Eudaven Ads
← Back to personas

The Plateau Veteran

Maya

A mid-life woman who has tried every diet, hit a hormonal wall,

Persona: Maya — The Plateau Veteran

1. Snapshot

2. Demographics

3. Psychographics

4. Daily life rhythm

5. Past attempts ledger

Approach Outcome Takeaway
Whole30 / clean eating Worked short-term, unsustainable "I can do strict for 30 days, then life happens"
Weight Watchers / WW Modest results, plateau "Counting points became its own job"
Noom Some loss, regained "It's just CBT-themed calorie counting"
Optavia / shake programs Lost, regained, hated it "I felt insane around food"
Intermittent fasting Worked at first, stopped "My body adapted"
Peloton / strength Fit but didn't move scale "I'm fitter and heavier — make that make sense"
Perimenopause supplements Marginal "Maybe placebo"

The pattern: she's done the work. Copy must NEVER imply otherwise.

6. Trigger moments

7. Jobs-to-be-done

8. Aspirational identity

"A calm, present woman who isn't fighting her body anymore. Who doesn't think about food all day. Who walks into a room and feels like herself."

This is not about being thin. It's about getting her back.

9. Self-talk (internal monologue Claude can mirror)

10. Spoken objections (top 5)

  1. "I don't want to be on this forever — what's the off-ramp?"
  2. "I'm scared of the side effects I keep reading about."
  3. "Will my insurance actually cover it, or is this another runaround?"
  4. "Compounded versions scare me — I want the real thing."
  5. "I've been burned by every program before — why will this be different?"

11. Hidden objections (unspoken — acknowledge gently)

The best body copy names one of these without making her say it.

12. Verbatim language (use in hooks/copy)

13. Anti-language (DO NOT USE)

14. Trigger keywords / scroll-stoppers

15. Search query patterns (GSA-ready)

Awareness: - "perimenopause weight gain treatment" - "why can't I lose weight after 40" - "GLP-1 reviews women 40s"

Consideration: - "Wegovy vs Zepbound which is better" - "GLP-1 telehealth branded not compounded" - "Calibrate vs Found vs [Eudaven]" - "Wegovy with insurance how much"

Conversion: - "Wegovy telehealth Aetna" (or specific carrier) - "Eudaven reviews" - "Eudaven cost" - "[competitor] cancel"

16. Trust builders

17. Information sources & social proof

18. Brand affinities (calibrates voice)

Lululemon · Athleta · Madewell · Trader Joe's · Whole Foods · Peloton · Whoop · Apple Watch · One Medical · Glossier · Aesop · Aritzia · Nuuly · Stitch Fix · Solidcore · Equinox (sometimes) · The Skimm · Cup of Jo · Goop (lapsed)

Voice should sit in the Aritzia / One Medical / NYT Well triangle — considered, warm, evidence-aware, never breathless.

19. Decision dynamics

20. Emotional arc

21. Story archetype

"The veteran finally gets reinforcements." Not a transformation story — she rejects that. It's a story of appropriate support arriving for someone who has earned it. Hero stays the hero; we're the ally.

Avoid: rescue narratives, makeover narratives, comparison-with-self narratives.

22. Eudaven fit

23. Funnel stage signals

24. Channel mix (Phase 1 priority bolded)

25. Visual cues (drives image-gen prompts)

26. Offer resonance

27. Regulatory flags (persona-specific)

28. Source notes & diversity caveat

Composite strawman built from competitive teardowns and public-domain market research patterns.

Diversity caveat: This v0.1 skews white, college-educated, suburban. Real customer research will likely necessitate splitting into: - Black women (different cultural body conversation, higher T2D prevalence, physician-of-color preference documented) - Latina women (family-system decision dynamics, Spanish-language ads, different community proof points) - Working-class variant (higher cost sensitivity, employer plan navigation more complex)

Validate against: real customer interviews (20+ patients), search query data, Meta audience insights, PHI-safe clinician intake themes.

Treat as v0.1 — refresh by month 3.

How to use

Persona detail

The full source-of-truth doc for one persona. This exact content is what gets injected into the brief_synthesizer system prompt.

Document structure

28 structured fields, grouped into:

  • Identity — demographics, segment tag, one-liner
  • Pain & promise — what they struggle with, what we promise
  • Objections — what they push back on (and how to answer)
  • Voice cues — words they use, words to avoid
  • Visual casting — used by visual_direction to bound the imagery

How to edit

  1. Edit data/kb/personas/<slug>.md.
  2. Run make kb-seed to re-parse + persist.
  3. Next brief synthesis for this persona will use the new content.

Threshold reminder

The brand_scorer enforces a per-persona threshold. If you tune this persona to be more sensitive (e.g., medical/legal scrutiny), bump the threshold in app/pipelines/nodes/brand_scorer.py.