Buy an under-marketed brand.
Switch the demand on.

A 12-expert business council (Drucker, Taleb, Porter, Duke, Christensen, Goldratt, Thompson, Levels…) stress-tests the deal for the risks not in the register. Surfaces 14 blind spots — the operator-capacity trap, the supply engine that walks out, three platform landlords, inherited warranty, and the missing Dean agreement — and what to add to the offer.

The skeptical read. Earnings-quality forensics, the red-flag register, valuation & return scenarios, financial-health scorecard, and a conditional, earnout-protected deal structure. The verified-downside view.

Three years of accountant P&Ls verify the IM's earnings — the "expense halving" was shareholder salary, not fiction; SDE reconciles at ~$172k. Plus three corrections that reshape the recommendations: advertising, storage and stock-at-cost.

Vendor finance, modelled with real numbers: three ways to structure the ~$135k goodwill, the ~$3,940/mo repayment, exactly where it sits in the P&L (interest vs principal), and whether you can pay yourself and service the deal in year one. 🔒 Internal — negotiating position.

The complete operating & growth plan behind the pitch — opportunity, market, competitive position, the five demand-side levers, channel roadmap, financial model, assumptions & risks.

The no-raise bootstrap reframe: don't buy the (largely aged) stock — take it on consignment, clear dead stock via $1 auctions, consolidate four storage units into one Auckland warehouse, and fund growth from cash flow.

The one named refurb rival, read cold from primary sources: a US-HQ'd, ~$71M-VC marketplace of 100+ third-party vendors (incl. a Dyson collection). Head-to-head on 11 dimensions, where each wins, and the twist — Reebelo is both a rival and a shelf TYDI plans to sell on. No new red flag; sharpens where TYDI must concentrate.

The brand-specialist play — focused single-brand NZ sites (Dyson first, Apple next) that beat the big-box generalists on SEO, conversion & ads. Plus the Google Ads / trademark compliance rules: what you can and can't do, and the one naming rule that matters.

Refactor TYDI to run on two operators + a picker — AI & Claude co-work as the operating layer, mapped function-by-function to the skills already in the brain. Plus the real edge: a catalog built to be sold by AI shopping agents.

The foundation: what the Trade Me API can actually automate (the whole per-listing lifecycle), what the best eBay refurb sellers do, and the photo/IMEI → live-listing pipeline that turns the gaps into a moat. Plus the two 2026 API-registration rules that are deal-checklist items.

The inventory spine: one system that owns the stock and links Trade Me ↔ Shopify — list once, sell once, withdraw everywhere. The modules, the one thing every off-the-shelf tool gets wrong (unique-unit vs quantity sync), and the build-vs-buy call.

The Phase-2 goal: one inventory hub as the single source of truth. Add a unit once, it lists to Shopify, Trade Me, Meta & Google; it sells once, and the system withdraws it everywhere, prints the dispatch on a barcode scan, books it to Xero, and fires the conversion back to the ad networks. Every API, the fulfilment workflow, the build path.

The deep-dive: email as a 15–25% revenue line TYDI isn't running. Segmentation, the automated flow library (cart, replenishment, win-back), a quantified potential model, and the dead-stock monetisation engine marketplaces structurally can't run.

Why organic traffic collapsed ~99% (4,345→40 keywords) — a fixable, un-redirected Magento→Shopify migration, not a penalty. The diagnosis, what's recoverable, the fix-vs-rebuild decision, and the recovery roadmap.

Zero presence on Google despite ~$30k/yr already spent on other advertising — ~12,000+ NZ "refurbished" searches/mo at cheap CPCs. Campaign architecture, unit-economics & ROAS model, and a costed 6-month budget (~NZD $27.5k, $2k→$6.5k ramp).

Two-part deep-dive: Trade Me listing optimisation (Best Match, the ranked CRO checklist, the new lower-fee economics) for the channel selling today — and website CRO (collection pages, product pages, advertorial LPs) for the 10×+ once it's measured and rebuilt. Trust is the lever.

A 19-year, 99%-feedback Top-Seller account (a moat) running a fast volume process — a 20-listing super-scraper audit shows 1 photo, zero attributes, no "Refurbished" in titles, no Sponsored Listings. The process analysis, improvement plan, Sponsored Listings economics & Best Match SEO.