about Yep, Ahrefs' search engine. Tim described it as R&D. Building a search engine forced improvements in crawling, storage, and understanding of pages, and those improvements fed back into the core product.
One concrete example: page history. Ahrefs can show previous snapshots of a page, closer to an archive view, even if it is currently more text/HTML than full visual snapshots. Those capabilities come from doing real search-engine work.
The customer-facing search engine front end is paused. Tim's argument was simple: conventional search as a consumer product is less compelling now, because AI changed the attention landscape.
But he hinted at what might be more important: search APIs. If AI systems need grounding in the live web, they need a way to fetch and rank information. The big incumbents have reasons to limit access. That creates an opening for independent search infrastructure.
Tim was clear he is not running that project, but it is hard not to see why Ahrefs would be well-positioned for it.
JavaScript rendering and the odd return of '2006 problems'
One of my favorite parts of this whole AI era is how it drags old problems back into view. We joked about the weird oscillation between modern web complexity and a return to simple HTML constraints, because many bots still struggle with JavaScript rendering.
Tim did not claim deep expertise here, but he noted that Ahrefs executes JavaScript on a meaningful portion of pages, and that the product even exposes reporting around links found in JavaScript contexts.
It is another reminder of how much infrastructure work sits underneath the tools we casually use. Rendering, crawling, interpreting, storing, deduplicating, and refreshing. It is not glamorous, but it is the foundation.
How Tim filters hype: blockchain, AI, and 'wait until it works'
Tim's mindset section was one of the most human parts of the conversation. He described his default stance as underpromise and overdeliver. He hates hype that does not land, because it wastes time and creates distrust.
He used blockchain as the cautionary tale. Big promises, few real outcomes. That experience shaped his skepticism toward AI too. He waited until the tools worked well enough to justify the attention.
I came at it from a different angle. I assumed some of the promises were impossible, so I was surprised when they worked better than my baseline expectation. Different psychology, same destination: both of us now see genuine utility in the tools, especially for building small apps, doing research, and generating variations during ideation.
Tim's three AI use cases in practice
Tim boiled his own AI usage down to three buckets:
- Vibe coding: building mini tools, automations, and simple apps that used to require engineering help
- Research: faster synthesis, fact-finding, and broad scanning, with the reminder to double-check anything sensitive or publishable
- Writing and ideation: generating variants for titles, outlines, and messages, then choosing what is better than the first draft
I added one more that I have felt personally: it is a safe place to ask the 'stupid questions' you would rather not ask in public. Tim agreed and shared a nice tactic: screenshot a phrase you do not understand from a conversation, drop it into an AI tool, and get the explanation without breaking the flow.
That is the throughline of the whole session for me. New surfaces, new tools, new jargon, but the same job: make sense of uncertainty, then do the work that holds up when the platform shifts. In search, that is where experimentation stops being a nice-to-have and becomes how you stay sane.
Put Search in Control Mode with SearchPilot
If you want to take that mindset from 'interesting ideas' to 'we know what worked', this is exactly what SearchPilot is built for. Search is often the biggest channel and the least understood, and SearchPilot makes SEO (and now GEO!) testable so leaders can move from guessing to knowing.
We run controlled experiments across category pages, product detail pages, navigation, and content, then deliver clear uplift with timelines and confidence. Teams move from quick validation to a steady test cadence to full control, turning search into a performance channel you can plan and fund.
For ecommerce teams focused on product grids, Merchant Center feeds, and variant handling, the first step is a focused test plan. Measurement tracks impressions, clicks, and revenue so leaders can see the real impact.
Stop trying to predict the future. Experiment to discover it. If you want tailored test ideas for your top PLPs and PDPs, schedule a demo and we will share a starter list and a clear path from validation to velocity to control.
%20(9).png)
%20(7).png)