Strategy decisions , 2026-05-31 (critique session)

Decisions Annette made while pressure-testing the master strategy.

1. Rebuilding Seminars = Case Study Zero

• Use Rebuilding Seminars as SEO Believer's FASTEST published proof (case study zero), not a stranger pilot.
• It's a live, already-signed client generating real data. Fastest path to a published before/after.
• Annette confirmed the Bob/Rebuilding relationship is no longer private (changed 2026-05-31), so a named SEO before/after is publishable.
• Action: build the prediction-scoreboard + baseline capture against Rebuilding Seminars first.

2. Bandwidth reality: 15 hours/week on SEO Believer

• Annette can commit ~15 hrs/week to SEO Believer (around menopause content, YouTube, Rebuilding ghostwriting, etc.).
• Bet: Claude Code multiplies that 15 hrs significantly (code-fixable wins are the leverage).
• Implication for sequencing: the "5 anchors at once" rail is too much concurrency for 15 hrs/wk solo. Serialize harder , one complete published loop first (Rebuilding = zero), gate the next anchor on the prior one's 30-day signal.

3. Open to hiring / partner

• Annette is fine hiring help or taking on a partner ONCE revenue supports a salary.
• This changes the long-game: bandwidth ceiling is liftable, not fixed. Plan can scale past solo limits when money is there.
• Near-term: still solo + Claude Code. Don't plan around hires that aren't funded yet.

Positioning decisions (2026-06-01)

• Scoreboard stays, but NOT the anchor. Annette likes the prediction scoreboard, agrees it doesn't need to carry positioning. It's the proof layer, not the headline. (Resolved Charlie's weak-point #3.)
• Lab-scientist identity is NOT a general selling point. Annette's read: medical technologist / lab background only sells to doctors, medical pros, or scientists. Do NOT anchor general service-business positioning on it. (Charlie's proposed flip is rejected for the mass market; keep it as a niche credibility note for medical/dental verticals only.)
• Denver: room for two. Annette unworried about Volume Nine copying. Denver is a huge marketplace; room for two excellent SEO/GEO firms, and V9 may not even want small service businesses. Copyability of the scoreboard is therefore acceptable.
• NEW differentiator candidate: a YouTube channel on local SEO + GEO strategies for service businesses. Annette is happy to create and host it. This is a content/distribution moat (audience + authority compounding) that fits her on-camera comfort and is harder to copy than a web page. ADD to strategy as a real differentiator lane.
• Proof asset: Bone Voyage 5-year Ahrefs history. DONE 2026-06-01. Verified figures (growth-to-peak, zero paid ads): DR 0.9 -> 62 (Jun 2023), organic keywords 0 -> 31,310 (Jun 2024), referring domains 0 -> 1,187 (Sep 2025), organic traffic 0 -> ~17,754/mo (Jun 2024). Charts in research/bonevoyage-graphs/. CANONICAL NUMBER: DR 0.9 -> 62. The "0 to 69" recollection is wrong per Ahrefs data; never cite 69. Domain: bonevoyagedogrescue.com.

Whitespace finding revised (2026-06-01, direct 6-engine test)

The proxy test OVERSTATED the AI vacuum. Direct queries (Perplexity, Claude, Grok, + ChatGPT sample) show Boulder dentistry/med spa candidates ARE named by at least one engine on high-intent queries. Strategic implications:

• Reframe the pitch from "AI can't see you (vacuum)" to "AI names you on ONE engine; we get you named on ALL + own the structured-data layer." More honest, still winnable.
• Anchor #1 target = a dated, schema-rich Boulder implant-cost page. Every engine anchors implant cost to ONE source (Boulder Dental Center "$3,150 start to finish"). Beating that single cost authority is the highest-probability AI-citation win in the test. Code-fixable, photogenic before/after.
• Claude = proof, not target for med spas (already names RESTOR favorably, never refuses local recs). Use as a demonstration asset since "cited by Claude" is success metric #1.
• Grok = Reddit/X corpus (cites forum chatter others miss). Separate citation path: get clients favorably mentioned on Reddit/X to win Grok.
• TODO: AI Overviews + Gemini not yet directly tested; signed-in Perplexity pass would give richer data. Untested verticals (LASIK, cosmetic surgery, IVF, orthodontics, wealth mgmt) still need whitespace checks before betting.

