Summarize this article with AI
Send it straight to your assistant and get the short version first. The link travels with the prompt, so the model reads the real page.
Opens in a new tab with the prompt already written. Nothing is sent until you press enter.
The short answer
Two equally good builders get different outcomes because one is easy to verify and the other is not. Nothing in the retrieval process assesses whether you are good at your job.
It is not a quality judgement
This is worth saying plainly, because contractors take it personally. Nothing in the retrieval and synthesis process evaluates whether your work is good. It evaluates whether you are describable and whether anything independent backs the description.
The best framer in the valley and the worst one are equally invisible if neither is legible.
The usual gap, in practice
| The visible company | The invisible one |
|---|---|
| States trade, scope and service area plainly | Leads with a slogan |
| Google profile matches the website | Profile last updated years ago |
| Reviews contain specifics | Reviews are one word |
| Listed on licensing board and two trade directories | Own domain is the only source |
| Foundation files present | robots.txt and sitemap both 404 |
Notice that none of those rows are about how well they build.
The engine lottery is real
We have measured an engine that returned hundreds of citations across a run while never once citing the business being measured, and observed the same pattern on an unrelated business the same day. That makes it a pattern rather than an artifact.
If you only checked that one assistant, you would conclude you were invisible everywhere. You would be wrong, and you might spend money fixing the wrong thing.
What to do about it
- Measure across several engines, with repetition, and keep the records.
- Fix legibility first, because it is cheap and it gates everything.
- Then fix corroboration, which is slower and matters more.
- Re-measure at ninety days with the identical method.
- Accept that one engine may never favour you, and stop paying to chase it.
The order to work it in
Once you accept that this is legibility rather than quality, the fix list stops being mysterious and becomes a sequence. Run it in this order, because each step makes the next one worth doing.
- Can you be read? Foundation files, crawlability, structured data. Days of work.
- Can you be understood? Trade, scope, service area and credentials stated plainly on the pages that carry them. A week.
- Can you be verified? Google profile, licensing record and directories all agreeing. An afternoon of reconciliation, then ongoing discipline.
- Can you be corroborated? Independent sources that describe you. Months, and this is the one with the longest lead time.
- Did it work? The identical measurement, ninety days later.
Almost every contractor who feels invisible is stuck at step one or three, and almost every proposal they receive is priced for step four.
Questions builders actually ask
Is my competitor paying for this?
Almost certainly not. In most local trades nobody has done this work, which is why the gaps are so easy to close.
Could a bad review be keeping me out?
Rarely on its own. Contradictory information across sources is a far more common cause than negative sentiment.
How long until I show up?
Foundation changes can register in weeks, corroboration over months. Ninety days is the first honest read.
Share this article with AI
Finished it? Hand the whole thing to a model and ask what it means for your business specifically, including anything in it you think is wrong.
Opens in a new tab with the prompt already written. Nothing is sent until you press enter.
