Finance
How Ashgrove got AI engines to quote its current fees
After a pricing change, AI engines kept quoting Ashgrove’s old account fee to prospects. Cited traced the errors to three stale sources, and five of six engines were quoting the current fees within 12 days.
12 days
to correct fee answers on five of six engines
In financial services, a wrong answer about fees is a compliance problem, not just a marketing one. Cited gave us the source behind every error, which meant we could actually fix it.
Nadia Haddad
Head of brand and content
The challenge
Ashgrove provides business banking for companies with fewer than 50 employees. In the spring it dropped its $15 monthly account fee and restructured its transfer pricing. Six weeks later, its support team was still hearing from prospects who had been told by an AI assistant that Ashgrove charged a monthly fee.
Compliance asked marketing to find every place the old pricing appeared. Nobody could say where the engines were getting it.
What they did
The team added 85 brand prompts to Cited covering fees, eligibility, security and integrations. After the first weekly run, Citation Sources traced 90% of the incorrect fee answers to three pages: an old pricing page still live on Ashgrove’s own site, a partner’s comparison table and an outdated review article.
Redirected the old pricing page and added a dated change log to the current one
Sent corrected pricing to the partner and the publisher, who both updated within a week
Added a plain-text fee summary at the top of the pricing page
Set Answer Diff alerts on every fee-related prompt
The result
Within 12 days, five of the six engines quoted current fees on every tracked prompt, and the sixth followed a week later. Accuracy on brand prompts rose from 62% to 97%, and fee-related support contacts from prospects fell by 70%.
Ashgrove now runs its brand prompts through Cited before and after every pricing or product change. The results go to the compliance team each month, alongside the list of sources that were corrected.