How the Response Index is measured
This page exists so anyone can check our work — or disagree with it precisely. Every signal below is publicly observable, which means you can reproduce any score in this index yourself.
The one rule that shapes everything
We never contact the businesses we measure. No calls, no texts, no emails, no form submissions, no mystery shopping. Measuring lead response time by submitting thousands of fake inquiries would produce a cleaner number, but it wastes the time of real staff, and at scale it runs into genuine legal exposure. So we built the index entirely out of what a business already shows the public.
The tradeoff is honest: this index measures responsiveness posture — how reachable a business is, and how it behaves when a customer speaks to it in public — rather than a stopwatch on a single lead.
What we collect
- Reply to reviews — Share of recent Google reviews that got any owner response.
- Median reply time — How long the typical owner takes to respond to a review.
- Tap-to-call on mobile — A phone number you can actually tap from a phone.
- Online booking — Any way to book without picking up the phone.
- Contact form — A form that captures a lead outside business hours.
- Homepage response time — How long the homepage takes to respond. Measured directly, not a Lighthouse score.
- Open Saturday — Listed weekend hours on Google.
- Open past 6pm — Reachable after the average customer finishes work.
Listings come from Google Maps. Everything else comes from a single polite request to the business’s homepage, identifying itself as TitanOpsResponseIndexBot. Response time is measured on that same request — it is server responsiveness, not a Lighthouse performance audit, and we do not describe it as one.
How the score is weighted
The Response Score runs 0–100. The weights encode a point of view: being reachable and replying matter more than having a fast website. A beautiful site nobody answers the phone for is the exact failure this index exists to expose.
| Signal | Points |
|---|---|
| Replies to reviews at all | 25 |
| Replies quickly | 10 |
| Tap-to-call on mobile | 12 |
| Online booking available | 12 |
| Contact form available | 6 |
| Mobile site speed | 20 |
| Open Saturday | 8 |
| Open past 6pm | 7 |
| Total | 100 |
Reply speed decays linearly: replying within a day earns full marks, and a median reply lag of 30 days or more earns none. Homepage response time works the same way — under 0.8 seconds earns full marks, 5 seconds or worse earns none. Grades are A (85+), B (70–84), C (55–69), D (40–54), F (below 40).
Businesses with no website are scored honestly rather than excluded — they simply cannot earn the web-dependent points. That absence is a finding, not a gap in our data.
What we exclude, and why
Two kinds of listing are dropped from every published figure. In both cases the honest answer is “we could not measure this”, and scoring them as failures would inflate how bad each market looks.
- Websites that block our crawler. Roughly a quarter of sites sit behind a firewall that rejects any request identifying itself as automated. We do not disguise our crawler to get around that. Those businesses are marked unmeasured rather than recorded as having no booking, no contact form and no tap-to-call.
- Listings with no reviews. Review response is our core signal, and it is undefined when there are no reviews. These are overwhelmingly ghost or unclaimed listings rather than operating practices.
Every market page publishes its coverage rate so you can see how much was excluded. This is a real limitation with a known direction: firewall-protected sites tend to belong to better-resourced businesses, so excluding them may bias scores slightly downward. We would rather state that plainly than hide it.
Sample size and publication
A city page is only published once we have measured at least 25 businesses in that market and vertical. A vertical-wide page requires at least 100. This rule is enforced by a database constraint, not by good intentions — it is not possible for this site to publish a market claim built on a thin sample.
We report medians rather than means for scores, because a couple of very good or very bad outliers should not describe a whole market.
Known limitations
It measures posture, not a stopwatch
We do not call, text, email, or submit forms to the businesses in this index. That means we cannot report "this practice answered in 4 minutes." We report how reachable and how responsive a business demonstrably is in public. Treat the score as a strong proxy, not a direct measurement of lead response time.
Review replies are a proxy, and proxies have error
A business could ignore Google reviews while answering every call perfectly. We think that is rare — the same staffing and attention gap usually shows up in both places — but it is a real source of error, and we would rather state it than bury it.
Only the most recent reviews are sampled
We read up to 15 recent reviews per business. A practice that replied diligently three years ago and stopped will score on current behaviour, which is what we want, but the sample is small for businesses with few reviews.
Homepage-only site checks
Booking links, tap-to-call, and contact forms are detected on the homepage. A business that hides its booking link three clicks deep scores as if it has none. We consider that fair — so does the customer — but it is a detection limit, not a judgment.
Google hours can be stale
Saturday and after-hours availability come from Google Business Profile listings, which owners do not always keep current.
Chains and multi-location practices
Each listing is scored independently. A practice with four locations appears four times, which can pull a small market toward that group’s habits.
Scope
The index currently covers 14 verticals across up to 70 US markets. Coverage expands as runs complete; a market with no published page simply has not been measured to threshold yet.
Corrections
If you own a business in this index and believe a measurement is wrong, email jamisonpressley2@gmail.com. We will re-measure and correct it. Individual businesses are never named on public pages — the index reports market-level aggregates only.
Using this data
You are welcome to cite these figures in articles, decks, and research with attribution to the TitanOps Response Index and a link back to the relevant page.