LIVE · re-measured every morning · Day –last measured –

– days ago, ChatGPT had never heard of this company.

0%→–%
Share of real buyer questions where ChatGPT names them unprompted
June 10: 0%. Not one unprompted mention.
Across all five AI engines, weighted by real-world usage: 2% → –%.
Measured against – real buyer questions across – DFW cities, asked fresh every day since June 10.
LAST MEASURED: –

Every number on this page is re-measured daily and published as-is — dips included. Scroll down and audit us.

See the data ↓
The climb · June 10 → today

Not one lucky screenshot. – scored answers a day, every day, for – days.

all five engines, audience-weighted · raw daily dots · 7-day average · axis runs 0–100% on purpose
Check the tape

Don’t take the chart’s word for it. Ask any AI engine yourself, right now.

Benchmark question · copy & paste
Is Varsity Zone HVAC in Frisco, TX a good company?

Why this question and not “who should I call for AC repair?” — because buyer-intent answers are probabilistic: the same question can name different companies on different runs. That’s exactly what the daily percentages above measure. This brand question is the one check that’s stable enough to try live.

The method
–
real buyer questions
×5
AI engines — ChatGPT, Gemini, Claude, Grok, Perplexity
–
DFW cities

= – scored answers. Every morning. Automatically.

The question set was standardized on June 29 and published in the daily data. Questions are asked fresh each morning through each engine’s public interface, with no conversation history. An answer counts as “named” only if it names the client unprompted.

“Who should I call for AC repair in Frisco?”
“Best HVAC company near Plano?”
“My AC stopped cooling — who’s reliable nearby?”

⇩ Download the daily numbers (CSV) — dips included
The client
Varsity Zone HVAC — Frisco, TX
5.0★ across 49 Google reviews
TX LICENSE TACLB00028792C · 10-year parts-and-labor warranty
View their public Google listing →

Varsity Zone was already a good company — licensed, warrantied, five-starred. AI engines just didn’t know it existed. That gap between being good and being findable is the only thing this page tracks.

Straight answers

Why isn’t the number 100%?

Because nobody honest can promise 100% — AI answers vary by design. We publish the real number instead of a rounded-up one.

Did you cherry-pick the questions?

The questions are plain buyer questions across 10 DFW cities — the set was standardized on June 29 and hasn’t changed since. Download the CSV — every day is in it, including the down days.

One good day is luck. Why trust a line?

Agreed. That’s why the chart shows every daily measurement as a dot, with a 7-day average on top, on an axis that runs from 0 to 100. The dips are in the data because the dips happened.

So what did you actually do?

We make the client visible in the sources AI engines actually read when they answer buyer questions. The specific playbook is the paid part — it’s the product. The measurement is the free part: the questions are ordinary, the scoring rule is public, and you can re-run the check yourself in the box above.

Your market · run your own numbers

What’s it worth when AI hands buyers your name?

A five-step funnel, top to bottom — each step is yours to set, and the money updates as you go.

Your trade
Presets are our estimates — everything below stays yours to change.
Market size
Tap one, then fine-tune in step 1.
Step 1 · your marketHomes that hire someone like you–
jobs a year
Households in your service area50,000
A DFW suburb runs 20–80k; a metro slice 100–300k.
Homes that hire an HVAC company each yearOur data10 in 100
Average job value$850
Blend repairs and replacements. HVAC replacement alone runs $5,500–$12,500.
Step 2 · the shiftThey ask AI who to call–
buyers a year
Buyers who ask AI first20 in 100
Surveys disagree because they count different things. Salesforce ’25: 7% of shoppers start with AI. BrightLocal Feb ’26: 45% used AI for local picks last year — 6% the year before. Our 20% is buyers AI actually steers. Drag it — or tap a label on the track.
Step 3 · the gateThe AI names you–
answers with your name
Share of AI answers naming you2 in 100
Invisible companies sit at 0–2%.
Step 4 · the phoneOnly some reach out to you–
contact you
1 in 4of named buyers reach out — our deliberately low assumption
Nobody has measured this. AI answers usually name 3–5 companies; a buyer who calls two gives you ~1-in-2 odds, and not everyone acts. (BrightLocal Mar ’26: 88% double-check before acting.) Set your own.
Step 5 · the closeYou win the job–
booked jobs a year
40%of contacts become paying jobs — you know yours, override ours
Presets anchor to ServiceTitan’s Oct 2022 call-booking report (3,000+ shops): HVAC books 38% of inbound calls, plumbing 43%, garage door 31%. Roofing runs ~27% industry-wide; painting has no published rate — 35% is our assumption. Booking rates include junk calls, so presets under-count real prospects — on purpose.
The bottom line · your market
Ceiling · revenue in play–
if you won every single named answer — exposure, not bookings; AI answers usually name more than one company
Total · booked-work estimate–
our deliberately low assumptions × your numbers — steps 4 and 5 are yours to set
Each +1 point of AI visibility ≈ – booked job ≈ –/yr in this market.
Invisible companies start at 0–2 points. Our client measures 20, live. That gap ≈ –/yr in booked work at your settings — measured for them, not a promise for you.
The link opens this page with your exact numbers pre-set — send it to a partner or your CFO.
BOOKED ≈ –/yr+1 PT ≈ –

Estimates from your inputs — sizing the channel, not promising sales: being named is visibility in the buyer’s answer, and the booked line applies our deliberately low assumptions, which are yours to change. Drag anything you disagree with.

Want to know your company’s number?

We measure before we take a dollar. Send your business name and we’ll run the same benchmark on you — and send you the raw scores. If your number is already good, we’ll tell you that too.

No call required. You get the numbers either way.
All percentages are the share of daily benchmark questions in which an engine’s answer names the client unprompted. Questions are asked fresh each morning with no conversation history. “Audience-weighted” combines engines in proportion to estimated real-world usage. AI answers are probabilistic — single days vary; read the trend. Results shown are for one client in one market and are not a guarantee of outcomes. Data last updated –.