Edwin Vlieg

The through-line

Everything Edwin Vlieg says circles back to a single conviction: get the product right and the rest follows. "We know if our product is good then the numbers will be good" [1], and in the same spirit, "Ik denk dat juist als je je product heel goed hebt staan dat het zichzelf wel verkoopt" [2]. The structural condition that makes this possible is independence. Moneybird has been bootstrapped and profitable since its second year, which he treats not as a badge but as a licence to optimise for product quality instead of margins and growth metrics [1]. Customers, not investors, paid for the company: "Onze klanten die betalen hebben ervoor gezorgd dat we groeien" [2].

The second thread is that scale has to be designed in rather than staffed up. Refusing phone support, automating sign-up, letting AI absorb the routine half of the support queue, running the company on 32-hour weeks and multi-month sabbaticals: these are all the same decision made repeatedly, that the organisation should stay calm and scalable even at the cost of some upside. "Ik lever dan liever in op een potentie om meer klanten te hebben dan dat ik weet dat we niet schaalbaar zijn" [2]. What has shifted over eighteen years is mainly his view of big moves. The 2012 rewrite and the German expansion are both filed under lessons learned, and his current instinct is incremental improvement plus a small number of explicit, resourced bets.

On bootstrapping and the flywheel

Moneybird never wrote a business plan. The founders built an MVP over one summer, charged €10 a month for anything over three invoices, and let paying customers prove the value before committing further [2]. Naïveté was an asset: had they written a plan describing a frontal assault on the big accounting incumbents, they or any investor would have lost their nerve. Instead the market corrected them from the other direction, "Uiteindelijk hebben boekhouders ons moeten overtuigen dat we een boekhoudpakket hadden eigenlijk" [2]. The idea itself came from the same place, noticing his own unmet need and that others noticed he had solved it: "Als ik zelf geen software heb kunnen vinden in Nederland die dit voor mij doet en het valt anderen op dat ik het kennelijk voor elkaar heb, misschien is daar iets" [2].

The early growth mechanism was crude and effective. "Elke euro die binnenkwam die ging gelijk weer terug in het Google account om meer klanten binnen te halen. Dus zo kregen we een vliegwieletje" [2]. That was only possible because the founders were students who did not need to draw an income [2]. He is clear-eyed about the counterfactual on venture money: with a board and investors to manage, he would probably have remained the product person and handed over the CEO title, because that work does not energise him, which is his broader warning to product founders who raise [1].

On not selling, and building the exit into the company

Moneybird will not be sold, on the argument that new management after an acquisition kills product innovation first, and staying in place is what lets him keep improving the thing [1]. The freedom an exit is supposed to buy is instead engineered into the company: both founders take multi-month sabbaticals, and he took four months last year [2]. The partnership itself is the long-term asset, and he jokes about its durability, "Ik zeg ook wel eens dat ik langer getrouwd ben met mijn compillon dan met mijn vrouw" [2]. The positive case is simply that there is more to do here: "So why not put the time in Money Bird and make it even greater" [1].

On the 32-hour week

He works 32 hours and thinks the standard is already too long: "I think 40 is too much. I currently work 32 actually" [1]. The claim is not that effort is optional but that time should be spent deliberately, "We of course we put in the effort, but yeah, we also prove that you need to be smart about the time, not filling all your weeks up" [1]. It sits alongside the sabbaticals and the async support model as part of the same argument, that a company can be built to run at a sustainable pace and still reach €20M-plus in revenue [1][2].

On scalability as a design constraint

Phone support was refused on principle. Email, and now AI, means the team can serve the next hundred customers the same way it served the last hundred without adding headcount in proportion [2], and asynchronous handling lets the team work on its own terms and keeps the organisation calm [1]. The same logic runs through onboarding: "Ik wil niet dat als een klant op vrijdagavond terwijl ik al een biertje zit te drinken op het terras zeg maar denk van ik wil klant worden, dat hij geen klant kan worden omdat we dat niet geautomatiseerd hebben" [2]. Where scalability and reach conflict, scalability wins [2]. The commercial side is built the same way: Moneybird reached €20M-plus in revenue through product-led growth with zero salespeople and zero account managers, on the view that a good enough product plus a marketing machine can stand in for a commercial organisation [2].

On AI: accelerator, not replacement

He is an advocate, to the point of saying that if he were starting a company today he would do a great deal of it entirely with AI: "Ik denk dat als ik aan nu een bedrijf zou beginnen dat ik heel veel dingen volledig met AI zou doen" [2]. But he tests the limits rather than assuming them. A workshop that forced engineers to use only Claude Code produced a clear verdict: "It's a very powerful tool that enables us to be go become faster but it's not like the engineer is not needed anymore" [1], with humans still required for performance and user experience. The effect is that the quality bar rises rather than the team shrinking. Small fixes that took a day now take an hour with Claude, so teams simply take on more, and juniors are still hired and trained, at a higher starting level [2]. For his own role it has been liberating: as CEO he can now ship finished product work himself instead of only internal prototypes, which keeps him connected to the product [1].

