Wireframes — Content & growth (W30–W34)¶
Visuals: wireframes-content.html · Slide deck
The engine that keeps the website, socials, and blog fed — and honest. Vendor watch (W34) turns supplier changes into proposed content; the doc library (W32) publishes homeowner how-tos to the web and into AI grounding; the blog studio (W33) and social studio (W31) carry the same rule as every AI surface (a human always publishes); and the experiments console (W30) tests models, ads, and landing pages in one framework with consent-aware guardrails.
W30 — Experiments console (A/B)¶
One A/B framework for three things that are usually tested in three different tools: AI model canaries (fed from the W29 lifecycle), ad creative, and unlisted segment-specific landing pages. Guardrails are platform-level: minimum sample before reading results, auto-stop on guardrail-metric regression, and consent-aware assignment — unlisted pages are targeted, never deceptive, and disclose tracking like any other page.
Visual: wireframes-content.html
W31 — Social studio¶
Company socials in one place: a shared calendar across Facebook, Instagram, Google Business Profile, and Nextdoor; AI-assisted drafts grounded on real jobs and the doc library; and an engagement inbox whose comments/DMs route into inbox triage (W26) so a social lead becomes a CRM lead. Job photos can only be used where W20 visibility is "marketing-approved" — homeowner consent travels with the asset.
Visual: wireframes-content.html
W32 — Doc library (homeowner how-tos, pushed to web)¶
A versioned library of customer-facing documents — maintenance guides, storm checklists, warranty care — published to the company website with one action. The same documents are embedded into pgvector, so the composer (W23) and homeowner Q&A (W12) cite them as grounding: write the answer once, and every AI surface uses the approved version.
Visual: wireframes-content.html
W33 — Blog studio (AI-assisted writing & tracking)¶
An editorial pipeline: ideas (seasonal, vendor watch, campaign needs) → AI-assisted draft grounded on the doc library and real job data → human editor review → publish to web → tracked through the attribution engine to leads and signed jobs. The editor gate is the same rule as everywhere else: AI drafts, a person publishes.
Visual: wireframes-content.html
W34 — Vendor watch → content pipeline¶
Monitors supplier catalogs (Beacon, ABC, SRS price/product files already integrated for ordering) and manufacturer feeds for new products, discontinuations, and spec changes. Each detected change proposes actions — doc library draft, blog post, social post, catalog update for the quote builder — every one landing in a human queue, never auto-published.
Visual: wireframes-content.html