Every roadmap I've ever inherited had fifty things on it. Fifty things, ranked by a mix of loudness, recency, and whoever got to the PM last. The fifty things were all, individually, reasonable. Together they were a kind of ambient noise — a to-do list masquerading as a strategy.
When I took on planning at Frame.io, I had a hypothesis I couldn't quite defend yet: a company should commit to roughly three bets a year, and actively refuse the rest. Not three themes. Not three pillars-plus-some-other-stuff. Three bets, and the discipline to say no to the forty-seven things that wanted to be the fourth.
It sounds almost aggressively obvious when you write it down. It is not what most companies do.
Why three
Three is small enough that every person in the company can hold it in their head on a bad Tuesday. That's the real test. If the thing you're building your year around can't survive a distracted 1:1, it's not a strategy — it's a slide.
Three is also big enough to matter. One bet is fragile; if it breaks, you have no year. Two is suspiciously neat and tends to collapse into a single priority and a hedge. Three forces honest tradeoffs without inviting the entire backlog to the table.
The point of the number isn't the number. The point is the refusal. Most of what makes strategy work is not the idea — it's the commitment to not do the other things, out loud, in writing, when it's uncomfortable.
What actually happened
When we reorganized planning around three bets, the first week was euphoric and the second was miserable. Euphoric because people felt, maybe for the first time, that leadership had actually chosen. Miserable because half the things people had been quietly protecting were now, officially, not happening.
That second week is where most companies lose their nerve. Someone senior walks into the room with a very good reason why their thing should be the fourth bet. The reason is always genuinely good. That's the trap. If your framework can't hold against a genuinely good reason, it's not a framework; it's a preference.
The hard part of strategy is never the insight. It's the commitment.
We held the line — mostly — and the releases that followed were the ones that actually moved the business. PLG. The core revenue features. The unsexy platform work that made the next year possible. None of them would have happened in a world where we tried to ship everything.
The quiet math
There's a version of this that reads as austerity: less is more, do fewer things better, etc. I don't love that framing. The point isn't to do less. The point is that attention compounds. Three bets, done with the full weight of the company behind them, produce a different kind of result than fifty things done with a tenth of everyone.
And the forty-seven things you said no to? Most of them quietly die. A few come back as next year's bets, sharpened by the time off. A couple turn out to have been someone's pet project and nothing more — and the act of not doing them is itself a gift to the team.
A small warning
"Three bets" is not a magic number. If your company is twelve people, it's probably one bet. If you're at real scale, it might be three per org, with the top-level three being categories, not projects. The useful question isn't "is it three?" It's: how few can you get away with, and still be honest about what matters?
That's the question I keep asking. Most years the answer surprises me with how small it is.