Inside the product he is deliberately conservative. Roughly 80% of the accounting is automated, but mostly through rules, because accounting is deterministic; AI is used to detect those rules and around the edges, finding invoices in mailboxes, or an MCP chat with your numbers [1]. In support, AI answers 60 to 70% of customer questions, so the support team has grown more slowly than the customer base, though the residual work has become harder and shifted towards knowledge management rather than vanishing [2].

On Shape Up, no backlog, and strategy bets

Moneybird has no real feature backlog. "You could create a backlog, but the backlog will be infinitely long and that's not really motivating for a team" [1]. Prioritisation comes instead from continuous customer conversations and memory, structured by Shape Up [1]. The rhythm is eight-week cycles, six weeks of project work and two of cooldown, with shaped pitches carrying an appetite and de-risked rabbit holes, so that a dependency such as API access must already exist before a project starts, and an open betting table that creates alignment across the team [2]. At the company level the same idea replaces the multi-year plan: strategy bets, each defining what success looks like, what the gamble is, and what they are willing to invest in money and people. Belgium and the payment account are the worked examples [2].

On rewrites and incremental change

The multi-year rewrite begun in 2012 is his clearest mistake. It cost years, during which no new features shipped on the original product, and those were the years of fastest growth [1][2]. He draws the lesson twice over: you do not always need to update your product to grow, and in hindsight he would have improved the existing product incrementally rather than rebuilding it [1][2].

On owning the customer relationship and becoming a bank

Selling directly to entrepreneurs rather than through accountants was never a deliberate strategic choice, but it turned out to be the decision that unlocked everything downstream, because owning the end-customer relationship is what made financial services such as bank accounts possible [2]. When the banks refused to cooperate on API access, the response was characteristically direct: "The banks didn't want to corporate with us. So we said okay let's start a bank ourselves" [1]. In practice that means a partnership with Adyen, who holds the licence and the compliance burden while Moneybird builds the banking experience, producing 24-hour account opening and fully integrated credit card statements [1].

On new markets and how regulation shapes them

Germany failed, and he is specific about why: at 10 to 15 people there was no critical mass, and a new market demands dedicated capacity and urgency you cannot spare while the home market is growing quickly [2]. Belgium, funded from cashflow, has been the corrective, and the central learning is that the accountant matters even more there than in the Netherlands [1]. The cause is regulatory rather than cultural. Belgium imposes hefty fines on VAT corrections while the Netherlands allows unlimited corrections without penalty, so Belgian entrepreneurs lean far harder on their accountant, and the entire go-to-market has to change accordingly [2].

On staying close to the product

Quality feedback loops are built into how the team works. Engineers rotate into support, fixing bugs including their colleagues', which creates a self-improving loop instead of code being thrown over the fence [1]. Moneybird still runs its own bookkeeping entirely in Moneybird, which means the company hits scale problems first, and most product improvements originate in Edwin doing his own VAT return [2]. His advice compresses to the same instinct: "Ga lekker beginnen, luister de podcast en praat met je klant. Ik denk dat het allerbelangrijkste is. Hoe groot je ook bent" [2].

Takeaways

• Bootstrapping is a product strategy: profitability from year two removed the pressure to optimise margins and growth metrics and left room to optimise product quality instead [1].
• Reinvest revenue directly into acquisition to create a flywheel, which is easiest when founders can go without an income early on [2].
• Refuse channels that do not scale. No phone support, no salespeople, no account managers, and fully automated sign-up so a customer can join on a Friday night without anyone present [1][2].
• Treat AI as a quality multiplier rather than a headcount cut: fixes that took a day take an hour, teams take on more, and juniors are still hired and trained at a higher level [2].
• Keep AI out of deterministic parts of the product. Automate accounting with rules and use AI to detect rules and handle the edges [1].
• Do not maintain a backlog. It grows infinitely long and demotivates the team; prioritise from live customer conversations inside eight-week Shape Up cycles with an explicit appetite [1][2].
• Replace the multi-year plan with strategy bets that state the success criteria, the gamble and the money and people you will commit [2].
• Avoid the full rewrite. The years Moneybird spent not touching its old stack were its fastest-growing, and incremental improvement would have been the better call [1][2].
• Do not enter a new market without dedicated capacity. Germany failed at 10 to 15 people because the home market absorbed all the urgency [2].
• Study the regulation before the go-to-market. VAT correction penalties in Belgium make the accountant a far more central figure than in the Netherlands [1][2].
• If a partner blocks you, take the position yourself: when banks refused API access, Moneybird built the banking experience on Adyen's licence and compliance [1].
• Build the freedom an exit would give you into the company, through sabbaticals and a 32-hour week, rather than selling to get it [1][2].