The Feature Prioritization Framework We Use With Every Client

The most common state of a client's feature backlog when we start: 47 items in a Notion table with no clear priority, three items marked "critical" that were added six months ago and never started, and a founder who says everything is important. This is not a roadmap. It is a wish list with no decision framework.

Every proposed feature gets scored on four dimensions, each 1-5. Revenue Impact: from "no measurable effect" (1) to "opens a new revenue stream" (5). When a client says a feature will "increase engagement," we push: how does that become revenue? What is the dollar amount of the delta? Most features that seem important become less so when you attach numbers.

User Demand: from "nobody has asked" (1) to "users are leaving because of its absence" (5). We validate with data, not opinions. "Our sales team says customers want X" scores 2 at best without tickets, surveys, or churn conversations to back it up.

Effort is inverse-scored because we bias toward smaller wins: under 8 hours scores 5, over 160 hours scores 1. We estimate in hours, not story points. Includes development, testing, review, deployment, and documentation. We add a 30% buffer because we have never seen a project come in under estimate.

Strategic Alignment: from "tangential to core value" (1) to "directly advances primary differentiator" (5). This prevents feature creep toward mediocrity. A project management tool does not need built-in chat.

Total score ranges from 4 to 20. We rank features, draw a line based on available capacity. Above the line gets built this quarter. Below 8 gets removed from the backlog entirely because it is not worth tracking.

The framework's real value is the conversation it forces. When a founder insists Feature X is critical, we score it together. Often they realize mid-scoring that it ranks below the feature they were ignoring.

We re-score quarterly because priorities shift. If your backlog has more than 20 items, you do not have a prioritization problem. You have a decision-avoidance problem. Score everything, cut ruthlessly, build the 3-5 things that score highest.