Balkrishna Attal

Blog / Resource

ResourceSeptember 27, 2026·11 min read

A Real PRD Template (With the Reasoning Behind Every Section)

A copyable PRD template, plus the reasoning most templates skip: why this exact order of sections is what makes a PRD a decision record instead of a justification.

— views
Quick answer

A good PRD (Product Requirements Document) has six sections: Problem, Goals, Non-Goals, Solution, Success Metrics, and Rollout Plan — in that order. The full copyable template is below. The most common mistake is starting with the solution instead of the problem, which is exactly backwards.

Problem Goals Non-Goals Solution Metrics RolloutPlan

Why this order matters

Every section in this template exists to force a decision before the next section can be written honestly. You can't write real success metrics if you haven't named a specific problem. You can't defend a solution if you haven't stated what's explicitly out of scope. Writing the sections out of order — solution first, problem retrofitted afterward — is the single most common reason a PRD reads like a justification instead of a decision record.

The full template

Copy this

Title: [Feature name] — PRD
Author: [Your name]   Status: Draft / In Review / Approved   Date: [Date]

1. Problem
What specific problem are we solving, for whom, and what evidence do we have that it's real? (Data, user quotes, support tickets — not just an opinion.)

2. Goals
What does success look like if this works? State it as an outcome, not a feature description.

3. Non-Goals
What are we explicitly NOT trying to solve with this, even if related? This section prevents scope creep more than any other.

4. Solution
What are we building? Include the core user flow, key screens or interactions, and 2-3 alternatives considered and rejected, with a one-line reason each.

5. Success Metrics
What number moves, by how much, in what timeframe, for this to be considered a success? Include a guardrail metric — something that should NOT get worse.

6. Rollout Plan
Phased rollout, feature flag strategy, who needs to be informed (support, sales), and the rollback plan if something goes wrong.

Common PRD mistakes

Writing the solution section first and reverse-engineering a problem statement to justify it. Skipping Non-Goals entirely, which is exactly how a two-week feature quietly becomes a six-week one. And writing success metrics so vague ("improve engagement") that no launch could ever definitively fail against them — which defeats the entire purpose of writing them down.

A worked example, filled in

Here's the same template filled in for a real (illustrative) feature, so you can see what "done well" actually looks like, not just the empty structure.

Example: Saved Search for a job-listing app

1. Problem: Users apply the same 3-4 filters every time they search, repeatedly, evidenced by session recordings showing 68% of returning users re-entering identical filter combinations from their previous session.

2. Goals: Reduce time-to-first-result for returning users. Increase searches-to-application conversion by reducing friction before the search even starts.

3. Non-Goals: Not building saved-search email alerts in this phase — that's a distinct feature with its own notification-infrastructure dependency, tracked separately.

4. Solution: A "Save this search" button appearing after a user applies their first filter (not after four, based on prior usage-pattern research). Alternative considered: auto-saving every search silently — rejected, since it clutters the UI with searches the user never intended to keep.

5. Success metrics: 15% of returning users create at least one saved search within 30 days. Guardrail: new-user search completion rate does not drop (confirms the feature isn't adding confusion for first-time users).

6. Rollout plan: 10% rollout for one week, monitored for the guardrail metric, then full rollout if stable. Support team briefed with three sample support responses in advance.

Notice this example is short — a strong PRD is closer to one page than five. Length isn't the goal; a next reader being able to understand the decision without a meeting is.

How long should a PRD actually be?

For most features, one to two pages is the right length — long enough to force the thinking in each section, short enough that engineering and design will actually read it closely instead of skimming. A PRD that runs to eight or ten pages usually means either the feature is too large and should be split, or the document has drifted into explaining implementation details that belong in engineering's own design doc, not the PRD.

PRD vs. one-pager vs. full spec

DocumentPurposeOwner
One-pagerPitch an idea before committing real time to itPM
PRDDefine the problem, goals, and solution once a feature is greenlitPM, with input from design and engineering
Technical design docDefine exactly how engineering will build itEngineering, informed by the PRD

A common failure mode is a PM writing a PRD so detailed it accidentally becomes the technical design doc too — which both slows the PM down unnecessarily and quietly removes engineering's own design ownership from the process.

Frequently asked questions

What sections should a PRD include?

Six, in this order: Problem, Goals, Non-Goals, Solution, Success Metrics, and Rollout Plan. Writing them in a different order, especially solution before problem, tends to produce a weaker document.

What is the difference between goals and success metrics in a PRD?

Goals describe the outcome you're aiming for in plain language; success metrics make that outcome measurable — a specific number, moving by a specific amount, in a specific timeframe.

Why does a PRD need a Non-Goals section?

Without an explicit Non-Goals section, scope creep happens by default, since nothing on paper stops a 'related' request from being folded into the current effort. Non-Goals is what a team points back to when that happens.

Who should write the PRD — the PM or the engineering team?

The PM typically owns and writes the PRD, since it starts with the problem and business context, but the Solution section should be developed together with engineering and design, not written in isolation and handed off.

Want this sharpened for your specific situation?

I take on a small number of career calls and resume reviews each month.

See how we could work together
BA

Balkrishna Attal

A product leader with a background in civil engineering. Writes about product management careers, skills, and the transition into product from somewhere else.