Prose's new VP, Brand is asked to make sure Prose is "accurately and favorably represented" in ChatGPT, Perplexity, Gemini and AI Overviews. But in the HTML prose.com sends, all 23 product pages, the reviews, about, FAQ and gift pages carry the homepage title, the homepage description and a canonical tag pointing to the homepage, with no text except "requires JavaScript". Google says not to let JavaScript change a canonical away from the HTML one, and the AI crawlers behind ChatGPT and Claude do not run JavaScript, so to them Custom Shampoo, Custom Serum and the 600,000-review page are all copies of the homepage.
First action
Run curl on /products/custom-shampoo and share the output with the SEO agency and engineering.
How to measure it
Product URLs indexed with their own canonical in Search Console
- Now
- –
- Target
- All 23
- By
- 60 days
Steps
Confirm the gap
Curl the 41 sitemap pages and inspect five in Search Console to see which canonical Google chose.
SEO/blog agency 1 day
Write the copy
A title and description for each product and content page.
VP, Brand Marketing 3 days
Render the head
Server-side title, description, canonical and Open Graph tags in the page template.
Senior Full-stack Engineer 1 week
Render the body
Product name, benefits, key ingredients, price from and review count in HTML.
Senior Full-stack Engineer 1 week
Add product data
Product structured data on each product page.
Senior Full-stack Engineer 2 days
Verify
Plain curl, Search Console inspection and a test question in ChatGPT and Perplexity.
SEO/blog agency 2 days
Track
Indexed product URLs, organic product sessions and AI citations monthly.
Analytics Lead monthly
Why it works
Server-render the head of every product and content page: a unique title ("Custom Shampoo, Made for Your Hair | Prose"), a description, a self-referencing canonical and Open Graph tags. Then server-render the core body text (what the product does, key ingredients, who it is for, the Prose Promise, review count) so readers that do not run JavaScript see the page. The React app can still take over in the browser. Check the result with a plain curl and in Search Console's URL inspection.
Search engines and answer engines can only recommend a page they can read and tell apart from the homepage. It is a small change to the page template that reaches 38 pages at once, and it is the base layer the VP's AI discoverability and SEO agency work depends on.
Ready to use
ChecklistPer-page HTML checklist
[ ] Unique <title> per page [ ] Unique meta description [ ] <link rel="canonical"> to the page itself [ ] og:title, og:description, og:image [ ] Product name, benefits and key ingredients in HTML [ ] Review count and Prose Promise in HTML [ ] Product structured data [ ] curl output shows all of the above
Risks and how to handle them
Duplicate head tags when React hydrates
Have the client keep the server-rendered tags rather than replace them.
Personalized prices differ per visitor
Render a "from" price and the public offer only.
Evidence
- HiringVP, Brand Marketing (posted September 21, 2026): owns earned media, community, creator advocacy, campaigns and social impact, manages "a team of four plus external agencies", oversees "the PR and SEO/blog agency" for "LLM-optimized earned presence", and owns AI discoverability: "Ensure Prose is accurately and favorably represented where customers increasingly discover beauty — AI assistants and answer engines (ChatGPT, Perplexity, Gemini, Google AI Overviews)", treating "AI surfaces as a brand channel to monitor and shape". The role also measures brand health (awareness, consideration, share of voice, sentiment, MMM) and will hire "the open Sr. Manager, Brand Advocacy".Source
- Your websiteRaw HTML of the sitemap pages (checked twice on October 22, 2026): 38 of the 41 non-ingredient URLs, including all 23 /products/ pages, the 3 /pages/ sets, /reviews, /about, /faq, /the-salon, /gift and the consultation pages, return the homepage title "Truly Custom Haircare and Skincare | Prose", the homepage meta description, a canonical tag of href="/" (the homepage) and no body text beyond "Prose.com requires JavaScript in order to run properly." Only the homepage, /ingredients and /blog/ carry their own canonical.Source
- Google Search CentralGoogle's JavaScript SEO guide: "you shouldn't use JavaScript to change the canonical URL to something else than the URL you specified as the canonical URL in the original HTML"; "The best way to set the canonical URL is to use HTML".Source
- Vercel researchVercel's study of AI crawler traffic: "none of the major AI crawlers currently render JavaScript", including the crawlers behind ChatGPT and Claude; Gemini uses Googlebot's infrastructure and does render it.Source
- HiringSenior Full-stack Engineer (posted August 14, 2026): Python/Django services and REST APIs, "React applications on the frontend", a "deliberately compact, senior engineering team" whose scope runs "from the consultation and e-commerce journey" to manufacturing and "touches everything from checkout conversion to factory throughput"; the platform is "evolving toward a modular architecture".Source