Summary

Odoo does ~650M ARR growing 50% a year for 20 years, from the Belgian countryside, with 5,000 employees and 50,000+ business customers — and Fabien Pinckaers refuses every exit: “We will never sell… there will never be an IPO.” Liquidity is engineered through secondaries every 3–4 years; the latest was €500M at a 5B valuation, Summit Partners selling to Sequoia, BlackRock, Mubadala, and CapitalG — with more buyers than sellers.
The inversion that breaks VC logic: Fabien wants the valuation LOW, because “in every transaction, I was purchasing shares. All the managers are always buying shares at Odoo. We never sold shares” — to the point of taking out big loans to buy at every round. He raised only twice (€3M at €10M post in 2010; €7M at €30M post on ~€20M revenue), kept 57%, and hasn’t raised primary since 2014.
The core thesis is commoditization of business software: SMBs run a mess of Slack, Mailchimp, WordPress and spreadsheets; ERP penetration is still under 10%; the market will consolidate like operating systems and Office into “one, two, or three players that will have the whole market.” Odoo’s wedge is doing what “Microsoft tried with Dynamics, likely SAP tried, and Oracle with NetSuite tried” and all failed at — every app, integrated, from €20/user.
Pricing was both his worst and best decision. A €10 × users × apps formula triggered a partner revolt and “probably lost one year of growth”; the 2022 cut from ~€120 to €20/user — made while everyone else raised prices into inflation — drove 2.8x more client acquisition. ACV is just 3,500 ARR, word of mouth is the main channel, and outbound only started months ago (30 people out of ~1,800 sales).
Management heresy at scale: no recruited VPs or managers ever (internal promotion only), no budgets shared with teams, no forecasts, no KPIs, no goals for most teams, no recurring meetings, teams of ~10 led by the best practitioner, average employee age 26. Hiring is a work test plus an IQ test (“the second best predictive KPI”), offer within five days. His US KPI parable: the imported star consultant scored worst on every metric — because he got the hardest projects.
SAP is the stated kill target: he bought sorrysap.com at 26 and used it as the homepage for version 14. Enterprise is opening now “thanks to SAP S/4HANA, which is a disaster — nobody wants to go,” and displacement is “just a matter of time… 10 or 20 years.” On AI agents he’s contrarian: Benioff’s Agentforce demo is “overrated” versus good UX, but AI in accounting hit 98.5% invoice recognition, better than humans — start from the problem, not “I want to do AI.”

Deep dive

1. The fat startup: build everything, take two years, nearly die

Fabien coded management software for his father’s friends at 13 (for €150 — “it was not so much about getting money, it was more about building something”), ran a Linux T-shirt e-commerce (“really painful to post all these T-shirts every day… for a student that wants to drink beer”), and an art marketplace selling 15,000 objects a month — more than eBay.be — that made only ~€5,000/month: “the business model was shitty.”
Odoo began as the anti-MVP: “I did the complete opposite. I did everything — accounting, logistic, purchase, everything at once, so it took me two years to build the first version.” The rationale came from a Paul Graham article: “If you want to build a big company, you have to fix a big problem.” Looking back, the integrated big picture “is probably what made Odoo today… it’s not a batch of different software.”
He bootstrapped by selling services — “which is what we’re told not to do,” Harry notes — grew to 100 people that way, and spent years near bankruptcy: “It’s not like there is one time. When you are close to bankruptcy, it’s for months and months… you get customers just because they pay faster, not because it’s a good deal.”

2. Fifteen years to the right business model

The 2010 pivot: kill all services overnight, become a software vendor. But the product was open source, so revenue was maintenance contracts — and customers churned in year two: “The software is good, I don’t need maintenance anymore… they just continued using the software but not paying me.” After two years he was near zero cash; during the open-core pivot he had roughly two weeks left.
The fix was open core — 80% of features free, 20% for a fee — and “then everything gets really good.” Once open core was working, revenue was ~€15M; the model is unchanged today. His summary of the whole arc: “The business model has been hard. Took me 15 years to get to the right business model.”
The pivot detonated the community he’d built on “we will always be open source, property is bad.” Partners defected to competitors; rebuilding trust took about a year. His stated biggest regret: “If I had pitched them correctly — ‘we are open source today, but maybe in the future we will need to monetize’ — I wouldn’t have had that many issues.” The kicker he still savors: open core let him 10x the development team, so “now I’m contributing way more to open source than before. It was bad computation from my side.”

3. Two raises, both at terrible prices — and that was fine

Round one, 2010: €3M at €7M pre from Sofinnova, out of ~10 offers — 30% of the company. “That was not a good valuation.” Round two was worse: €7M at €23M pre on roughly €20M of revenue, with exactly one offer (XAnge). Investors “were scared about everything we had to pivot… we had growth, but not exceptional growth, regular growth.”
The consequence of eating dilution early and never again: no primary raise since 2014, and the company was profitable, with 57% founder ownership. Growth was never a bang — “we basically grew at 50% per year, year after year… it’s not huge. It’s just that I did it for 20 years.”

