Qualify who is reading

The store shows four paid levels identically to whoever arrives, which sells to everybody and therefore to nobody. Five audiences — founders, investors, C-level executives, cybersecurity professionals, and risk and governance or GRC professionals — each with its own path and its own subset of the offers. The front door stays exactly as it is.

3 of 5 done. Status: next.

From the memo Five audiences, decoupled views, and the version the board can read (2026-09-16) — https://store.sgit.ai/admin/memos/2026-09-16-five-audiences/↗

• QB-1 · Decide what happens to the three buyer groups — next — The store already has three, cut by situation rather than by role, each with a page in the nav and in the sitemap and claims citing them. Five new ones is a replacement and the memo does not say whether the old three are retired, re-cut or kept alongside. Nothing else in this workstream can start without the answer.
• QB-2 · Write the five audiences down as data — done — Shipped in v0.2.7 as data/audiences.yml. One record per audience: who they are, what they arrive asking, which level the view leads with, which are quieted, and the one published vault that speaks to them.
• QB-3 · Hidden, or shown differently — answered — done — Shown differently. Nothing is hidden from anybody. Each view leads with the level that usually fits the question somebody arrived with and quiets the ones that usually do not; every level stays present, reachable and priced on every view. The alternative would have meant this site showing different catalogues to different readers, which is a thing that has to be said out loud before it is built. check_the_five_audiences_hide_nothing is what stops quiet quietly becoming absent — that change is one CSS rule away and would look like a tidy-up in a diff.
• QB-4 · Build the five paths — done — Shipped in v0.2.7. Five paths at /are/, generated — a sixth audience is a file.
• QB-5 · Run a synthetic pass over the five paths — queued — Named in the memo as the feedback loop for this work. The method exists and has been run once; the second run is the one that tells us whether it generalises. Blocked on: QB-4.

This document is released under the Creative Commons Attribution 4.0 International licence (CC BY 4.0).