AI Agents Struggle to Access B2B Software Pricing Data, Report Says

Image credit: Search Engine Journal

AI agents face significant challenges extracting B2B software pricing and features from company websites, frequently resorting to potentially outdated third-party data, a new report by Siteline revealed Tuesday.

The report highlighted that website errors and obscured information often impede AI agents like Anthropic‘s Claude from accurately interpreting pricing structures, creating a reliance on external sources for critical business intelligence.

Siteline researchers tested a Claude agent on 100 leading B2B software products, conducting 534 attempts to locate monthly prices and feature details.

Approximately 30 percent of the agent’s runs encountered at least one error, according to the report. A quarter of these errors involved access denials due to bot blocking or unreadable web pages.

These errors created a substantial content gap, with error-prone runs sourcing 58 percent of their information from third-party sites such as G2 and Vendr. In contrast, error-free runs pulled only 12 percent of content from external sources.

David Kaufman, CEO of Siteline, said that while many companies aspire to be AI-ready, their websites often present significant hurdles to AI agents attempting to gather pricing data.

The study found that 65 percent of software plans displayed readable prices directly on company websites. However, 14 percent of plans provided no pricing information, instead routing users to a sales contact, which the report described as a “dead end” for AI agents.

The report suggested that companies could improve AI accessibility by rendering pricing and features server-side and highlighting key details early on. AI agents typically process only the first 15,000 to 20,000 tokens of a web page.

Companies like Vercel, Zendesk, Coda, Braze, FullStory, Databricks, and Cloudflare were among those whose pricing information was analyzed during the research.

The findings underscore a growing concern for businesses and AI developers as AI agents become increasingly integral to market research and competitive analysis.

Source: Search Engine Journal