Add persona-hub architecture spec (the top-level organizing layer) (#24)
Charles's "step back" insight: instead of scattered best-X pages, build buyer-journey HUBS ("The New AR-15 Owner", "PCC Owner") that chain the sub-decisions (which rifle/optic/can/light/trigger...) into one page,
Charles's "step back" insight: instead of scattered best-X pages, build buyer-journey HUBS ("The New AR-15 Owner", "PCC Owner") that chain the sub-decisions (which rifle/optic/can/light/trigger...) into one page, with a live "complete build cost right now" rollup.
Grounded in GDA's own AI-referral data: ChatGPT is ALREADY assembling these journeys by hand from scattered spoke pages (AR-15 sub-decisions = 40%+ of AI traffic; best-556-suppressor alone = 32%). The hub is the page the bot is trying to assemble. GDA already has the spokes + 4 proto-hubs (best-home-defense-setup, best-hunting-setup, etc.).
Wins on 4 axes: SEO (hub-and-spoke consolidates PageRank — fixes the "many pages, low authority" problem WITHOUT new pages, via internal linking), AI (the assembled journey answer is maximally citable), live-data moat (dynamic build-cost rollup competitors can't produce), and JTBD (single decision surface for the whole journey).
- New spec: persona-hub-architecture.md — hub anatomy (real synthesis +
live build-cost rollup, NOT a doorway link-list), internal-linking rules, anti-doorway guardrails, AR-15-owner hub as first build (by data), persona build order, how the other specs slot under it.
- Reframed decision-guide-agent (fills spokes) + comparison-generator
(X-vs-Y within a spoke) as spoke-fillers under the hub layer.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>