atrivo.labs (DevHumor) gets lots of views but few follows because the account does not read as legit/established. An official-looking logo with a verified star on every post signals credibility and lifts view-to-follow conversion.
Success criteria
Conversion ≥0.1% → win, roll the badge to all posts. Lift but <0.1% → keep badge, stack end-of-post tagline. Baseline 0.03% (recompute on reach — views double-count).
Key metrics
View-to-follow conversion %, followers gained, views/reach per post
Baselines
As of 2026-06-23:
- Followers: 1,000+ (crossed 1k on 2026-06-08)
- Conversion: 0.03% (views→followers) — recompute on reach before locking; views double-count, reach is the honest denominator
- Target: ≥0.1% (a ~3.3× lift)
Results
12 badge posts, Jun 23 – Jul 1. Total views 17,970 (avg 1,498/post). Per-post 'Follows' insight, manually verified in-app: 8 total → attributed conversion 0.045% vs 0.03% baseline (~1.5× lift), below the ≥0.1% target. Account-level numbers were dominated by older viral content: ~400k background views in the same window at baseline conversion (~120 follows) explain nearly all follower growth (+~130–146). With n=8 follows (±5–6 at 95%), the lift is not statistically separable from baseline.
Conclusion
Per the pre-registered middle branch: legitimacy may help — keep the official badge on all posts (it costs nothing and didn't hurt), stack the end-of-post follow tagline next, and don't expect branding alone to 3× conversion. For a rigorous v2, run badge on/off ALTERNATION so both arms share account momentum and bio-path noise, and score with per-post Follows ÷ views. Method lesson for all future conversion experiments: scraped account-level conversion is an upper bound only; IG per-post Follows is the verdict-grade metric (manual in-app tally or professional-dashboard export at close-out).
Log
Baselines drafted, badge change defined (official logo + verified star on every post).
Registered in the Experiments register from the draft lab note (#9, still in Blog Posts as Draft).
Series identified and tagged in Content: 12 posts, 2026-06-23 11:54 (“It’s paradoxical, yet it is true”) → 2026-07-01 08:30 (“Why it’s always like this?”). Posts after Jul 1 are outside the experiment chain.
Metrics policy: views-first (scrapeable twice daily); reach only as quarterly export calibration. Baseline 0.03% stays views-denominated — consistent with the tracker data.
In-app cross-check (Kish): account-level views Jun 23–Jul 1 ≈ 400,000 (dominated by older viral posts still circulating). Per-post 'Follows' insight summed across the 12 experiment posts ≈ 8. Sanity check: 400k background views × 0.03% baseline ≈ 120 follows, consistent with observed account growth (~+130–146 in the period) — old content at baseline conversion explains nearly all follower growth. Experiment posts' own attributed conversion: 8 ÷ 17,970 ≈ 0.045% — a ~1.5× lift over the 0.03% baseline, below the 0.1% target. Note: n=8 follows, so noise is large (±5-6 at 95%); result is not statistically separable from baseline.
CLOSED. Per-post Follows manually verified in-app across all 12 posts: total 8. Verdict recorded as Inconclusive (lift observed, below target, n too small).