PROGRAM

After a decade of consistent traffic growth driven by search engine optimization (SEO), traffic to the UCSF Profiles expert finder system and others in our network has shown declines for the first time ever, back toward pre-pandemic levels. While this may reflect reduced interest in UCSF COVID-19 researchers, we're curious about another possibility: are search engines and AI bots increasingly delivering our content without sending users our way?

Google, Bing, and other search engines often display knowledge panels or AI-generated summaries directly at the top of the results page, especially for higher-profile researchers. The links to our site are still there, but hidden under the search engine's summary. For years we worked to be the #1 search result, but is the resulting traffic waning in an increasingly no-click world? Are users still benefiting from our work, just invisibly? And how should expert finder systems measure success in this new environment?

We're also exploring another angle: our own performance. Poorly behaved bots, some associated with newer AI companies, may be placing increasing load on our systems. Sites like Wikipedia have reported substantial strain from new bots. Expert finder systems, featuring rich structured public data, may be similarly vulnerable, affecting both performance and the quality of our analytics.

In this session, we'll share what we've seen, and make space for candid conversations:

• Are popular profiles losing traffic while lesser-known ones hold steady?
• Is our content particularly ripe bot fodder because it is slow moving and reflected in other sources like PubMed and NIH RePORTER?
• How much of our traffic is real users vs. bots?
• How can we measure impact beyond web clicks? If people see our content, but not on our site, is that a win or a loss?
• Should the expert finder community rethink exposure strategies in the AI era?

We don't have the answers, but together, we can start making sense of this new world.