The starting point: technically clean, but anonymous

GraftPal is an AI-powered CRM for small businesses that replaces a stack of separate tools for leads, invoices and proposals. Its site was the cleanest technical base of the five brands in these success stories. The problem was not access. It was identity: AI engines had little structured information telling them what GraftPal was, and in the category answers buyers saw, bigger CRM brands filled the space.

What the AI Discoverability audit found

WriteWorks audited a sample of graftpal.com. The approximate readiness score was already excellent.

  • Readable by every crawler. Pages were server-rendered and fast, robots.txt allowed everything, and no AI user agent was refused.
  • No entity markup where it mattered most. The homepage and its market versions had no JSON-LD at all, raw or rendered.
  • Freshness signals missing. 23% of the audited pages, mostly solution and integration pages, carried FAQPage markup with no dateModified.
  • One hub needed JavaScript. The resources hub showed only 14% of its main content in raw HTML and had no ItemList.

What changed after the fixes

  • Resolved: entity markup. Every market homepage now carries complete Organization markup (name, URL, logo and sameAs) in its server HTML: 0% → 100%.
  • Resolved: freshness dates. FAQ pages with a dateModified went from 0% to 100%, and all audited articles and FAQs from 20% to 100%.
  • Resolved: the resources hub. It now sends 94% of its content in the raw HTML, up from 14%, with CollectionPage and ItemList markup.
  • Resolved: expected schema. Key pages carrying the schema expected for their type rose from 92% to 100%.
  • Still open. The status pages still show 36% of their content without JavaScript, unchanged; they were not part of this round.

Crawlers also get pages faster: the median server response is now 6.7 times quicker, and pages answering in under 0.8 seconds rose from 96% to 100%.

How these numbers were measured

Before = the first WriteWorks AI Discoverability audit, after = the re-audit once the fixes were deployed, on the same pages with the same rules. Percentages are shares of the audited pages, not of every URL on the site. A fix is marked Resolved when the re-audit shows it fixed, Improved when it is partly fixed, and Open when nothing changed. Speed comparisons come from one vantage point and are indicative. AI platform coverage is the share of the AI platforms measured that cited the site, at the baseline and six months on: a timeline, not a claim about cause.