DocsHow to write a PRD in Worklayer

How to write a PRD in Worklayer

Use Case: When you need to spec a new feature or initiative and want a structured PRD in minutes instead of hours Time to Complete: 10-15 minutes Prerequisites: Workspace set up, context files added to Context/Product/, PRD template available in Templates/Work/

Quick Answer

In Worklayer, you can write a PRD in 10-15 minutes by referencing your stored context files (personas, metrics, goals) and using the PRD template. The AI generates a structured draft grounded in your actual product data that you review and save to Outcomes/PRDs/.

Step-by-Step Instructions

Step 1: Attach Context Files

In the AI chat, use @ to reference your stored context files. This ensures the PRD draft is grounded in your actual product data, not generic assumptions.

Example:

Use @Context/Product/user-problems-and-jtbd.md, @Context/Product/personas-and-use-cases.md, and @Context/Product/product-metrics-and-funnel.md to draft a PRD for [feature name].

Step 2: Reference the PRD Template

Add the PRD template to your prompt so the output follows your standard format and includes all required sections.

Example:

Use @Templates/Work/prd-template.md to structure the PRD.

Step 3: Specify the Feature and Goal

Clearly state what feature you're speccing and what business goal it supports. This keeps the PRD focused and actionable.

Example:

Draft a PRD for "Onboarding Checklist Feature" that increases Day-1 activation from 0% to 50%.

Step 4: Send the Prompt and Review

Combine all elements into one prompt, send it, and review the AI-generated draft. Check that metrics, personas, and success criteria match your actual data.

Example full prompt:

Use @Context/Product/user-problems-and-jtbd.md, @Context/Product/personas-and-use-cases.md, @Context/Product/product-metrics-and-funnel.md, @Context/Company/goals-and-okrs.md, and @Templates/Work/prd-template.md to draft a PRD for "Onboarding Checklist Feature" that increases Day-1 activation from 0% to 50%.

Step 5: Save to Outcomes

After reviewing and refining, save the PRD to the correct location so it's discoverable and versioned.

Example output location:

/Outcomes/PRDs/onboarding-checklist-prd.md

Example: End-to-End Workflow

Let's write a PRD for "Onboarding Checklist Feature."

Context:

  • Personas: Overwhelmed PM, Non-Technical PM (from Context/Product/personas-and-use-cases.md)
  • Problem: Users abandon the app during their first session (60% drop-off rate)
  • Goal: Increase Day-1 activation from 0% to 50% (from Context/Company/goals-and-okrs.md)
  • Metrics: Track completion rate, time-to-first-value, and Week-1 retention (from Context/Product/product-metrics-and-funnel.md)

Prompt:

Use @Context/Product/user-problems-and-jtbd.md, @Context/Product/personas-and-use-cases.md, @Context/Product/product-metrics-and-funnel.md, @Context/Company/goals-and-okrs.md, and @Templates/Work/prd-template.md to draft a PRD for "Onboarding Checklist Feature" that increases Day-1 activation from 0% to 50%. Include: - Problem statement grounded in current metrics (60% drop-off) - Success criteria with specific targets (Day-1 activation: 50%, Week-1 retention: 40%) - User stories for Overwhelmed PM and Non-Technical PM personas - Rollout plan with phased release (Alpha → Beta → GA)

Result:

AI generates a complete PRD with:

  • Executive Summary: "Onboarding Checklist Feature reduces first-session abandonment by providing guided 5-step setup"
  • Problem Statement: "60% of new users abandon app during first session due to unclear next steps"
  • Success Metrics: Day-1 activation target: 50%, Week-1 retention: 40%, Time-to-first-value: 5 minutes
  • User Stories: 5 stories covering Overwhelmed PM ("See guided checklist on first session") and Non-Technical PM ("Connect tools with one click")
  • Rollout Plan: Alpha (10 users, 2 weeks) → Beta (50 users, 4 weeks) → GA (all users)

Save to /Outcomes/PRDs/onboarding-checklist-prd.md

Time saved: 2-3 hours of manual writing → 15 minutes with Worklayer

Tips & Best Practices

  • Store personas once: Save personas to Context/Product/personas-and-use-cases.md so you don't re-enter them every time you write a PRD.
  • Reference multiple context files: Use 2-3 context files per PRD (personas, metrics, goals) for richer drafts that feel grounded in real product data instead of generic assumptions.
  • Review metrics section carefully: AI may hallucinate baseline numbers if your context files show "TBD" or placeholders. Always verify metrics against your real data sources before sharing the PRD.
  • Use specific feature names: Instead of "the new feature," say "Onboarding Checklist Feature" so the PRD is immediately clear to stakeholders and engineering.
  • Update context files regularly: If your Context/Product/product-metrics-and-funnel.md is outdated, the AI will generate success metrics based on stale data. Review and update context files quarterly.

Common Mistakes to Avoid

  • Forgetting to attach context files: If you skip @Context/Product/personas-and-use-cases.md, the AI generates generic personas ("As a user, I want...") instead of using your actual target users (Overwhelmed PM, AI-Curious PO). Always attach relevant context.
  • Using outdated context files: If your Context/Product/product-metrics-and-funnel.md shows "Baseline: TBD" or "Current: 0," the AI can't generate realistic success metrics. Update context files before writing PRDs.
  • Not specifying the business goal: If you don't mention the goal ("increase Day-1 activation from 0% to 50%"), the AI may generate a PRD without clear success criteria. Always state the goal explicitly in your prompt.

Related Articles