[ Answers · AI search ]

Why does AI get my company's pricing wrong?

By Ben Tabas, Intrica Research · Updated

Answer

We found no public study that measures how often AI gets prices wrong, so the causes below come from adjacent evidence. Ahrefs, July 28, 2025, found AI-cited URLs were on average 25.7% fresher than organic Google results. Old prices on third party pages and prices that only load with JavaScript are two ways a wrong number reaches the answer.

Four sourced reasons

  • Freshness: Ahrefs analysed 16.975 million cited URLs and found cited URLs were on average 25.7% fresher than organic results. AI engines favor recently updated pages, so an undated pricing page can lose to a newer page with a different number.
  • Third party pages: in Intrica's exchange safety study, published Oct. 16, 2026, 52.1% of ChatGPT's sources were the exchange's own site. On Perplexity, Gemini and Google AI Overviews that share was 8.2% to 10.2%, and Reddit, Trustpilot and crypto media such as Coin Bureau were among the most cited sources. Those pages carry more of the answer, and their prices may be old.
  • Agents fall back to other sites: Growth Memo, July 13, 2026, tested 100 B2B products. Pricing and features had a 79% first-party answer rate, against 93% for integrations. Even where a numeric public price was shown, agents still cited a third party in 18% of runs. Of 580 third party pricing citations, 46% were directories.
  • JavaScript: Vercel, Dec. 17, 2024, found none of the major AI crawlers it studied rendered JavaScript, apart from Gemini. A price that appears only after scripts run may be blank to them.

What to change

  1. Put every price in plain HTML on the page, with the plan name and billing period.
  2. Date the pricing page with a visible last updated line and keep it current.
  3. Publish one facts page, such as /facts, that states current prices, plan limits and company details in plain text.
  4. Find third party listings and reviews that show old prices, and ask each site to correct them.
  5. Ask ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews and Google AI Mode for your price each month and log the answers.

Whose pages the engines cite

In Intrica's exchange safety study, 52.1% of ChatGPT's sources were the exchange's own site, against 8.2% to 10.2% on Perplexity, Gemini and Google AI Overviews. The study asked four AI engines if 13 crypto exchanges are safe, Aug. 11 to 21, 2026.

See the data →

See what ChatGPT and Google AI say about your company today, including the price.

Run the free Answer ShareBook a call

Sources

  1. New Study: AI Assistants Prefer to Cite "Fresher" Content (17 Million Citations Analyzed) · Ahrefs, July 28, 2025Supports: 16.975 million cited URLs were on average 25.7% fresher than organic search results.
  2. Where AI agents get stuck on your site · Growth Memo (Kevin Indig), July 13, 2026Supports: 79% first-party answer rate for pricing and features vs 93% for integrations, 18% third party citation with a public price, 46% of 580 third party pricing citations from directories. Checked against the full article on Oct. 29, 2026.
  3. The rise of the AI crawler · Vercel, Dec. 17, 2024Supports: None of the major AI crawlers studied rendered JavaScript, while Gemini uses Googlebot's infrastructure.

Related questions

Intrica gets crypto, fintech and AI companies named in AI answers. AI search optimization →