What Comes After Product Management
I've been thinking a lot the past year about the decision layer of the AI race. You can both be excited for the potential of what's to come, while also asking hard questions about what we're doing and who gets to make the decisions.
The other question I'm circling is about trajectory. If you've spent any time with agentic tools, these workflows can really juice work you'd have on the backburner for years making it cheaper than ever to complete them. This is truly exciting! Silly projects I've had in half-done states for years launched in days! (Like this website for tennis rankings in Oregon)
Where's the management layer of all of this? All of the hullabaloo around the rise in product management, the traditional frame of product is not going to keep up with this. I'm not advocating for more designers here – though they'd help – I think we should consider a secret third thing. Trajectory.
Product management asks: what should we build next and why? That assumes you control the roadmap. That the organization supports your execution. That's not where most teams actually live.
Trajectory management asks: given where we are and what we can't change, what move creates the most value and positions us best for the next move?
It's reading where things are actually headed and intervening before drift becomes disaster. Identifying which variables you can influence. Sequencing decisions to maximize learning. Building momentum through small moves that compound. Anybody on the team at any role can do trajectory management with context, whereas product management requires 1) a hero 2) someone who thinks they're the CEO 3) forces someone to do more with less 4) or burns people out even on the most successul engagements.
The best people already do this regardless of their role and title. They just don't have language for it, and organizations don't reward it because the work is often like the best team development work, you just do it and people don't realize what the impact was until you're gone and not doing it anymore.
I'm going to be writing more about this. I keep seeing posts and articles about failed projects, and they're only going to accelerate as we delegate crucial decisions away from people and onto machines.
Matteo Wong and Charlie Warzel published a piece in The Atlantic this week that is worth reading slowly. Token prices fall as models improve. The faster they fall, the harder it becomes to service the debt that built the infrastructure producing the tokens. One analyst quoted in the piece calls this "a death spiral to zero."
What the piece leaves mostly implicit is the public dimension: who already paid for the build-out, regardless of how the financial bet resolves.
The answer is: you did.