4. Pricing: the worst decision and the best one

The worst: a price of €10 × number of users × number of applications, growing so aggressively that partners refused to sell it. The trap with partner networks — “you never know if it’s because the price is wrong or because of the resistance to change” — kept him defending it for nine months before reverting. “We probably lost one year of growth.”
The best, in 2022: cut the all-apps price from ~€120/user/month to €20, targeting small clients, “at the time where everybody was increasing their price with inflation.” Result: 2.8x more clients acquired since that date, and the up-and-to-the-right chart Harry saw in the deck.
Cheap is the point, not a weakness. Against Marc Andreessen’s “big company, big prices” doctrine: “If you want to commoditize the market and have every company on your software, you have to have a very small price… What’s the price of Microsoft Office? It’s not a very high price, but everybody uses them.” And why chase a cash cow at all? “We don’t have cash issues. We are profitable. Why would I charge more if I have all the money I need to accelerate?”

5. The commoditization thesis: business software ends like Office

The market structure today: big players (SAP, Microsoft) do many apps “but it’s shitty, it’s slow, it’s expensive,” while point tools (Slack, Mailchimp, Trello) do one thing well. “Nobody succeeded to do the benefits of both… Microsoft tried with Dynamics, likely SAP tried, and Oracle with NetSuite tried.” The rate on the market is still very low, less than 10%.
His history lesson: operating systems consolidated from MS-DOS/DR-DOS/Unix/Minix to three; office suites from Works/WordPerfect/Lotus 1-2-3 to two. “Management software will follow the same trends… there will be one, two, or three players that will have the whole market. That’s my goal at Odoo — just to survive.”
Horizontal over vertical, deliberately: “Instead of being the best in one industry, we wanted to be the best in applications… It’s like a phone. Whatever your industry, you use the same phone.” Why incumbents with money and talent can’t catch up: “It’s more about having a few genius guys rather than an army of 10,000 developers… As companies grow, they defocus themselves. We have been focused all along on this one singular thing.”

6. Management without managers: no budgets, no KPIs, no recurring meetings

Odoo has never recruited a VP, manager, or team leader — internal promotion only, even for new country subsidiaries. Rationale: external managers import habits that dilute culture, and “there is no way you can make a team of developers become better if you are not the best developer of the team yourself.”
No budgets shared with teams, no forecasts (the CFO keeps one “just to check that the cash is okay”), no KPIs, no goals except sales targets. “If you give budget to people, they will just spend the budget… If you don’t, they will have to spend responsibly.” Measuring efficiency? “You don’t need to measure. You know. When you work with people, you know who’s good or not.”
The KPI parable: his US director insisted on customer-satisfaction and delivery-time KPIs. After three months the worst scorer was the star consultant imported from Belgium to rescue the office — “because we gave him all the shitty projects that were complex to deliver. It was obvious to the whole team that this guy was the best. Yet on the KPIs, he was the worst.”
Recurring meetings are banned outright: “Recurring meeting is a way to manage people, and we try to avoid managers… they work with each other every day, so they don’t need.” And most CEOs get the job wrong: “They are constantly meeting people, going to events — I think the company needs them more than the external people.” His own split: 50% in the product, 50% fixing whatever internal process is weakest that month.

7. Hiring: a work test, an IQ test, an offer in five days

The process: watch a developer develop or a salesperson demo, plus an IQ test — “the second best predictive KPI on the performance of the person,” after the core skill itself. No resumes: he had to retrain recruiters who over-weighted the 40-minute resume chat over the two-hour test. One meeting, decision immediately, “usually we recruit within five days. And we beat the market because we are fast, just because of that.”
Hire spiky, not round: “If you look for people that are perfect in everything, you will get people that are average in everything… If you want people who transform your company, they have to be extremely good at something, and you will have to accept their weaknesses.” Culture-fit failures are “less than 1%”; over 90% of dismissals are competence.
His confessed weakness is firing: “I’m always too late… When there is a doubt, there is no doubt. As entrepreneurs we are always optimistic — when you start to think negatively about someone, that’s probably because it’s too late.”
The average age at Odoo is 26, and he rejects the lazy-European-youth trope: “The new generation, they want to work. It’s just that they need a purpose… They won’t work if they feel like they don’t have an impact.” Training: one month of dedicated onboarding, then coaching by team leaders running groups of ~10.

8. Geography: find the person first, and pick tier-two cities

Expansion rule: “We first find the person, and only after the person we define where we go.” Hence an office in Buffalo — “nobody goes to Buffalo” — because the right guy wanted to live there. The San Francisco office is the counterexample he regrets: “You are next to Google and Microsoft and Apple. How can you get the best people with that?” In Buffalo, “nobody’s leaving… retention is the key of growing your business” — a salesperson with five years’ experience sells double a two-year one.
When a region stalls, he relocates himself. India had been stuck at 200 employees for seven years under one of his first employees; Fabien moved to Gujarat for a year with his wife and kids, replaced nobody (“they were all very good, they just needed a little bit of help”), and took it from 16 new clients a month to ~800, 200 employees to 800. Earlier he’d done two years in the US the same way.

