about the ac group

We have been earning attention since 1999. Only the engine changed.

Most firms in generative engine optimization showed up in the last two years. We have been competing for human attention online for more than two decades — through three distinct eras of how that attention is won. This is the story of a company that kept one goal and changed its method three times, and exactly why that long arc makes us better at getting you cited by AI today.

the through-line

One goal, since the start: be the answer worth choosing.

In 1999, "the answer" lived on a web page a person found through a directory or a link. In the Google era, it lived in a ranking. Today it lives inside an AI-generated response, cited from a handful of sources. The surface kept moving. The job did not: understand the question better than anyone else, and answer it so well that the system, human or machine, has to point to you.

That continuity is the whole point. We did not pivot into GEO because it was the trend of the year. We arrived at it the same way we arrived at SEO two decades ago — by following where attention went and learning, in depth, how to earn it there.

That history shapes how we work, beyond what we claim. We start every engagement with evidence — a baseline of how the AI engines actually describe a client today — because two decades of search work taught us that opinions are cheap and measurement is not. We write with data and verifiable claims rather than adjectives, partly because it is honest and partly because it is what gets cited. And we hold our own site to the standard we sell, so the method is visible rather than asserted. None of that is a recent pose; it is how a company behaves when it has had to earn attention through three different machines.

the three eras

How the company evolved without losing itself.

Each era added a craft. None replaced the one before. That accumulation is what a client hires.

1999 useful content

Earn attention

We started before Google ran the web, when discovery meant hand-curated directories like Yahoo and DMOZ, a handful of early crawlers, and meta keywords you typed in by hand. There was barely an algorithm to game, so the discipline was pure: publish something genuinely worth reading, for whatever market needed it, or be ignored. We wrote for readers first because readers were the only audience that existed — no ranking system was going to rescue thin work. That instinct, useful or invisible, never left the company, and it turned out to be the one rule every later era kept rewarding.

2004+ the search era

Rank on Google

As Google became the front door to the internet, the craft changed and we professionalized with it: technical SEO, information architecture, link authority, and search intent. We lived through the updates that separated operators from manipulators — Florida, then Panda on thin content in 2011, Penguin on link spam in 2012, Hummingbird on meaning in 2013 — and each one quietly rewarded the same thing we had done since 1999. We learned to make content rank without gaming it, and built the deep search craft that GEO now stands on. The teams that only knew tricks did not survive those years; the ones who understood why Google kept moving did.

2022 the inflection

ChatGPT arrives

When the first mass-market LLM shipped in late 2022, we did not treat it as a novelty or a threat to wave away. We treated it as the signal we had been watching for: search volume would soften, and the highest-intent questions — the ones that used to convert — would move into AI chat. While the industry argued about whether any of it was real, we started shifting toward GEO and AEO, mapping how engines choose what to cite. By the time Google announced SGE in 2023 and shipped AI Overviews in 2024, we were already doing the work, not scrambling to name it.

2026 today

Specialized in GEO/AEO

Today we engineer brands to be cited and recommended by AI, on a foundation of content and search craft 27 years deep. The mechanics are new — entity grounding, structured answers, earned corroboration, share-of-voice across five engines — but the goal is exactly what it was in 1999: be the most useful, most verifiable answer to the question a person is actually asking. The engines kept changing the rules; the thing they reward underneath never did. That is the whole reason a long arc beats a fresh logo here.

why we moved early

We read the ChatGPT moment for what it was.

When ChatGPT reached the public, most marketers filed it under "interesting toy." We saw a change in where questions get asked. If people could get a direct, synthesized answer without ten blue links, then the click economy that funded the web for twenty years would start to erode, and the brands cited inside those answers would inherit the attention that used to flow to rankings.

So we did what we had done once before, when Google changed the game: we went deep on the new mechanics early. How models retrieve. How they decompose a question into sub-questions. Why they favor definitive, well-structured, entity-rich sources. By the time "GEO" had a name and a conference circuit, we had already been practicing it.

Conviction is easy to claim after the fact. Ours is visible in the work: a site engineered to be cited, pricing published in the open, and a standing challenge to test us in the engines before you spend a euro.

