What gets shipped, when, and in what order.
The sequence is built around two things: getting traffic before summer peak (Q2/Q3 2026), and putting the foundation in place for a Q1 2027 ramp when you said you'd think about DTC. Soft moves, no expensive bets.
Your rule
"I'd rather have traffic and not need it, than need it and not have it."
So we push SEO ahead of inventory. By the time you decide whether to leave the FBA sandbox, the traffic will already be there.
Days 1 to 14: the trust-and-hygiene sprint
- Fix the contact form (or kill it). Currently every prospect message is silently dropped.
- Replace the fake address, fake stats, and Pexels "team photo" on the live site.
- Build a real robots.txt and sitemap.xml. Both return 404 right now.
- Upload disavow.txt to Google Search Console (the 92 spam domains).
- Add Organization + Product schema to the live page (30 min job, moves AI score 6 to 14).
- Add meta description, canonical, og:image (15 min job).
- Fix the Amazon brand-field "NHI" to "ChillSwift" on B07BNCJ1SJ via Brand Registry 2.0.
- Kill the dead
prelaunch.chillswift.comDNS record.
Acceptance: AI visibility score moves from 6 to ~25. Trust signals stop costing you Amazon Brand Registry audit risk. Disavow file is in front of Google.
Days 15 to 45: the new-site ship
- Ship the new Astro site (already built, preview at chillswift-preview.pages.dev).
- Publish the 4 product pages (cooling vest, dog collar, towel, packs) with full Product schema.
- Publish the 5 article drafts I already wrote (heat exhaustion, dog walks, heatstroke signs, vest mechanism, vest duration).
- Add 301 redirects from the 6 high-value dead Shopify and WordPress URLs to the new pages.
- Submit sitemap to Search Console. Force a fresh crawl.
Acceptance: 9 of the 20 target pages live. 5 of 18 striking-distance keywords starting to move. First AI citations should appear within 30 days of these going live.
Days 46 to 90: the content + AI evidence push
- Build the remaining 5 informational guides (outlined, not drafted: construction workers, motorcycle riders, MS heat intolerance, hot pavement test, are cooling vests worth it).
- Build the 3 AI evidence pages:
/cooling-vest-comparison-2026/,/cooling-vest-vs-cooling-shirt-vs-cooling-towel/,/dog-cooling-collar-vs-cooling-vest-vs-mat/. Schema-loaded, citation-bait. These are the highest-leverage AI moves. - Build the 3 cluster hubs (cooling gear for humans, cooling gear for dogs, heat safety).
- Run the 50-prompt AI citation bank monthly. Log results. Track citation rate.
- Render the 5 reel storyboards if you green-light the budget.
Acceptance: 20 of 20 pages live. 50-prompt baseline complete. AI citation rate target: 20% by day 60, 50% by day 90.
Q1 2027 roadmap (if traffic proves out)
This is the optional layer. Only triggered if Q4 2026 traffic data says it's worth the spend. No commitments now.
- White-label editorial backlinks via Julian Goldie at $600 to $800/m. Real editorial links, not PBN spam. Background only. Your call.
- DTC layer on top of FBA. Add Shopify or simple Stripe checkout to the new site. Keep Amazon as the primary fulfilment path. Test 5% of traffic.
- Email capture from informational guides. Build a simple list. Summer 2027 promo runs to that list, not just Amazon search.
- Expand the keyword footprint into adjacent niches (cooling for motorcycle, MS, pet, athlete) once core terms are ranking.
What's NOT in the plan
- Paid Amazon PPC. You haven't said you want it and the organic lift will reduce the need.
- Influencer outreach. Premature until citation rate is over 20%.
- A pivot away from FBA. You said no, and the plan respects that.
- Anything that needs more than ~5 hours per week of your time. The whole thing is designed to run mostly in the background.
The honest assessment
You have a real product, real reviews, real Brand Registry, and no SEO debt that wasn't inherited. The site is a hand-coded HTML page that doesn't reflect any of that. Fix the trust-and-hygiene layer this month, ship the new site next month, layer in the AI evidence pages the month after. By summer peak you should be visible. By Q1 2027 you should have enough data to know whether DTC is worth it.