€2M ARR on €1.2M: How Promptwatch Sold Its €6M AI Search Optimization Round

Twelve months after launch, Promptwatch was booking €2 million a year in recurring revenue. It had raised just €1.2 million to get there. Last week the Amsterdam startup added a €6 million seed round, and the most revealing part of the deal is not the amount. It is what its two lead investors chose to talk about publicly: not the technology, not the size of the market they are chasing, but a single ratio that told them everything they needed to know. Promptwatch got funded for how little it had spent.

Promptwatch is a bet on a shift you have probably already felt. For twenty years, brands knew exactly how customers found them: keywords, rankings, click-through rates. Now people ask ChatGPT, Claude and Perplexity instead, and those models recommend brands without ever explaining why. Gartner expects organic search traffic to fall by 50% or more by 2028. The entire discipline of SEO, a multi-billion-dollar industry, is being quietly rewired into something with no dashboard.

Gijs de Groot and Klaas Foppen started Promptwatch in 2025 because that dashboard did not exist. De Groot, 26, is the commercial half: a business-economics degree from Tilburg, a master’s in financial economics from Erasmus, a stint in sales at Sparkbase, and enough of a public profile to speak at BrightonSEO, the SEO industry’s main stage. Foppen is the technical half, with a build history across Homestra, Sparkbase, Podvine and Foodsy. The pair are not first-timers. Before Promptwatch they had already built and, reportedly, sold Sparkbase together, only a few months after launching it, then turned straight around and started again.

That history matters for how you read the raise. A pair who has already taken a company from zero to an exit knows what a real business looks like, and knows the difference between vanity growth and revenue. It shows in the choices: they did not raise a large seed to buy growth, they built the growth first and raised against it. Investors pattern-match hard on this, and Promptwatch gave them the pattern they wanted.

What they built is deliberately not another monitoring tool. Promptwatch calls itself an end-to-end “AI Search Optimization” platform, or GEO, for Generative Engine Optimization. It ingests more than 100 million real prompt-and-response data points every day to show a brand how the models actually describe it, finds the gaps, generates AI-optimized content to fill them, and then publishes that content straight into WordPress, Webflow and Framer. It plugs into Claude and ChatGPT through the Model Context Protocol. The pitch is not “here is your visibility score.” It is “here is your visibility problem, and here is the fix, already shipped.”

When the round was announced, the founders said the expected founder things. The interesting quotes came from the money.

Alexander Kölpin, managing partner at the lead fund, seed + speed Ventures, explained the decision in one line: “Promptwatch convinced us because its experienced founding team has already achieved more than 2 million euros in ARR with limited capital and personnel resources.”

Then André Hammerer, managing director at co-investor Blum Ventures, published his own note. His version: “What convinced me most is the execution: €2M ARR within 12 months of launch, achieved with €1.2M in funding and a small team in Amsterdam.”

Read those twice. Two different funds, deciding independently, each reached for the same fact and framed it the same way. Not the market size. Not the technology. The ratio — revenue produced per euro consumed. When two investors who did not coordinate their diligence both write down the same number as the reason they wired money, that number is the round. Everything else in the data room was confirmation.

This is the quiet lesson under the whole story. Promptwatch did not win a €6 million round by out-narrating a $17 billion market. It won by handing investors a metric so clean that the diligence wrote itself. The founders did the expensive, unglamorous work of getting to €2 million ARR on a shoestring first, and let that single ratio do the persuading. The deck was a formality.

Most founders pitch investors who stopped writing checks six months ago. Fund Momentum tracks 970+ actively deploying VC funds with real GP signals — who's bullish, what they're writing, and how to pitch them. Now also available as an MCP API so your AI tools can pull investor intelligence directly. Stop guessing. Start closing.

The €6 million seed was led by seed + speed Ventures, a Berlin-based fund that backs enterprise and B2B software across the DACH region. Blum Ventures joined, with its deal driven by Hammerer alongside Philipp M. Roth and Wen-Long Cheok. And Arches Capital, which had led Promptwatch’s €1.2 million pre-seed back in September 2025, doubled down. The advisory firm torq.partners supported the process. Total disclosed capital now sits at roughly €7.2 million.

Look at the shape of that syndicate, because the sequencing matters more than the sum. The most powerful name in it is not the new lead. It is Arches Capital, the pre-seed investor coming back for more. When the fund that watched you operate for a year decides to write a second, larger cheque, it hands the incoming lead the single most credible piece of diligence available: the insider is not cashing out or sitting still, the insider is buying more. Arches doubling down did more to de-risk seed + speed’s decision than any customer reference could.

The new leads brought the thing Promptwatch specifically needs next: seed + speed’s enterprise-software DNA in DACH, and a European syndicate positioned to help push into the United States. The use of funds is unambiguous and milestone-shaped rather than vague: open a New York office, deepen the relationship with a US customer base that already includes marquee names, build out the agentic AI features, and hire into engineering and go-to-market. This is not a “we need runway” raise. It is a “we have a working machine and we want to point it at America” raise, and the money is scoped to that.

