Every firm in this category publishes numbers. Almost none of them publish how the number was arrived at, or when, and a fair share are unverifiable by construction. So here is ours, including the times we measured our own work and did not like the answer. If something below is out of date, that is on us and we would like to know.
0
characters
Method
Fetched ctrltechhq.com without executing JavaScript, stripped scripts and markup, counted the visible characters. The site is client-rendered, so the served HTML carried the schema and the meta description and not one sentence of copy. Google renders JavaScript and saw a normal site. ChatGPT, Claude and Perplexity largely do not.
Measured 2026-08-14
3
files missing
Method
All three returned HTTP 200. All three were serving the single-page-app shell, because the host rewrites unknown paths to index.html and none of the files had ever been created. A crawler asking for robots.txt got a web page.
Measured 2026-08-14
1 of 5
buyer prompts
Method
Ran five buyer-intent prompts a Dayton owner might use, covering web design and AI automation. We surfaced on the branded query only. Roughly seventeen competitors appeared across the other four; we were in none of them.
Measured 2026-08-14
38
sites audited
Method
Ran our presence audit against every company with a website in an 83-row Dayton scan. 41 had a URL, 38 responded. 25 of 38 carried no structured data, 29 had no llms.txt, 10 had no real sitemap, and 2 actively disallow AI crawlers in robots.txt, most likely without knowing it. Median readable text was 5,084 characters.
Measured 2026-08-14
36 of 38
beat us
Method
Same run as 04, with our own domain in the sample. At zero readable characters we placed below every site that returned any text at all. We fixed ours the same day, which is the only reason this entry is not still true.
Measured 2026-08-14
$419
per month
Method
A market looked wide open: weak competition, nobody defending the search results. We checked demand before building anything and found roughly 40 searches a month, worth about $419 a month at market click rates. A comparable local category in the same city was worth $21,648. An undefended market is usually an empty one.
Measured 2026-07-14
$1.49
per metro
Method
One prospecting pass over a metro, joining public hiring signals against public ad-transparency records. That pass returned 83 companies in Dayton with a dated reason to call each one. Cost is the actual metered API spend, not an estimate.
Measured 2026-08-13
186 → 364
pages to clicks
Method
A competitor in our category runs 186 pages, a hand-written llms.txt, a section built specifically for AI crawlers, and an MCP endpoint. Their ranking keyword set is 33 terms and about 364 monthly clicks, every one a question, none with commercial intent, and a large share of it students asking whether their university can detect ChatGPT. We were about to build a 25-page version of the same idea.
Measured 2026-08-14
You will not see "up to 80% cost reduction" here. Competitors publish figures like that with no method, no date, and no source. If we cannot show how a number was arrived at, it does not go on the site.
Nobody controls these systems and their output is not deterministic. We can tell you what an engine returned on a given day and we can fix what is mechanically broken. Anyone promising you a ranking in ChatGPT is selling something they cannot deliver.
We run paid campaigns and back-office work for clients and we have their performance data in front of us. It is theirs. It appears here only if they say so, in writing, by name.
Where we have no result yet, the section stays empty. An empty section costs us a sale. A fabricated one costs more.
If a number here is out of date, that is on us. Tell us and we will fix it or pull it.