Benchmarking cheat sheet
Benchmarking cheat sheet
Use this quick reference to benchmark Walnut demos and playlists, diagnose performance gaps, and choose the right optimization lever — without digging through every dashboard.
1️⃣ Before You Benchmark (Don’t Skip)
- Start with purpose: Confirm why this asset exists (discovery, conversion, onboarding, expansion) and what “success” should drive (explore, book, trial, adopt).
- Wait for processing: sessions finalize after inactivity; Insights refreshes periodically; CRM/MAP sync may lag.
- Use a stable window: last 14–30 days
- Segment before you compare: External vs Internal, Identified vs Anonymous (use Hide Bounced Sessions when diagnosing mid/late-funnel behavior).
- Check identification coverage: aim for 70–80% identified sessions (use Uncover for company-level + forms/URL params for contact-level).
2️⃣ Core Benchmark KPIs (Use These Everywhere)
| KPI | What it tells you | If it’s low… |
|---|---|---|
| Completion Rate | Story quality & relevance | Value arrives too late, flow is too long, or navigation is unclear |
| Bounce Rate | Hook & gating effectiveness | Opener is weak or gate appears too early |
| Median Time Spent | Depth of attention (outlier-safe) | Confusion, fatigue, or unclear “what’s next” momentum |
| CTA Conversion (FAB or other CTA) | Intent to act | CTA is too late, too generic, or not aligned to viewer stage |
| Engagement Score | Overall asset health vs top performers | Structural, pacing, or relevance issues |
| Identified vs Anonymous | Attribution readiness | You’re missing ROI signals and undercounting high-intent behavior |
3️⃣ Benchmarks Change by Funnel Stage
| Funnel Stage | Primary Benchmark Focus |
|---|---|
| Top of Funnel | Bounce ↓, early value/early completion ↑, identification coverage ↑ |
| Mid-Funnel | Completion ↑, median time ↑, CTA conversion (FAB) ↑, engagement score ↑ |
| Bottom of Funnel | Return sessions ↑, CTA conversion (FAB) ↑, late-stage screens/sections viewed |
| Post-Sale | Repeat sessions ↑, completion consistency ↑, playlist completion ↑ |
4️⃣ Which Tool to Use (Fast Diagnosis)
| If you see… | Use this tool | To answer… |
|---|---|---|
| Low completion | Guides Funnel / Screens Funnel | Where does attention drop (step-by-step or screen-by-screen)? |
| High bounce | Insights Summary + First Screen/First Step | Is the opener, promise, or gate creating early exits? |
| High time, low completion | Sessions Table + Session Journey | Are viewers stuck, looping, or getting lost? |
| Low CTA conversion (FAB) | Top Screens + Guides Funnel | Is the CTA too late, unclear, or missing the peak value moment? |
| Playlist drop-off | Playlist Item Insights | Which item breaks momentum (or fails to earn continuation)? |
5️⃣ Guided Demo vs Playlist Benchmarks
| Asset Type | Primary KPIs | Optimization Focus |
|---|---|---|
| Guided Demo | Guide completion (chapters), steps viewed, median time, bounce, CTA conversion (FAB) | Narrative pacing, annotation length, gate placement, CTA timing/copy |
| Playlist | Playlist completion, avg/median time, played %, item-level completion | Sequencing, first-item strength, scope control (3–5 items), “recommended next” momentum |
6️⃣ Common Fixes (Metric → Action)
- Bounce high: strengthen the opener (clear promise + value fast); delay gating until after early value (often steps 3–5).
- Completion low: move the “aha” earlier; shorten paths; trim/merge redundant guide steps.
- Time high, completion low: simplify navigation; clarify “what’s next”; reduce modal/text friction.
- CTA conversion (FAB) low: move CTA closer to the peak value moment; use action-driven copy (Book, Start, Unlock, Continue); offer more than one CTA moment (not only the end).
- Playlist exits early: start with the most visual/fast-to-value item; rename items by outcome; limit to 3–5 items.
7️⃣ Advanced (Optional): ROI Benchmarks
- Stage conversion (with vs without demo views)
- Average sales cycle length by demo activity
- Win / loss by demo engagement
- Open opportunities with low engagement (re-activation targets)
Rule of thumb: Only use ROI benchmarks once identification coverage and integrations are solid (otherwise attribution will be incomplete).
Was this helpful?