NEW VERTICAL: HRT providers / testosterone (TRT) clinics , nationwide/international (2026-06-01)

Annette's idea, and it's her strongest vertical fit. Why it's special (not just another vertical):

• The credential flips here. Annette's "lab scientist only sells to medical buyers" objection REVERSES for HRT: she's a medical technologist AND an HRT patient AND a menopause subject-matter expert (SmartStrongAlive, Blueprint biomarkers, sub-therapeutic-estrogen discovery). For an HRT/TRT clinic, "my SEO consultant actually understands estradiol/testosterone/pellets and patient search behavior" is a major trust signal. Being Annette IS the moat in this vertical.
• Breaks the geography cage legitimately. HRT/TRT is increasingly telehealth (ships nationwide, video consults). Buyers research heavily on AI before booking = the exact high-consideration-on-AI behavior the GEO bet targets. Bigger market (national/intl) AND more AI-shaped than local verticals. Cleanest expression of the GEO thesis.
• Dodges the untested-vertical risk: huge informational search volume, authority/structure/freshness (GEO levers) win here.

Watch-outs:

• YMYL + ad-policy. HRT/testosterone are "Your Money or Your Life" topics: higher Google E-E-A-T scrutiny; testosterone trips ad-policy/controlled-substance filters. For GEO/organic this is an ADVANTAGE (rewards real expertise Annette has), but thin/hype content gets buried. Higher bar; her scientific voice clears it.
• Attention overlap with SmartStrongAlive / menopause work. Mostly synergy (shared research/voice), but keep SEO Believer client work distinct from her own publications.

STATUS: adopt as a priority vertical. DECISION (2026-06-01, refined): NATIONWIDE pool of LOCAL (in-person, brick-and-mortar) HRT/TRT clinics. NOT just Front Range, and explicitly NOT statewide/national telehealth (avoids the Hims/Defy battlefield). The play: target in-person HRT/pellet/TRT clinics in ANY US metro using the local-SEO+GEO playbook (each clinic is a geographic map-pack/"near me" business the heat-map engine handles). This expands the prospect pool from a few Boulder clinics to thousands of local clinics nationwide while keeping the exact same local playbook + cusp + code-fixable filter. Geographic reach is national; the SEO work is local per clinic. Win a case study anywhere, then the model replicates city to city. Whitespace test: run LOCAL HRT queries across several representative US metros (not just Boulder) to map where the AI/local opening is widest.

Vertical list tiered by evidence (FINAL, 2026-06-01 , extended test done)

Ranked by case-study attractiveness (opening width x deal value x code-fixability):

