SuppScore · Prepared by Whirlwind
Three facts about where this market and this build actually stand today. Not a recommendation yet, just what is, before we get into what to do about it.
80 to 90 percent of head supplement searches already trigger an AI-generated answer before any website shows (NIH 905 of 1,000 keywords, Examine 894 of 1,000, ConsumerLab 897 of 1,000, across the sites we tore down). Most of the sites that should own that answer are locking themselves out of it. Drugs.com blocks roughly twenty AI crawlers outright. Open Food Facts blocks GPTBot, ClaudeBot and PerplexityBot by name. Examine and the NIH get turned away by their own bot protection without meaning to. Of every site studied, exactly one welcomes AI crawlers on purpose.
The build we audited serves every page through a client-side gate. Real content only appears once a browser sets a flag in local storage after the page has loaded, which no AI crawler can do. As reviewed, no assistant can currently read a single page on the site, regardless of how good the content on it is. This was true in the version we audited. It takes one command, run against a live page, to confirm whether it still is.
The ranked "best of" pages, your own idea, meant to catch ad traffic, need a live mapping of which scored products belong to which goal, sleep, energy, gut health and so on, to generate at all. That field doesn't exist in the data model yet. This isn't new scope and it isn't a template question. It's a genuine gap, found while specifying the page properly, and it would have stopped this exact idea from working whether or not we'd been brought in.
Real UK monthly search volume, already mapped in the keyword universe. Not projections.