EffiWise

Updated 2026-09-15 · n = 151

43% of independent insurance agencies are never named by AI.

Every agency that ran the free EffiWise audit May through September 2026 is in this data: 151 scored reports across 118 agency websites, mostly in the Southeast and Southern Plains. No agency is named. Every figure carries its denominator. Where our own tooling could not measure something, it is counted as unmeasured, never as a zero.

01

AI Shortlist Battle

Seven buyer-style prompts per agency, run through AI search for the agency's own market. We record which agencies get named.

43%
of agencies were never named by AI, in any prompt
65 of 151 audits
30%
of AI answers named the agency being audited
309 of 1024 prompt answers
14%
were never named while a competitor was named every single time
21 of 151 audits
14
median AI Shortlist score out of 100. The worst of the three pillars.
Local Findability 40, Authority Signals 57
02

Authority signals on the agency website

Read directly from the HTML of up to eight pages per site, the same way an AI crawler reads it.

6%
show a license number anywhere on the site
9 of 151
62%
name at least one human agent
93 of 151
68%
have any schema markup at all
102 of 151
35%
tell the machine they are an insurance agency, in code
53 of 151 use the InsuranceAgency type
50%
list even one carrier by name in readable text
76 of 151. Logos alone are invisible to AI.
60
median mobile PageSpeed score out of 100
A quarter of sites scored under 47
03

Can AI read the site at all?

Measured on the 97 audits that carry the machine-readability check.

13%
of sites needed JavaScript to show any content. Our crawler, like AI crawlers, saw an empty page.
13 of 97
31 vs 50
median score, unreadable sites vs readable sites
n = 11 and 79
34%
of sites were custom or unidentifiable builds. Every unreadable site was one of them.
33 of 97
04

Directory presence

Where our checks found the agency. Absence in our check is not proof of absence, but AI engines run the same kind of check.

41%
had a Facebook page we could find
62 of 151
6%
had a Yelp listing we could find. Yelp reviews now feed ChatGPT directly.
9 of 151
1%
had a Better Business Bureau profile we could find
2 of 151

How to read the scores

The median overall score is 42 out of 100 across all 151 audits. The first 60 audits (May 2026) ran on an older scoring regime that treated a missing Google Business Profile lookup as a deficiency; from June onward the median is 49. Scores are computed in code from measured facts. AI prose never moves a number.

Where does your agency land?

The audit is free, takes about three minutes, and shows exactly which agencies AI named instead of you.

Run my free audit
Method

Each audit fetches up to eight pages with a Chrome user agent and parses JSON-LD, carrier names (text and image alt), license patterns, and NAP consistency in code. Google Business Profile data comes from the Places API where available. PageSpeed comes from Google's PageSpeed Insights API. The AI Shortlist Battle runs seven region-aware buyer prompts through AI search with live grounding and records the named agencies. Counts reflect what our checks surfaced; absence in a check is reported as “not found” and never as proof the agency is not listed. Cite as: EffiWise Agency Audit corpus, 2026-09-15.