1. Dentistry implant-cost (Boulder) , lead. Single cost authority (Boulder Dental Center $3,150) is displaceable by a dated schema-rich page.
2. Orthodontics braces-cost (Boulder) , NEW strong add. Same single-cost-authority shape ("braces cost Boulder" anchors to ONE aggregator, dentistsranked). Second photogenic displace-the-source play.
3. Med spas (Boulder) , keep, proof vertical (RESTOR-type candidates).
4. HRT/TRT, local in-person, franchise-light metros , keep, PORTABLE play. Confirmed: AI surfaces local clinics, ZERO telehealth (Hims/Defy) in answers. Real competitor = multi-location franchises (Gameday Men's, Ageless Men's).
• CITY SELECTION METHOD (corrected 2026-06-01): The 5 sampled metros (Boulder/Denver/Phoenix/Nashville/Tampa/Austin) were a DIAGNOSTIC SAMPLE to detect the franchise-vs-independent pattern, NOT the target list. Nationwide reach = score ALL top-100 US metros by population, then rank by (a) market openness (franchise density in map pack , a MAPS question best answered by the DataForSEO heat-map engine, not AI-engine queries) and (b) WARM-ACCESS overlay: cities where Annette has family/network boots-on-the-ground.
• Warm-access cities (priority for first client, regardless of openness score): Miami (daughter), St. Louis (daughter), Greenville SC (daughter), Boston (sister), Washington DC (brother), Boulder (home). These get an in-person reference / intro advantage. True first-target = a city that is BOTH open AND warm-access.
• Build: ranked top-100 metro scaffold (free, now) + browser spot-check of warm-access cities (free, now). Full 100-metro openness scoring = first real job for the DataForSEO engine once funded ($20-50 top-up).
• HRT FIRST-TARGET SHORTLIST (2026-06-01, browser spot-check): see research/hrt-metro-targeting-2026-06-01.md.
1. St. Louis , TOP PICK. Cleanest opening (deepest independent bench, only 1 franchise location) AND daughter on the ground (warm intro). Best market-openness + warm-access overlap. Pursue client #1 here.
2. Washington DC (open, zero Gameday/Ageless, brother there).
3. Boston (open, franchise-light, sister there).
4. Miami (open-to-mixed, Gameday foothold, daughter there).
5. Greenville SC (MIXED , Gameday has 2 locations locking the men's/TRT lane; women's BHRT lane still open; daughter there).
• Franchise tell (automatable signal): Gameday Men's Health LOCATION COUNT locks a metro. Renew Vitality appears everywhere but never dominates. The heat-map engine scores openness by franchise map-pack density.
• CAVEAT: St. Louis + Boston were thin samples (1 query each). Confirm with a fuller read before pitching.
5. Cosmetic surgery , opportunistic only (open but crowded, high-trust).

CUT / closed (do not pitch as case-study bets):

• Wealth management (regulated, saturated, compliance limits publishing).
• Personal injury law (directory + old-firm saturation).
• LASIK , closed (Dishler + national chains).
• Fertility/IVF , closed (Conceptions/RMA + CCRM own Boulder+Denver).

Honest boundary (unchanged): these are citation-layer/consideration-phase wins, NOT proof AI out-converts Maps today. Lead client cases with the Maps volume win; layer the AI-citation win as ahead-of-the-curve proof.

Build-vs-rent: RankGridEngine (decided 2026-06-01, OVERRIDES panel consensus)

The 4-LLM panel said "rent Local Falcon, don't build the engine." Annette correctly identified the flaw in that consensus: it assumes HUMAN labor economics (build = weeks of $70-150/hr dev time). With Claude Code, the build is cheap (tokens, not Annette-hours), so the "don't build" logic does not apply.

DECISION: BUILD RankGridEngine (Claude Code does it cheaply). BUT the panel's TECHNICAL findings still hold , they are about correctness, not cost:

• Codex's 15 flaws are the BUILD SPEC, not reasons to avoid building. Must-fix before trusting output: send search_places=false + search_this_area=true; record rank_group filtered to type==maps_search (NOT rank_absolute, which is ad-contaminated); structured per-task statuses (ok/empty/timeout/error) so API failures never render as "not ranked" cells; real cost-guard with hard cap + --approve-cost flag; canonical place identity (cid > place_id > normalized name+address+phone) so franchise duplicates collapse; explicit depth=20 if 21+ is the sentinel; postback/resumable ingestion for large scans (national = ~49k tasks, NOT cheap , re-estimate).
• VALIDATION GATE: before any client territory recommendation, validate the engine's output for ONE city against a manual Google Maps check OR a single-month Local Falcon trial ($25, used as a RULER to confirm accuracy, then cancel). Build right + sanity-check once. This is the part of the panel's advice that survives.

Revision to Section 6 (case-study play)

• Was: "Start with 5 anchors, one per top vertical."
• Now: "Case study zero = Rebuilding Seminars (already signed). Then ONE new Boulder anchor (med spa or dentistry, proven-open). Gate anchor #2 on anchor #1's 30-day signal. Scale concurrency only when revenue funds help."