Hot take. Prioritization by number is kind of silly and a waste of time?

Think about it: how many times does “prioritizing your roadmap” devolve into just assigning numbers to projects to assert an arbitrary level of importance to compare against?

I’ve come to realize it’s just an ineffective way of setting direction.

Not because priorities don’t matter (they do!) but perhaps only at the highest level. Anything lower that (especially a long backlog) feels awful to prioritize this way, because you can’t prioritize correctly unless (1) you have a really clear end state in mind and (2) you know what’s blocked by what. Most teams skip that step and go straight to ranking, which is why the arguments never really resolve. Everyone’s optimizing for something slightly different and there’s no forcing function.

Because everyone’s optimizing for something a bit different, the ranking often becomes political, frustrating, even psychologically unsafe if taken to an extreme.

What have I found way more effective? Sequencing.

Tada! Not that exciting, but hear me out: Instead of trying to stack-rank and horse-trade things, define your ideal end state and work backward. Think through not just the hard dependencies, but also the smaller features or launches that compound as you build on top.

The roadmap starts to naturally write itself. No numbers were harmed in the making of this roadmap.

(The one side effect is that sometimes I feel a bit like Charlie from Always Sunny explaining his workboard 🤪 )

There are multiple “sequences of work” happening right now across the EPD team at @Buffer, and I don’t know that I’d even say any are tangibly more or less important than another. But we have bandwidth to tackle the ones we can, and the launches unlock & compound on one another - sometimes across workstreams.

And now we’re shipping more quickly & frequently than ever before, at higher quality more of the time, and the value is compounding. I think it’s one of a few reasons we’re growing faster now 😉