Eudaven Ads
← Back to brief

Persona

David

David · 52 · Phoenix, AZ

Brief angle

“How to tell a serious compounded program from a warning-letter operator.”

Variant hook

“Patient-specific compounding requires patient-specific clinician evaluation. Both should be visible to you.”

Operator action

Regenerate this variant

Re-runs the copy polish, compliance check, brand scorer, and synthetic perf scorer on the same hook. Creates a new version (v2) as a child of the current one. Costs ~$0.04 in OpenRouter calls and takes ~10 seconds.

Version timeline

1 version · newest first

v1

current May 11, 00:12
pending brand
Compliance 0.88
Brand 0.86
Perf —

“Patient-specific compounding requires patient-specific clinician evaluation. Both should be visible to you.”

At Eudaven, every compounded GLP-1 prescription is preceded by a licensed clinician review of your labs, history, and clinical context. Your medication is filled by a named 503A partner pharmacy with COA documentation available. You should be able to trace both — the prescriber and the pharmacy — before you start.

Talk to a licensed clinician

How to use

Variant detail

Full state for one AdVariant: copy, scores, review history, and image assets per version. This is the page reviewers and operators work from most.

How to use this page

  1. Read the hook at the top — that's the headline shipping to Meta.
  2. Scroll to Copy blocks for the polished body + CTA. Character limits are enforced (hook ≤40, body ≤125, CTA ≤20).
  3. Versions shows the history — newest first. Click any to compare diffs.
  4. Assets (loads in below) shows concept thumbnails + final 1:1 / 4:5 / 9:16 renders. Hover the section's ? for the cost breakdown.
  5. If scores look bad, click Regenerate to spawn a new version with the same hook but fresh polish.

Score reference

Compliance — pass/warn/blocker against GLP-1 rules.

compliance_score = 1.0 − Σ(blocker·0.50 + required·0.25 + warning·0.10)

Brand — Sage Vitality voice rubric. Personas have thresholds: Maya/David 0.80, Jess 0.85, Rachel 0.95.

Synthetic perf — judge-predicted performance score; advisory only until real metrics are ingested.

Related pages

  • Review — clinician / brand / legal sign-off queues
  • Concept pick — choose the visual concept for final renders