Share

Esri

Esri's top 10 neighbors span journalists, science magazines, NPR personalities, and a technology peer — a mix that resists any single label and reflects the flat shape of this audience.

Audience similarity here measures how closely two entities' audiences resemble each other in composition. The top 10 scores run from 0.87 down to 0.85, a narrow band with no single dominant neighbor. WIRED Science leads at 0.87, followed closely by Peter Sagal (0.86), Ken White (0.86), Kai Ryssdal (0.86), and Scott Simon (0.86). The subcategory breakdown tells the story: four of the top 10 are Journalists (Kai Ryssdal, Scott Simon, Mary Louise Kelly, and All Things Considered falls under Podcasts and Radio), two are Magazines (WIRED Science and Nature does not appear in the top 10 — correcting: the Magazines subcategory is represented by WIRED Science at 0.87), one is a TV Personality (Peter Sagal), one is a Professional (Ken White), one is a Government Official (CDC Director), one is Podcasts and Radio (All Things Considered), and one is a fellow Technology brand (Asana at 0.85). The cross-kind pattern is pronounced: nine of the ten neighbors are outside the Technology subcategory, with Asana the lone peer. The dominant cluster is public-radio journalists and science-adjacent media — a profile more consistent with a civically engaged, data-literate professional audience than with a software peer set.

The flat shape and the dominance of journalists and public-media personalities over technology neighbors suggests Esri's audience is defined less by software category affinity and more by a broader orientation toward data, science, and public affairs.

Playground →Docs

Nearest neighbors by audience shape

Cosine similarity over Persona Live audience composition.
"Entities whose overall audience profile most closely matches this one — people who follow one tend to follow the other."

Raw data

Attention graph /similar sample results  (free):
Full Attention graph /similarity /affinity /demographics available with  API key: