Every roadmap I have ever shipped — and I have shipped them across several companies, my own and other people's — was presented to the world as a plan and functioned, in practice, as a negotiation. The sooner a product manager makes peace with that, the sooner they stop being disappointed by their own documents. [author assertion]
A plan assumes the future will cooperate. A negotiation assumes interests conflict and must be rebalanced as facts arrive. Roadmaps do the second of these while dressed as the first, which is why they generate so much theatre: the quarterly review where everyone pretends the document was a forecast, when everyone in the room knows it was a settlement. [author assertion]
Look at any roadmap item and you will find a bid. Sales is bidding for the feature that closes next quarter's deal. Engineering is bidding for the rewrite that keeps the system standing. The founder is bidding for the narrative that makes the company legible to investors. The dates attached to the items are not predictions; they are exchange rates between these currencies, quoted with false precision because nobody wants to write "it depends" in a box that will be screenshotted. [author assertion]
What good product managers actually do, in my experience, is keep the negotiation honest. They make the trade-offs explicit — this for that, now for later, this customer for that segment. They re-open settled items when the facts change, without treating the re-opening as an admission of failure. And they protect a small core that is not up for trade at all: the problem being solved, and for whom. A roadmap where everything is negotiable is not a roadmap; it is an auction catalogue. [author assertion]
The failure mode I have committed most often, and watched most often, is defending the document instead of the negotiation. By the third quarter you are no longer managing a product; you are managing the gap between a slide and the truth, and the gap eats exactly the attention the product needed. The teams I have seen escape this trap share one habit: they publish the roadmap with its assumptions attached. When an assumption dies, the renegotiation happens in the open, on the record, without a culprit. [author assertion]
None of this means abandoning commitment. A negotiation that never settles is just a meeting, and teams need settled stretches in which to do deep work. The craft is in choosing the grain: fix the near term in detail, hold the far term in themes, and be explicit about which is which. Most roadmap arguments I have refereed were really arguments about grain — one side defending a promise, the other defending a possibility, both using the same boxes. [author assertion]
Here is my practical rule, bought at some cost. A roadmap should state what we believe, what we are trading, and what would change our minds — and it should be reissued, cheerfully, whenever any of the three moves. Teams forgive changes. Customers forgive changes. What nobody forgives is a surprise they can prove you saw coming. [author assertion]
