Sameer Al-Sakran

Somewhere in the Metabase admin panel, past the settings most people never open, there used to be a small call-to-action. Click it, pay $300 a month, and the Metabase logo would vanish from your embedded charts. No salesperson called. No demo was booked. There was no support email to write to. That buried button was the entire go-to-market motion, and it pulled in close to six figures in recurring revenue before anyone at the company would have called it a business.

This is how Sameer Al-Sakran runs things. The co-founder and CEO of Metabase has built one of the most widely used open-source business intelligence tools on earth, and he did it largely by refusing to behave like the founders around him. While others raced to monetize, he waited. While others hired account executives, he watched a self-service product outgrow the one his salespeople were selling. The pattern is consistent: trust the product, distrust the playbook.

Metabase today is used by more than 70,000 companies and runs at eight-figure annual revenue with a team of roughly 100 to 140 people. It connects to your database and lets ordinary people - not just analysts - ask questions and get charts back. That sounds simple. Making it simple was the whole point, and the hardest part.

The side project that wouldn't stay small

Al-Sakran has spent most of his working life around machine learning and data. Before Metabase he was a data scientist in residence at General Catalyst, CTO at Blackjet, and held engineering and research roles going back through companies like imeem. He studied at Stanford. He has, by his own framing, spent a career watching smart people struggle to get a straight answer out of their own data.

In 2014 he was CTO at Expa, the startup studio, building shared infrastructure for the portfolio. The recurring frustration was always the same: every new company needed a way to answer basic questions from a database, and every existing option was too heavy. So he built a lighter one. It was meant to be plumbing, not a product.

After six to nine months it had outgrown the studio. Metabase spun out, raised a roughly $2M seed, and added a second seed of about $2.5M two years later. By 2021 it had raised a Series B, bringing total funding to around $51M. The numbers are tidy. The strategy behind them was anything but conventional.

The discipline of not selling

For four years, Metabase charged nothing. This was not neglect. Al-Sakran actively turned customers away. One prospect arrived with a 50-page European contract. Another wanted to embed the product on terms he felt would distort what it was. He said no to both. The reasoning was that revenue taken too early, on the wrong terms, would quietly bend the product toward whoever was paying loudest.

When money finally came in, it came through that buried admin-panel CTA - the smallest possible commercial gesture. No team, no funnel, no campaign. Just a feature people would pay to switch off, sitting where only the people who needed it would find it.

The competitive strategy matched the sales strategy in its refusal to play the usual game. Instead of obsessing over rivals, Metabase ran what Al-Sakran calls a taste test: put the tool next to Tableau, Looker, or Superset, let a user try all of them, and win on which one they actually want to open tomorrow. The goal was to be the daily driver, not the loudest billboard.

The detour that nearly cost him

Then he listened to the advice. Investors and advisors pointed at the obvious move: build an enterprise edition, hire account executives, sell up-market. Metabase did exactly that, and growth nearly stalled. The directly sold business was real but heavy, and it pulled energy away from the thing that worked.

What revived it was a cloud self-service option. Customers could sign up and run Metabase without a contract or a call, and that motion quickly dwarfed the enterprise sales effort. Al-Sakran has since said that if he had trusted the self-service instinct from the start, he might have saved years. The detour, though, taught him something durable about the gap between a generic playbook and his own company's actual strengths.

Pricing, pride, and giving in

For two years he experimented with nearly every pricing model except the one customers kept asking for. Per-server. Per-dashboard. Per-database. Per-installation. Everything but per-user, because he feared per-user pricing would cap adoption - charge by the seat and people add fewer seats. Customers disagreed, repeatedly. Eventually he relented. His verdict afterward was blunt: once he accepted per-user pricing, "everything became easier," and negotiations that used to drag simplified overnight. It is one of the rare cases where the crowd was right and he says so plainly.

What scaling actually broke

Ask most founders about their hardest growth phase and you get a story about money or hiring at scale. Al-Sakran's answer is more specific and more interesting: the jump from 15 to 100 people. The reason is counterintuitive. The company's single greatest strength was its tightly woven product-design-engineering loop, the thing that made the software feel coherent. That loop was precisely the thing that resisted being split into separate functional roles. The weak areas were easy to delegate to strong outside hires. The crown jewel was the hard part to scale, because scaling it meant pulling it apart.

The person behind the patience

The patience is not an accident of temperament so much as a practiced one. He keeps an empty notepad that he fills every morning, and keeps a regular movement discipline going to stay, in his words, in motion. He says he enjoys being corrected, which is a strange thing for a CEO to admit and a useful one to believe. His book of choice is Edward Gibbon's "The Decline and Fall of the Roman Empire," though he warns that his reasons for liking it are "tedious."

His one piece of founder advice rejects the genre's usual optimism. The trait that separates the founders who make it, he says, is "irrational perseverance" - the willingness to keep going past the point where a reasonable person would have quit. Paired with his other favorite idea, playing your own chessboard, it forms a coherent worldview: pick your own game, then refuse to stop playing it.

It would be easy to read Metabase's growth as luck dressed up as strategy. The four-year wait, the buried button, the enterprise detour reversed by self-service - any one of them could be told as a near-miss. Told together, they describe a founder who kept returning to the same small set of beliefs and kept being proven right by them. Version N+1, as he puts it, just needs to be slightly better than version N. Do that long enough, on your own board, and you wake up with 70,000 companies.