If you are still optimizing purely for Google rankings in 2026, you are playing the last game. Marius Meiners, co-founder and CEO of Peec AI, built a company to $8.6 million ARR in 14 months by building infrastructure for the next one. I caught his episode on The SaaS Podcast and it is one of the more practically useful founder stories I have heard this year.

What GEO and AEO Actually Mean

GEO — Generative Engine Optimization — is the practice of making your brand visible inside AI-generated answers from ChatGPT, Perplexity, Gemini, and their successors. AEO — Answer Engine Optimization — is a related term for the same underlying shift: search is increasingly happening inside conversational AI interfaces, not ten blue links.

Most founders treat this as a future problem. Marius treated it as a present opportunity the moment ChatGPT launched search capabilities in October 2024. He noticed that the sharpest SEO practitioners in the world were already obsessing over it. That signal was enough.

“The smartest SEO experts in the world were already obsessed with it. The signal was loud.”

The insight is simple: if buyers in your category are asking AI assistants for recommendations, and your brand does not surface, you are invisible at the moment of highest intent. Unlike traditional SEO, there is no public ranking algorithm to reverse-engineer. You need tooling to even know whether you exist in those answers.

That is the gap Peec fills — tracking brand mentions and sentiment across ChatGPT, Perplexity, and Gemini, then giving teams levers to improve visibility.

The V0 MVP and the LOI Filter

What makes this story useful for founders is not the outcome — it is the process.

Marius had not written production code in four years when he had the idea. He built a working prototype in V0 in 1.5 days. Not a mockup. A functional product he could put in front of potential customers.

Then he did something most founders skip: instead of asking “what do you think?”, he asked people to sign a letter of intent. Non-binding, no money involved — but requiring someone to actually read terms and digitally sign filters out a lot of polite noise.

Eight people signed. That was enough to start writing real code.

“Nothing beats someone paying. That’s the only real validation you can get.”

LOIs are not payments, but they are the closest low-cost proxy. They force the prospective customer to take a concrete action, which distinguishes genuine demand from encouraging conversation. This is especially critical in B2B, where people are professionally socialized to be nice about your idea.

Pricing as Market Capture, Not Revenue Optimization

Competitors in the AI search tracking space were pricing at €500 per month and above, targeting enterprise marketing teams with long sales cycles. Marius priced Peec at €85 per month.

The math sounds like it sacrifices revenue. The actual outcome: 2,000+ customers in 14 months, 300+ new customers per month, 30% of revenue from word-of-mouth, and $29M raised across a seed and Series A without slowing customer acquisition.

Aggressive mid-market pricing works when the category is new and trust is unproven. At €85, the question becomes “why not try it?” rather than “is this worth a committee decision?” You compress the sales cycle to near zero and generate the case studies you need to move upmarket later.

The Channel Validation Loop

Here is the part I find most compelling from a systems perspective: 20% of Peec’s own new customer conversions come through AI search. The product they sell — AI search optimization — is generating a fifth of their own pipeline through the channel it optimizes for.

That is not a coincidence. It is proof of thesis, and it is the kind of flywheel that is hard to argue with in a sales conversation.

“People really don’t want to buy stuff they don’t urgently need in B2B settings. It’s about finding demand much more than creating supply.”

If your product solves a problem, the best demo you can give is using your own product publicly and visibly. Peec showing up in AI search results for queries about AI search optimization is the case study and the distribution channel at the same time.

Scrappiness Has an Expiration Date

One thing Marius is direct about: the operational mode that gets you to product-market fit is not the mode that scales past it.

“Being scrappy at the start is super important…but you’ve got to shift into deploying capital very quickly when you find product-market fit.”

From zero to early traction: compress everything, move fast, spend nothing. From PMF to scale: hire deliberately, deploy the capital, build the systems. Founders who stay in scrappy mode after PMF often hit a ceiling not because the market dried up but because the organization cannot metabolize growth.

Peec went from two people to 55 in roughly 14 months. That kind of scaling requires letting go of the early operating style — deliberately and quickly.

What To Do With This

If you are a founder or marketer who has not audited your AI search presence, do it this week. Ask ChatGPT, Perplexity, and Gemini what tools they recommend in your category. See if you show up. See what they say about your competitors.

If you are not in those answers, that is your GEO gap. Tools like Peec will track it systematically; for now, manual spot-checks at least tell you where you stand.

The founders who build GEO into their content and product strategy now will have a compounding advantage over those who wait until AI search is as crowded and contested as Google.

The full episode is worth your time: The SaaS Podcast Ep. 481 on Overcast.