There is a quieter reason the early move mattered. Citation share compounds. The first brand to be consistently named in a category becomes the reference others get compared against, and models reinforce that lead every time they answer. Moving early was not about bragging rights; it was about claiming ground that gets harder to take later. We tell clients the same thing now that we acted on then, the window is open, and it does not stay open forever.

what we believe

Four principles that survived every era.

Useful beats clever

Every era rewarded the same thing underneath the tactics: content that actually helps. Tricks expire; usefulness compounds. AI citation is just the newest proof of an old rule.

Own your roots

GEO is not a replacement for SEO; it is built on it. We do not hide our search heritage to look new — we lean on it, because the teams that understand the whole chain make better calls.

Evidence over adjectives

We write with data and verifiable claims, not grandiloquence. It is more honest, and it happens to be what AI engines cite. We hold our own pages to that bar.

Measure or it did not happen

Visibility work earned a bad name by being unmeasurable. We instrument citation from day one, set targets, and report against them every week.

who we work with

Built for B2B SaaS that sells in two languages.

We focus on B2B SaaS companies between roughly €1M and €20M in ARR — past product-market fit, serious about growth, and underserved by the enterprise agencies that price their programs from fifteen thousand dollars a month. These are teams that know AI search is changing how buyers find them and want a partner who can act now, not a year into a learning curve.

We also do something most of our field does not: we work in English and Spanish with equal rigor. The Spanish-language AI search market is large, growing fast, and technically underserved. A SaaS brand selling into both North America and Spanish-speaking markets usually has to stitch together two vendors. We run both from one team, with research done natively in each market rather than translated after the fact.

What we are not is a fit for everyone. If a team wants the cheapest possible retainer, or a vendor who will promise a number without measuring a baseline first, we are the wrong call, and we will say so early. We work best with people who want the work done properly and want to see the evidence, the same standard we have held since there was barely an algorithm to answer to.

why the heritage matters to you

A 27-year head start is not nostalgia. It is a moat.

There is a real, current debate in our field: critics argue that GEO is just SEO with a new name. They have a point worth taking seriously, and our history is the cleanest answer to it. We are not threatened by the overlap, because we lived it. We know exactly where SEO ends and where the new work begins, because we built the SEO foundation ourselves and then extended past it.

A GEO-only agency founded last year has to learn the search fundamentals that everything still rests on. An old SEO shop that bolted "GEO" onto its deck has to learn the new citation mechanics. We are the rare case that did both, in sequence, in real conditions. That is the difference between a vendor reacting to a trend and a partner who saw it coming.

There is a practical payoff to that for clients, beyond the story. When the tactics of the month change, and in a field this young they change constantly, a team grounded in fundamentals adjusts calmly instead of chasing every new claim. We have watched enough cycles to tell a durable shift from a passing one, and to keep a program steady through both. Twenty-five years does not make us old; it makes us hard to surprise.

There is a practical reason that matters to you, not just to us. A team that has only ever known AI search treats every shift as a crisis, because it has no prior pattern to compare against. We have watched discovery move three times, so we can tell the difference between a genuine structural change and the noise of a busy news cycle. That judgment is hard to buy and harder to fake, and it is the quiet thing you are actually hiring when you hire for experience.

See how we run a GEO program →

where this is going

The next era is already forming. We are watching it the same way.

Search did not end; it changed shape. The same is happening again as AI assistants gain memory, take actions, and shop on a buyer's behalf. The brands that machines trust today will be the defaults those agents reach for tomorrow, and the gap between cited and uncited will compound. Being in the answer is becoming the whole game.

Our job is to keep reading the signal early — as we did in 1999, in the Google era, and again in 2022, and to translate it into work that earns our clients attention wherever buyers go looking. The tactics will keep changing. The discipline underneath them, the one we have practiced for more than two decades, will not.

If there is one thing we would want a prospective client to take from this page, it is that longevity here is not a vanity metric. In a category where most competitors are younger than the technology they sell, having watched attention move three times is a practical, daily advantage — it is the difference between guessing where buyers go next and having a tested method for finding out. That is what we bring, and it is precisely what we have always brought.

put us to the test

Don't take the history on faith. Ask an AI about us.

Open ChatGPT, Claude, Gemini or Perplexity and ask which GEO agencies to consider for B2B SaaS. If we have done our job, we are in the answer. Then let us run the same test on your brand — free, in 48 hours, no sales call.

Get your free AI visibility audit