30 June 2026Agent Name
Agency Name · 5 areas measured · ChatGPT, Gemini, Claude and Perplexity
About this report
What this is. A live measurement of how ChatGPT, Gemini, Claude and Perplexity answer when buyers and sellers ask about 5 areas you work: 40 captured answers, every entity mention verified against the answer text. No analytics product records this behaviour; the only way to see it is to ask the platforms and capture what they say, which is what this system does.
What it covers. Part 1: where you stand, name by name and area by area. Part 2: why, through the three things AI weighs (coverage, messaging, social proof). Part 3: the plan, with your numbers filled in. The appendix holds the full underlying detail.
How to read it. Each section leads with its finding in large type, then the data behind it. If you only skim, read the findings and Part 3.
A baseline, not a verdict forever. AI answers move as you act, as rivals act, and as the models themselves are updated. This is your measured starting position; movement against it is the number that matters.
27 / 100
Getting seen
One number for how often AI names you, weighted by how high on the list you land.
When buyers ask AI for an agent in your areas, you show up in 30% of answers. When you are named, you sit around #3.2 on the list.
Right now AI names Competitor 1 more than you. This report shows exactly where, and how to close it.
AI doesn't judge, it repeats whoever is best covered, best described and best reviewed across the web. Three levers, and here is where you stand on each.
36 / 100
Sample coverage finding.
44 / 100
Sample messaging finding.
31 / 100
Sample social-proof finding.
The truth, name by name
Who AI actually hands buyers in your areas: the full leaderboard, the fight in each of your patches, and how the four AIs each treat you.
Who AI hands buyers instead of you
This is the list AI reads out when someone asks who to use in your areas, in order. You sit at #5: the names above you are the ones buyers hear first. Your row is highlighted.
Competitor 1 is the name AI hands out in your market. You are #5.
Area by area: who owns each of your patches
Your visibility is not one number, it is a fight per area. For each area you paid to measure: AI's current pick there, and where you sit.
You are winning 0 of your 5 areas. The other 5 belong to someone else.
Red rows are areas where you may be missing appraisal opportunities: when a seller there asks AI, the name in the right-hand column is who gets put forward. The plan below has a play per weak area.
The four AIs don't agree on you
Each one reads different sources and names different people. Here is exactly where you stand on each, side by side. The low ones are where you have the most ground to make up, and each engine's source list shows how.
Claude barely knows you: named in 3 of 10 answers.
Same you, four verdicts. The reason is the sources column: each AI trusts different sites, so the fix is to be strong on the ones your weak engines read. The plan at the end covers exactly that.
In their own words
Not our summary. This is what the AIs literally answered when we asked about your market, copied straight out of the run.
Sample chatgpt answer. This placeholder demonstrates where the model's real response appears in a purchased report.
Three things decide this
The same three diagnostic chapters included in every purchased report, shown here with fictional placeholder names, sources and measurements.
Coverage
BuildingAI cannot name who it cannot find. Before it answers, it reads a handful of sites about your market; being present on them is what makes you nameable. Coverage is simply: of the sites AI actually reads, how many carry you?
Sample coverage finding.
The sites deciding this without you
Every row is a source AI read while answering WITHOUT you. This is your coverage to-do list, worst first: get present and reviewed on these and you enter the answers they feed.
citation-1.example fed 13 answers that left you out.
Counted per answer, across every question we asked. A large red number with a small grey one means that site talks about your market constantly, and barely about you. How to work the list: a directory or review site, get listed and reviewed there; a rival's own website, you can't join it, so match it with your own site and the rest of this list instead.
What each engine reads, side by side
The four AIs trust different sites, so coverage is a per-engine job. For each engine: its top sources for your market. Be visible on each engine's list, especially your weakest engine's.
