A Clipus report is structured for skimming first, drilling second. The top of the page is the headline. Everything below is evidence.
Section 1 — Executive summary
Three sentences. Read this first.
Section 2 — Findings
Each finding has:
- A claim ("Your pricing page does not surface the Free plan above the fold")
- A confidence band (low / medium / high)
- Evidence (DOM snippet or screenshot)
Section 3 — Sources used
A live list of the source types that contributed to this report. The set varies by report type:
- · Website — DOM + OG meta from your landing pages
- · Reviews — G2, Capterra, TrustRadius reviews
- · Competitors — Top 3 competitor landing pages
- · Market — Competitor landing changes over time
- · Performance — Your own video/social performance trend
- · Internal — Product analytics, CRM, usage data
- · DOM Time-Series — Your product DOM changes over the past 30 days (rollouts, A/B tests). Refreshed weekly via Continuous Intelligence.
- · API Intercept — Live product data captured from API responses
- · Pattern Library — Comparable precedents from other SaaS products
Section 4 — Suggested next actions
Surface Reports focus on positioning and content angles. Deep and Comparative reports can also show an activation rail when there is a recommended move ready for production.
The activation rail summarizes:
- Recommended move — the next video idea Clipus thinks is worth acting on.
- Source evidence — the findings or moves that support the recommendation.
- Output capability — whether the draft should use Proof Mode or Polished Hybrid.
- Studio activation — the button that starts a report-backed Studio draft.
If you schedule the idea from Calendar instead, Calendar keeps the lineage visible so the item does not look like a generic suggestion.
When confidence is low
Low does not mean wrong — it means we found one or two signals, not many. Treat low-confidence findings as hypotheses to verify with your own product metrics.