The hard part of AI search isn’t measuring it. It’s making the number mean something.
At a glance
- Company: Slopelift, a search agency running SEO and GEO programs for e-commerce and content-driven brands across Austria and Europe
- Interviewee: Marietta Robitza, Head of SEO & Chief Knowledge Officer
- Challenge: AI search introduced a new channel – and a whole new vocabulary of KPIs – that neither the agency’s clients nor the wider market knew how to interpret or act on
- Solution: OtterlyAI for AI-search visibility tracking, citation auditing, and content optimization
- Results: ChatGPT identified as the #2 search channel and the largest source of measured AI referral traffic; that traffic rising since March; AI-sourced visitors showing higher engagement than the site average
Ask most agencies about AI search and you’ll get a story about tools – which platform tracks visibility in ChatGPT, who scrapes citations, whose dashboard is prettiest.
Marietta, Head of SEO and Chief Knowledge Officer at Slopelift, learned that the tooling was never the real problem.
“Where it gets difficult,” she says, “is how do we interpret the data.”
A brand-visibility score is easy to produce and easy to display. Getting a client – and their management – to understand what it means and what to do about it is the actual job. That gap, between a metric existing and a metric mattering, is the story of how Slopelift built its GEO practice.
Slopelift saw it coming before the tools existed
Slopelift’s interest in AI search predates the software built to measure it. The team had been dissecting Google’s featured snippets in fine detail — how they were assembled, why Google surfaced particular pages — long before generative results went mainstream.
“We already started actually before AI Overviews were launched,” Marietta says. So when AI Overviews rolled out — “a big step in Austria, and in Europe in general” — Slopelift didn’t scramble. It asked a sharper question than most: for each individual client, what does AI search actually mean for their audience? Does it even affect them? And can it be turned into a source of traffic rather than a threat to it?
What they found didn’t match the headlines
The prevailing narrative was collapse — zero-click searches everywhere, organic traffic in freefall. Slopelift’s data told a more precise, more useful story.
For e-commerce clients, the disruption was minimal. “Very little,” Marietta says — no meaningful spike in zero-click searches. For sites leaning heavily on informational and blog content, clicks did soften, “but definitely not as much as for publishers.”
“There was no spike in zero-click searches for e-commerce. For content-heavy sites clicks went down a little — but definitely not as much as for publishers.”
That nuance mattered commercially. Instead of alarming every client with the same doom slide, Slopelift could tell each one what was actually happening to them — and where the opportunity was.
The real bottleneck: a channel with no shared language
The hard part wasn’t loss. It was literacy.
AI search arrived with an unfamiliar vocabulary — brand coverage, share of voice, citations — and no established framework for reading it. “It’s new for everyone,” Marietta says. Slopelift’s team had to learn how each KPI was calculated, then face the harder question every client eventually asked: okay, what does this tell us now?
For an agency whose remit explicitly includes training and education — Marietta owns the learning programs for every discipline at Slopelift, not just SEO — solving that interpretation problem wasn’t optional. It was the product.
Why Slopelift chose OtterlyAI
Last year the requirements came into focus. Slopelift needed to measure its own and its clients’ visibility in ChatGPT and other AI platforms — and, critically, it needed something clients could hand upward to their own leadership.
OtterlyAI met both. “You get the nice dashboard views, you get easy-to-understand diagrams,” Marietta says. In feedback from the teams using it, the same points came up repeatedly: competitor logos lined up side by side, performance visible at a glance, “you can really easily grasp everything.”
For a practice built on making complex things legible, that clarity was the whole point — not a feature.
Where it clicked: connecting visibility to revenue signals
The breakthrough came when Slopelift stopped presenting visibility scores in isolation and started pairing them with what the traffic actually did.
A brand-coverage figure on its own — “say you have a brand coverage of, I don’t know, 89%” — left clients unsure what to do with it. But shown alongside measured AI referral traffic and engagement data, the same number suddenly had weight.
“It really starts to click with them. ‘Okay, that channel is actually interesting for us.'”
That reframing — from abstract score to business channel — is what moved GEO from a curiosity to a line item clients wanted to invest in.
(Note: the 89% above is Marietta’s illustrative example, not a specific client result.)
How OtterlyAI shows up in the day-to-day
Three uses, all concrete:
Citation and reference cleanup. On several projects, the team reviews the external sites feeding AI answers to find where a brand is missing — or where it’s cited with an outdated URL or old logo. “That was quite easy to find out and also easy to use,” Marietta says: low effort, obvious payoff.
Content optimization. On-page content work is one of Slopelift’s largest services, and the agency now runs pages and copy through OtterlyAI’s content GEO checker as a standard step, applying its recommendations before anything ships.
From “what is this” to “what do we do.” OtterlyAI’s recommendations feature closes the loop on Slopelift’s original bottleneck — interpretation. “That’s the next step, right? How do we use the data? That’s something that’s really helpful.”
The results, stated honestly
The signals are early but consistent.
ChatGPT has become the clear number-two search channel behind Google — the larger share, and the source of the majority of AI referral traffic Slopelift measures, which is why the agency focuses its AI-search effort there. That traffic was flat through last year and has been climbing since around March.
The more interesting finding is quality, not quantity. Compared to the site average, AI-sourced visitors “tend to spend more time on the site and tend to engage more” — early evidence, Marietta notes, that “AI traffic users are highly qualified.” She’s careful to caveat it: so far that’s one client, with more to evaluate. But it points the right way.
“The engagement rates are higher. Users coming from AI traffic tend to spend more time and engage more — which confirms the theory that AI traffic is highly qualified.”
What Marietta expects next
Her outlook is level-headed. She thinks the whole industry will grow fluent in GEO KPIs over the next year as the vocabulary becomes routine. She’s not writing off the corporate website either — “in 2027 you can still safely use your website” — and expects it to stick around longer than the hype suggests.
Her real hope is about quality. She wants content teams to resist churning out AI slop and “not ruin the internet again, like some SEOs did 20 years back” — to return to unique content that’s genuinely useful. “If that’s not the goal, what is the goal?”
For Slopelift, that principle and the tooling now reinforce each other. With OtterlyAI, the agency doesn’t just measure AI search — it can explain it, act on it, and show clients why it’s worth the investment.
Want to make AI search measurable – and meaningful – for your brand or agency? Learn more at OtterlyAI.