- 1citation-1.example8
- 2citation-2.example5
- 3citation-3.example3
- 1citation-1.example8
- 2citation-2.example5
- 3citation-3.example3
- 1citation-1.example8
- 2citation-2.example5
- 3citation-3.example3
- 1citation-1.example8
- 2citation-2.example5
- 3citation-3.example3
Where the recommendation actually comes from
AI does not decide who is best. It repeats what these sites say. Get onto them and you become the answer; stay off and you stay invisible, no matter how good you are in real life.
Citation 1
citation-1.example
Citation 2
citation-2.example
Citation 3
citation-3.example
Citation 4
citation-1.example
Citation 5
citation-2.example
Citation 6
citation-3.example
AI never cites your own website. That is a gap worth closing: your plan covers it.
The searches AI ran about your market, that you never saw
Before it answered, AI quietly searched the web. These are the patterns across every retrieval query it ran, captured at the moment of response: the exact machinery deciding who gets named.
AI ran 39 searches to size up your market before it answered a single question.
Perplexity does not expose its searches; counts cover the other three. Not every search fits a recurring pattern, so the rows sum to less than the total. Read the table as the levers in action: reviews, superlatives, directories and recency are literally what AI goes looking for.
Messaging
BuildingAI repeats what the web says about you, in the web's words. If your positioning is stated clearly and consistently everywhere AI looks, it gets repeated back to buyers; if it is vague or scattered, AI describes you vaguely or not at all.
Sample messaging finding.
What AI says about you when you're not in the room
Ask AI about you by name and this is what it tells a buyer, in its words, not yours: whether it even knows who you are, the tone it takes, and the exact descriptions it repeats.
AI knows who you are. It tells your story your way only 2 times out of 4.
“Sample evidence quote 1”
“Sample evidence quote 2”
The questions buyers ask, and who got the answer
These are the 6 questions buyers ask AI before they ever call an agent. 3 came back with you in the answer. In the other 3, potential clients were recommended other agents.
3 buyer questions were answered without you in the room.
Your private action plan
The paid report turns the findings above into this same week, month and quarter structure. Exact actions, targets and competitive strategy stay private.
This week
Your highest-priority immediate action
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
The first measured gap to close
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
A quick win tied to your run
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
This month
The questions and areas to target next
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
The platform-specific opportunity to pursue
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
The evidence required to close the gap
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
This quarter
Your compounding priority and target
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
The longer-term competitive opportunity
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
What to re-measure in the next report
The purchased report fills this row with the exact action, named target and supporting evidence from your fresh run.
The foundations checklist
The purchased report includes the complete execution checklist beneath the timed plan.
Guidance from this report, in one place
Every recommendation is collected here in the real report so it can be worked from one section.
The public sample preserves the complete report structure while withholding the exact actions, competitive targets and implementation guidance customers pay to receive.
Where this leaves you
- This conclusion states exactly where you stand overall, how many of your measured areas you hold, and who currently sits ahead of you.
- It identifies your largest measured opportunity and explains why that is the first place your private action plan focuses.
- It closes with the order of execution and the numbers to track when the report is run again.
Preview only. Your purchased report fills this conclusion with findings from your fresh run; no customer conclusion or strategy is published in the sample.
Nothing is a black-box summary
The appendix uses this same layout in the purchased report. The questions, descriptors and quotes below are fictional placeholders.
Appendix: every direct answer about you, in detail
The full record behind Part 2's messaging chapter: each platform's answer to each direct question, how well it knew you, and the words it used. Kept in full so nothing in this report is a summary you cannot check.
Described as: Sample descriptor 1
“Sample evidence quote 1”
Described as: Sample descriptor 2
“Sample evidence quote 2”
Described as: Sample descriptor 3
“Sample evidence quote 3”
Described as: Sample descriptor 4
“Sample evidence quote 4”
See the complete structure.
Buy the findings and the playbook.
The sample shows what every section delivers. Your purchase fills those sections with a fresh measurement of your name, your areas, your competitors and the exact work required to move your result.
Get your private report
Social proof
BuildingWhen AI weighs who to recommend, review volume and rating are its safety check: heavily-reviewed names are the low-risk answer. It is not judging your work, it is counting the proof.
Sample social-proof finding.