Sharper findings worth flagging.
The non-obvious stuff. Things that came up while digging that might not be in the main rings but could swing the needle if you act on them. Some are gifts, some are warnings.
1. The "NHI" brand on your flagship Amazon listing
This is the single highest-leverage fix in the entire audit.
Your B07BNCJ1SJ listing (the cooling vest, your most-reviewed product) shows brand "NHI" not "ChillSwift" in 3 separate Amazon search results. The 4-pack replacement listing (B07BND7352) is correctly titled "ChillSwift 4 x Cool Vest Replacement Packs for NHI Cooling Vest", which confirms the relationship but documents the brand-field mismatch.
Every AI engine pulling Amazon data attributes your flagship product to a different brand. Until this is fixed, no amount of SEO will connect ChillSwift to its bestseller in AI answers.
Fix: Amazon Brand Registry 2.0, listing details, change brand field. Should be a 10-minute fix if Brand Registry plays nice.
2. A 404 URL is still being cited by AI
The query "chillswift vs ezcooldown cooling vest comparison" in AI engines surfaces https://www.chillswift.com/products/ultralight-ice-cooling-vest as a cited source. That URL is a 404.
AI engines are still pulling and citing a dead URL from your old Shopify era. The migration broke the 301 chain. The fact that this dead URL is the cited source means AI memory has cached it. Adding a single 301 from that URL to a real product page recovers a piece of cited ground for free.
3. mycoolingstore.com gets 6 AI citations with zero buying guides
The most-cited domain on AI engines for cooling-vest queries (after Amazon) is a Philadelphia family-owned Shopify retailer. They get cited because they have 364 referring domains and product taxonomy. They have no comparison guides. No "how does it work" content. No FAQ pages.
One properly-structured 1500-word "cooling vest comparison 2026" page with schema, a table, and real specs would beat them on AI citation in months. This is the cleanest lane I've seen on the new-site brief.
4. Your Amazon competitors lead title with ICE PACK COUNT
All 3 Rob-confirmed competitor ASINs lead their titles with the number of ice packs (24, 30, 12). This is the differentiator on Amazon for this category. If your B07BNCJ1SJ title doesn't lead with pack count, you're losing scroll attention to listings that do.
Worth a 10-minute A/B test once Brand Registry brand-field is fixed.
5. The site has migrated CMS THREE times in 18 months
PrestaShop until Oct 2023. Shopify until Feb 2025. WordPress until May 2025. Now hand-coded static HTML. Each move broke the URL chain from the previous one.
The April 2025 rankings cliff isn't the Shopify-to-WordPress switch like the original brief assumed. It's the WordPress-to-static switch a month later. Worth knowing because the URL recovery map needs to handle WordPress-era URLs (/author/mondaynomore/, /cooldogvest/) too, not just Shopify ones.
6. Your category on Amazon is Tools and Home Improvement, not Sports or Clothing
All 3 confirmed competitor ASINs sit in Tools and Home Improvement. This matches industrial-worker buyer intent. Worth confirming whether B07BNCJ1SJ is in the same category. If it's filed under Sports or Clothing, you're competing against a different set of listings and getting different traffic.
7. The 2 real backlinks you have are your own forum posts
Of the 155 backlinks in your profile, only 2 are real and editorial. Both are your own PrestaShop support forum posts from 2023. Everything else is spam.
The good news: this means once we disavow the 92 spam domains, the backlink profile is effectively clean. There's no toxic real-link cleanup. The bad news: you start the link-building game from zero. The disavow + future white-label editorial backlinks via Julian Goldie (Phase 7 of the 90-day plan, your call) is the cleanest restart possible.
8. The contact form pretends to work
JavaScript setTimeout fakes a "Thanks! Your message was sent" message. There is no fetch, no endpoint, no email handler. Every prospect message gets a fake confirmation and you never receive it. This is currently the most damaging single bug on the site.
You might have prospects who think you ignored them. Worth fixing before the new site ships, even as a stopgap.
9. Soft commercial nudge (not for the orb hub, for Phase 2)
If the new site proves out (citation rate above 20% by day 60, real organic traffic by day 90), there's an option to layer in proper editorial backlink building via Julian Goldie at $600 to $800/m white-labelled. Background only. Your call. No pressure, no pitch. I'm mentioning it here, not on the main orb, because it's not relevant until Phase 3.