The founders’ framing stayed disciplined. As de Groot put it: “This funding validates our belief when we started Promptwatch that businesses care about how AI describes their brand and that AI Search will become a meaningful revenue channel.” Note what he did not claim: no talk of dominating the category, no hand-waving about being first. The claim was narrow and already proven by the ARR line. In a market this hyped, the restraint is itself a signal.

The number that made the round believable came from a customer, not a founder.

Monks, the global marketing agency whose client roster includes Netflix, Asana and BMW, started using Promptwatch after noticing something it could not explain: clients were losing market share for reasons none of its existing tools could surface. The old SEO dashboards showed nothing wrong. The brands were simply being cited less often by AI models, invisibly, with no keyword or ranking to point at. After adopting Promptwatch, Monks reported a 72% increase in AI Search citations across its clients, and now runs centralized client reporting through the platform.

That is the whole thesis compressed into one deployment. A sophisticated agency, serving some of the largest brands on earth, could not see the problem until Promptwatch showed it to them, and then could not fix it as fast without the automated publishing loop. Duolingo, Fireflies, ABN AMRO, WPP and iO Digital are on the customer list too, but Monks is the one that matters, because it comes with an attributable number. “1,840 organizations use us” is a logo wall. “Monks saw a 72% lift in citations” is proof. The founders had both, and they led with the one that could be measured.

There is a second thing the ARR line is quietly buying, and investors clearly saw it: data. Every one of those 1,840 customers feeds the same engine, and the engine now processes more than 100 million prompt-and-response pairs a day. That volume is not a feature you can clone by copying a screenshot; it compounds. The more brands Promptwatch optimizes, the sharper its picture of how the models actually behave, and the better its generated content performs, which wins more customers. A monitoring tool sells a dashboard. Promptwatch is accumulating a proprietary map of how AI systems form brand opinions, and that map gets more valuable with every day it stays ahead. The €2 million of ARR is the visible layer. The data flywheel underneath it is what a seed investor is really pricing.

Raising (soon)? I’ve helped 50+ startups to accelerate their fundraising campaigns and match them with the right investors. If you want hands-on support - from pitch refinement to warm intros to term sheet negotiation - let’s talk.

1. Raise on a ratio, not a narrative. Promptwatch’s entire round rested on one legible number: €2M ARR built on €1.2M of capital. Both leads named it independently. When you can compress your traction into a single ratio an investor can repeat to their partnership without a slide, you have pre-written their diligence. Do the unglamorous work of reaching a clean metric before you open the round, then time the raise to land on it. A vision deck persuades one person in the room. A ratio persuades the ones who are not in the room.

2. Turn your existing investor’s follow-on into your lead’s diligence. The strongest signal in Promptwatch’s syndicate was Arches Capital coming back from the pre-seed with a bigger cheque. The investor who has watched you operate for twelve months re-upping is more convincing to a new lead than any reference call. Engineer for it: keep your earliest backer close, updated, and impressed, so that when you raise the next round their continued conviction is on the table as evidence. An inside follow-on is not a nice-to-have. It is your cheapest, most credible due diligence.

3. One measured customer beats a thousand logos. Promptwatch works with 1,840 organizations, but the round turned on Monks and a specific figure: a 72% lift in AI citations. Logo walls signal adoption; a single attributable customer result signals value. Before you raise, get one marquee customer to let you quote a hard number tied to an outcome. Investors discount aggregate usage stats because founders inflate them. They cannot discount a named agency reporting a specific percentage. Find that one number and put it at the center of the story.

4. Sell the loop, not the feature. In a category where Peec AI and Profound have raised far more, Promptwatch got underwritten on being end-to-end: it measures, generates and publishes, closing the whole loop instead of owning one step. That is harder to copy than any single feature and easier for an investor to believe as a moat. When you pitch into a crowded space, do not compete on the feature everyone lists. Define the workflow you own end-to-end and make that the category. Investors fund defensible position, and a loop is more defensible than a widget.

5. Make capital efficiency an explicit part of the pitch. Promptwatch did not hide that it is a small, lean, repeat-founder team that has sold a company before. It foregrounded it, and its investors repeated it back. Serial founders who build efficiently command a valuation premium, often 1.5 to 2x a first-timer at the same traction. If that is you, say so plainly, in numbers: revenue per euro raised, headcount, burn. Do not let capital efficiency be a thing investors have to discover in the data room. Put it in the first paragraph, because in a market drunk on mega-rounds, restraint is a differentiator.

The money is pointed at America. A New York office, a push into US enterprise and agency accounts, and a deeper set of agentic features are the near-term milestones. The open question is not whether the European machine works, it clearly does, but whether it travels, and whether an efficient team can hold its ground as much better-funded rivals crowd the same category. The next round will be priced on the answer. For now, Promptwatch has done the rare thing of making frugality look like the flex.

The paid Playbook goes deeper: the full cap-table breakdown, the capital-efficiency math and dilution estimate, an honest competitive read against Peec AI and Profound, what this round does not yet answer, and the exact “raise on a ratio” outreach templates.