Why this matters
As a Product Analyst, your job is not just to find insights, but to move the product forward. Clear recommendations turn analysis into action that teams can execute.
- Roadmap decisions: Translate trends into prioritized actions.
- Experimentation: Suggest testable changes with expected impact.
- Stakeholder alignment: Give product, design, and engineering a single, specific next step.
- Accountability: Define owners, timelines, and metrics so progress can be measured.
Concept explained simply
A clear recommendation is a one-sentence decision plus a short justification and a way to measure success.
Think: Action + Why + How well know it worked.
Mental model: A-O-W-M-I-R-C-N
- Action: What exactly will we do?
- Owner: Who is responsible?
- When: By when will we ship/decide?
- Metric: What outcome metric will we move?
- Impact: Whats the estimated effect?
- Risks: What could go wrong?
- Confidence: How certain are we?
- Next step: What is the immediate next action (often a test)?
Use this fill-in template
Recommendation: [Action] owned by [Owner] by [When] to improve [Metric] by ~[Impact]. Confidence: [High/Med/Low]. Risk: [Top risk]. Next step: [Run A/B test / ship MVP / validate with users].
Worked examples
Example 1 Sign-up conversion drop
Insight: Sign-up conversion fell from 43% to 37% after adding the address field; mobile drop was largest (down 8 pp).
Clear recommendation: Remove the address field on mobile sign-up and collect it post-onboarding, owned by Growth Engineering, within the next sprint, to raise sign-up conversion by ~4 pp. Metric: sign-up completion rate (mobile). Risk: increased fraud; mitigate with device fingerprinting. Confidence: Medium. Next step: A/B test with 50% traffic.
Example 2 Paywall engagement
Insight: 62% of readers bounce on the paywall; users who see a 3-article teaser before the paywall convert 1.8x more than those who see it immediately.
Clear recommendation: Shift paywall to show after 2 articles for new visitors, owned by Monetization PM, by end of month, to lift trial starts by ~1015%. Metric: trial starts per 1k sessions. Risk: ad revenue dilution; monitor RPM. Confidence: Medium. Next step: 2-week experiment on new visitors only.
Example 3 Onboarding completion
Insight: Users who finish the checklist in 24h have 2.4x week-4 retention. Step 3 (import contacts) has 35% drop-off.
Clear recommendation: Make Step 3 skippable and move it after the aha moment, owned by Onboarding squad, by next release, to raise day-1 checklist completion by ~8 pp. Metric: day-1 completion rate; guardrail: week-1 retention. Risk: fewer connections created; add follow-up nudge. Confidence: High. Next step: Ship change as a guarded rollout (10% 50% 100%).
How to craft clear recommendations (step-by-step)
- State the decision in one sentence starting with a strong verb (Add, Remove, Increase, Defer, Test).
- Assign an owner and a timebox (team and date/sprint).
- Name the primary metric and expected magnitude (a range is fine).
- Call out a key risk or assumption and how youll monitor it.
- Set the immediate next step (experiment, MVP, or rollout plan).
- Keep evidence short: 24 bullets under the one-liner.
Evidence bullets to support your recommendation
- What changed and by how much (include baseline).
- Where the effect is strongest (segment/device/geo).
- Why this action is the simplest plausible fix.
- Any guardrail to protect (e.g., retention, revenue, support tickets).
Self-check checklist
- Is the action unambiguous and testable?
- Is there a named owner and a timeframe?
- Is the metric singular and clearly defined?
- Is impact a rough range, not a vague wins ?
- Is at least one risk/assumption explicit?
- Is there a concrete next step (test or ship plan)?
- Can someone unfamiliar with the context execute it?
Common mistakes (and fixes)
- Vague verbs ( Improve onboarding ) Write a specific change ( Move Step 3 after Step 1 ).
- No owner/timeframe Add the responsible team and sprint/date.
- Too many options Recommend one path; put alternates in backup.
- Metric soup Pick one primary success metric and one guardrail.
- No risks Name the top risk and how youll watch it.
- Unbounded impact Give a realistic range (e.g., 35%).
How to self-check quickly
Read only the first sentence aloud. If a teammate can say What do I do by when, and how do we judge success? youre clear. If not, tighten it.
Practice & exercises
Complete exercises 12 below. Use the template and checklist. Note: The quick test is available to everyone; only logged-in users get saved progress.
Tip: Recommendation one-liners you can adapt
- Add X to Y for Z segment to improve M by ~AB% by [date].
- Remove/Defer X from Y to reduce friction and lift M by ~AB% by [date].
- Test X vs Y for Z users; pick winner on M with guardrail G in [2 weeks].
Practical projects
- One-page Action Memo: Create a memo with 3 recommendations for your product area. Include owner, timeframe, metric, impact, risks, and next steps.
- Before/After Rewrite: Take three past insights and rewrite them into crisp recommendations. Share with a peer for feedback.
- Stakeholder Dry Run: Present one recommendation in 3 minutes using only the one-liner and 3 bullets of evidence.
Who this is for
- Product Analysts turning insights into product changes.
- Data-minded PMs and designers who want actionable, testable next steps.
Prerequisites
- Basic metric literacy (conversion, retention, revenue, confidence basics).
- Comfort with A/B testing or phased rollouts.
Learning path
- Write one-liner recommendations from 3 existing analyses.
- Add owner, timeframe, metric, and impact ranges.
- Prioritize with a simple ICE or RICE score (your rough estimate is fine).
- Prepare a test or rollout plan with guardrails.
- Present and collect feedback; refine wording to be crisper.
Mini challenge
Scenario: Checkout page load time increased by 300ms last week; drop in mobile conversion is 2.1 pp, strongest on 3G networks.
- Write a one-sentence recommendation using the template.
- Name owner, timeframe, metric, impact, risk, and next step.
Show an example answer
Reduce checkout bundle size by deferring non-critical scripts on mobile, owned by Web Perf squad, within 2 sprints, to raise mobile checkout conversion by ~1.52.5 pp. Metric: mobile checkout conversion; guardrail: refund rate. Risk: feature regressions; mitigate with automated tests. Confidence: Medium. Next step: 2-week A/B with 50% traffic.