9. Never sell, never IPO — and want the price low

The €500M secondary at 5B (Summit Partners selling to Sequoia Capital, BlackRock, Mubadala, CapitalG) “starts to be a correct price” — Harry argues it’s still cheap for 550M revenue growing 50% with good margins. Fabien’s reveal: “I always wanted the price to be low, because in every transaction I was purchasing shares. All the managers are always buying shares at Odoo. We never sold shares” — funded with a big loan because “we don’t have money.”
No IPO, ever: “Public companies tend to refocus on the short term… I don’t want to refocus everybody on earning calls and quarterly results. I like to say anything — I could say I think Odoo is worth way more than 5 billion, maybe 7 billion. If I was public, I couldn’t say that.” The cap table signed up for this: liquidity via secondaries every three to four years, “and they all agreed… if you have a really good business growing like crazy, profitable, it’s gonna be easy to sell.”
On his paper $2.8B: “It’s not like I have money in my bank account. I own a company that’s huge, but that’s it… I have a very small car. I don’t have a house. I live in the house of my wife.” Any spare cash goes into more Odoo shares: “I don’t know another company that can give me 50% year after year — and I’m quite sure about it.” In the quick-fire, offered OpenAI at 160 billion, Anthropic at 40, or X at 50: “I don’t like both of them. They are both risky.”
If handed 2.6 billion, he says it would not change his life: “I’m still the same man than 10 years before.” He drives a small car, has no house, and lives in his wife’s house.

10. Sorry, SAP: the enterprise door is open

At 26 he bought sorrysap.com “because I thought that one day I would be much bigger than them.” When they released what he guessed was version 14 — “much better than SAP at the time,” in his opinion — “we replaced the home page with one big word: sorry SAP.”
SAP remains “the clear and nearly only leader for large companies,” but on displacement: “I don’t know how they can survive what’s coming… we are stealing so much clients from them that it’s just a matter of time. But it’s gonna take time — 10 or 20 years, I don’t know.” The immediate catalyst: “Thanks to SAP S/4HANA, which is a disaster. Nobody wants to go. That opens the door to new entrants like us.”
Enterprise required a separate department and culture adjustment — Odoo’s “extremely efficient, direct” teams clash with big-company culture often in the first or second month of a project. The strategic tell: service is only 18% of revenues by design, in a market “owned by service companies” where an ERP costs millions to implement — Odoo’s methodologies are built to sell as little service as possible. Salesforce’s fate, in one line: “They will continue, but other challengers like us will grow much faster.”

11. AI: overrated as theater, transformative where the problem is real

On Benioff’s Agentforce demo — calling a help desk to change an order — his takedown: a good portal reached from your email “is way more efficient… I’m pretty sure if I call this help desk with my shitty English accent, he won’t recognize me.” The principle: “In a lot of ways, UX — having a very, very good product — is more efficient than AI.”
Where it’s real: accounting. “We have now 98.5% recognition on invoices and bills, which is better than the humans, and that transformed the market. Accountants don’t need to record data anymore. They just validate.” Method over hype: “A lot of people start with the solution — ‘I want to do AI.’ If you start from the problem, where people are wasting time, then you get to good use cases. Sometimes it’s just UX, sometimes it’s AI, sometimes it’s a different algorithm.” On an AI winter: “It’s gonna be interesting. But for sure there will be also very disruptive use cases.”

12. Culture: autonomy, responsibility, evolution — and pick your own title

The three pillars: “autonomy, responsibilities, and evolution. Everybody says that, but for us it’s true.” Responsibility is trained by refusing to decide for people — the barcode-scanner story: if a salesperson asks permission to buy a $50 scanner, “if you answer, he will always ask you. If you tell him ‘you’re smart, decide,’ the way you behave with people will force them to take their responsibilities.” A purchasing process to approve a $30 scanner “will cost you more than $30.”
Titles are self-selected and printed on business cards as requested — funny or grand — “because I want to kill that. What matters is what you do, not the title you have.” Average salaries are published on the website; pay varies by geography and steeply by competence: “The best people should have much more than the other ones.”
The through-line on scale: companies fail by defocus — “there is like a dark force that pushes everyone to complexify the company, and you always have to fight back.” Odoo at 5,000 people: one product, three big departments, 400,000 lines of code. And the founder’s closing self-diagnosis: “You have to be obsessed. I don’t think you can succeed if you are not obsessed… If you look at it like just a job, I don’t think you can get such a fast-growing company.”

Claude Cooperates! Exploring Cultural Evolution in LLM Societies, with Aron Vallinder &Edward Hughes LSTM: The Comeback Story? [Prof. Sepp Hochreiter]