Every number you'll ever see from us was measured, not modeled.
Most marketing tools estimate. They model your traffic, guess your rank, and score you with math nobody can audit. We do something slower and better: we go look. This page is exactly how · for every signal we measure · in plain language, including what we can't know.
A rank isn't a number. It's a map with up to 49 points on it.
Google answers "plumber near me" differently on every street, because where the searcher stands is half the question. A single "you rank #4" is only true in one spot · so we don't stop there.
How we measure: every week (daily on higher plans) we run your real searches as real Google queries pinned to real coordinates across your service area · up to 49 points on a grid. We record the position at every point, which page of yours ranked, and what else was on the page: ads, and whether an AI answer box sat on top.
What we can't know: two searches minutes apart can disagree by a spot or two. That's Google, not us · so verdicts lean on multi-week trends, and single reads are presented as what they are: one real look on one real day.
We ask the assistants every day. And we keep their exact words.
"Who should I call for pest control near me?" is being asked to ChatGPT thousands of times a day · and somebody gets recommended. That's the second search engine now, and most tools either ignore it or sell you an "AI visibility score" nobody can audit.
How we measure: a panel of ~25 questions real buyers ask, put to ChatGPT, Gemini and Google's AI Mode every day. Every answer is stored word for word before anything is computed. Then we simply count: who got named, in what order, whether your site was cited.
The honesty rule: shares always come with their denominators ("named in 5 of 42 analyzed answers"), and an answer we couldn't fetch shrinks the denominator visibly · it never pads the score.
"For pest control in that area I'd suggest calling Redd Pest Solutions or Synergy² · both have strong recent reviews…"
Measured against your market, with a three-state answer that never bluffs.
Everyone knows their star rating. Almost nobody knows the number that decides the map: how fast reviews arrive compared to your competitors'. We record both · yours and the field's median and leader · every week.
The directories: we check the fifteen that actually carry weight (Google, Yelp, Facebook, BBB, Bing, Apple Maps and nine more) and answer three questions per directory: are you there, is the information right, and if not · here's the exact door to fix it, with your correct details ready to paste.
The three states: found · checked-and-absent · could-not-verify. That third state exists because pretending to know is worse than saying we couldn't check. Incumbents advertise "300+ directories"; most of those have never sent anyone a customer. Fewer, honest, weighted.
One real weekly sweep:
Google ✓ found · correct
Yelp ✓ found · wrong phone number
BBB ✗ missing · claim link attached
Apple Maps ? could not verify
Every "missing" ships with the claim door and your correct details. Every "could not verify" says so instead of guessing.
Your site graded on what moves local results. Your links counted with the receipts.
The site: every week we grade the checks that actually matter for a local business · speed from Google's own test, the titles and headings that win clicks, whether Google is allowed in at all, security, structured data. Each finding comes with a plain explanation and a fix path.
The links: backlinks are the web's votes. We compute your Lighthouse Authority score from two independent crawls of the web · and unlike the scores other tools sell, ours expands into the actual list of sites linking to you. A score you can't open into its evidence is an opinion.
Also watched: your top rivals' homepages, fingerprinted weekly · so the week a competitor starts investing in their site, you hear it from us first.
Why our authority score has its own name:
Ahrefs has DR, Moz has DA, Semrush has AS · each is that vendor's own math on that vendor's own crawl. Ours is Lighthouse Authority: computed from the open web graph plus a live link index, defined publicly, and always shown with the links behind it. Same family of number, honest provenance.
Who's winning your market's clicks · and which searches they own that you don't.
How we measure: one weekly sweep records the complete visible field · every business that appeared for your searches across your service area, with positions, reviews and strength. From it we compute share of voice (how the clicks likely divide, using public click-through curves) and the gap list: searches where a rival ranks and you don't, sorted by how winnable they look.
The line we won't cross: we publish shares of a measured field, never invented visit counts. Independent studies put the big tools' traffic estimates at roughly 50-60% median error. A share of what we actually saw is defensible; a made-up absolute is marketing.
Demand we saw, not demand we imagined:
Keyword ideas come from what Google itself displayed in your market · the related searches and question boxes on your area's real result pages · sized with honest ranges (like "1K-10K a month"), never fake-precise counts. The industry's exact-looking numbers are wrong most of the time, and everyone in it knows.
Rankings are the means. The phone ringing is the point.
How we measure: a tracking number and a tiny script on your site turn every call, form and phone-tap into one ledger, each entry stamped with what produced it: an ad click, your Google profile, a search, a referral. It's the bridge between "your visibility improved" and "your phone rang because of it" · the claim no ranking tool can make.
The honesty rule: "unknown" is a real bucket. An event we can't attribute honestly says so · no source ever gets invented to make a chart prettier.
The promises, all in one place:
- Absence is data: not showing up gets recorded and shown, never hidden.
- Every share comes with its denominator.
- Shares of measured fields, never invented visit counts.
- AI answers stored word for word · every score opens into its evidence.
- Could-not-check is always distinct from failed.
- Bad news is included. That's the product.
The method, applied to you, in sixty seconds.
The free grade runs a slice of everything above · your site, your Google profile, the directories, one real search · and emails you what ChatGPT said.