Concept

Product Strategy Stack

conceptproduct-strategyroadmapokrsgoal-setting

Product Strategy Stack

A five-layer model for product planning developed by Ravi Mehta, where each layer constrains the one below. The framework’s most distinctive claim is that goals belong at the bottom of the stack — derived from the roadmap, not the other way round.

The five layers

Mission — why the company exists. Changes rarely if at all. Everything in the stack must be consistent with the mission, but the mission itself does not change in response to product decisions.

Company strategy — how the company wins in its market. Directional and multi-year. Defines the bets the company is making about where to compete and how to differentiate.

Product strategy — the bets the product team is making to advance company strategy. This is where the PM has the most leverage. A good product strategy articulates which problems to solve, which users to prioritise, and which capabilities to build or not build.

Roadmap — the concrete plan of work over a defined horizon. Specific, sequenced, and tied to the product strategy. Determines what will actually be built.

Goals — the measurable outcomes the roadmap is intended to produce. Set after the roadmap, not before.

The contrarian insight: goals last

Conventional planning practice starts with goals (OKRs) and derives work from them. Mehta inverts this sequence. The argument: goals set without a roadmap are guesses. If the team has not decided what to build, any metric is equally valid. Goals set after the roadmap are commitments — they reflect an honest assessment of what the decided work should produce.

This does not mean goals are unimportant. It means they are most useful as accountability tools for a plan that already exists, not as drivers of a plan that has yet to be formed.

Velocity vs latency

A related diagnostic from Mehta:

  • Velocity: how much the team ships per unit time. Large teams with mature infrastructure have high velocity.
  • Latency: the time from forming a hypothesis to getting data on it. Startups validate in days; enterprises often take months.

High velocity with high latency means shipping a lot while learning slowly. The Product Strategy Stack implicitly addresses latency by forcing the team to be explicit about what they are trying to learn at each layer before committing to specific work.

Source

See Ravi Mehta on the Product Strategy Stack, Twelve PM Competencies, and Selective Micromanagement.