In brief: A new Polish expert site in one of the most saturated B2B categories — sales consulting — began being cited in Google AI answers within 60 days of launch, with no podcasts, no industry media, and no large budgets. Proof that a small company can realistically compete for a place in AI answers today — if it knows how to claim it. That site is karoltabis.com — the founder's own proof for Discovery Stack, before work begins with your company.
This is not a case about us. This is a case about what just happened to your company.
It's 8:47 AM. Your ideal customer — a CMO at a B2B company, exactly the one you've been trying to reach for months without success — opens ChatGPT and asks about the problem you solve better than anyone else in your category.
AI answers. It lists three brands. Yours isn't one of them.
It's not there because your competitor figured out something you haven't yet: the era of "being visible in Google" as the primary B2B marketing currency has ended. Your customers click through search results less and less. They ask artificial intelligence and make decisions based on what they hear.
The common assumption is that in this new era, only established market leaders with massive SEO budgets stand a chance.
This case study completely disproves that assumption.
What happened
In March 2026, a new expert site launched in one of the most saturated B2B sales categories — sales consulting in Poland. No large budget. No paid campaigns. No podcasts. No industry media mentions.
After 60 days, Google AI began systematically citing its content — alongside global B2B software brands and leading Polish industry publications.
Next-generation algorithms don't search for the richest domain. They search for the best, most understandable answer. This changes the rules: a well-designed small-company site can realistically compete for a spot in AI's answer alongside a global leader — provided someone knows how to design it.
What we observed (precisely)
As of May 26, 2026 — verified live across leading generative engines.
Three representative situations:
-
For the query
buying committee, the site is one of the primary sources Google AI builds its answer from. Even more striking — within the generated text, the model explicitly recommends the reader deepen their knowledge there. This is a rare signal, visible only for sources AI treats as authoritative. -
For the query
how to increase B2B sales, content lands directly in the AI Overview summary, and the page ranks #2 in organic results — just below a global B2B software leader. -
For the query
B2B prospecting system, Google AI Mode pulls several independent fragments from a single article, placing the author alongside the most established CRM brands in the market.
These are not one-off appearances. The site ranks for over 70 categorical keywords, and across leading language models, nearly 200 independent references and citations to the author have been registered across contextual queries.
This site did NOT appear in industry media. It had NO podcasts. It sent NO newsletter blasts. None of the classical "presence" playbook. And yet — Google AI selects it as the answer to questions your customers ask every day.
What this means for your company
This case isn't a curiosity. It's a signal you need to read carefully — because it concerns every marketing decision you'll make in the next twelve months.
Your customers no longer search for your brand — they search for a solution to their problem. The question isn't whether your company is "visible in Google." The question is: when AI builds an answer to a question your customer asks every day — does your brand appear as one of the names mentioned, or not at all?
Every month of delay has a cost. The first companies in each B2B category are taking their position in AI answers right now. That position, once claimed, is difficult to dislodge. Language models learn slowly and are reluctant to change their minds. Whoever becomes the default answer now will hold that position for the next 5–10 years.
Traditional metrics are decoupling from reality. Site traffic, click-through rates, and search positions — these are increasingly weak indicators of your real influence on buying decisions. The new currency is share of AI answers — and if you're not measuring it, you don't know where your company actually stands.
How we know this works for you
The site in this case study is karoltabis.com — the personal site of Karol Tabiś, the creator of Discovery Stack. We publish this case openly for one reason: it's our own proof that this works, before we start working with your company.
What you saw above is a fragment of what's possible. A full Discovery Stack implementation — which we run for a select group of companies each year — scales these results in ways no other B2B marketing approach available today can match.
We don't promise percentages. We show what's possible — and help you check where your company stands today.
What the first step looks like
You fill out a short Discovery Stack Diagnosis form. Five fields, five minutes of your time.
Within five business days, you receive a two-page analysis:
- Where your company appears today in AI answers for queries your customers actually ask
- Who AI mentions instead of you
- What this state of affairs is costing you over the next 12 months
The decision on next steps belongs to you. The Diagnosis is a starting point, not a commitment.
Key takeaways
- A new site can enter Google AI answers in 60 days — without podcasts, industry media, or large budgets.
- AI picks the best, most understandable answer — not the richest domain — a small company can realistically compete with a global leader.
- A place in AI answers, once claimed, is hard to take back — models change their minds slowly (a 5–10 year window).
- Traditional metrics poorly reflect real influence — the new currency is share of AI answers.
- The first step is measurement, not optimization — where your company stands in AI answers today.
Evidence based on at least 96 measurement points across leading generative engines (as of May 26, 2026). Case open — updates in progress.