If you found us in your server log.
Something called TrueMeasure-Inspector read a page on your site. This page explains what it is, what it took, and every way to make it stop. You do not have to write to us to be left alone, though you are welcome to.
TRUE MEASURE WEB DESIGN LLC runs True Measure Inspector: a free inspection anyone can run on their own website, and a census built by inspecting business websites one at a time.
To know whether a finding means anything, we first have to know what ordinary business websites look like. So we read a lot of them and keep the results. Your site was almost certainly one of those readings.
What it took from you
Name
TrueMeasure-Inspector/2.0
Full user-agent
Mozilla/5.0 (compatible; TrueMeasure-Inspector/2.0; +https://truemeasurewebdesign.com/inspector)
Requests per business site
Up to 9, plus any redirects your server sends us through. First robots.txt, the file of rules for automated readers. Then the page, robots.txt again, two sitemap addresses (lists of your pages), one stylesheet (the file that sets your look), your icon, your sharing image, and one address we expect to be missing.
Business directories
We also read one public business directory, a chamber of commerce member list, to find business websites. We read it a few times a year. There we send at most one request a second and obey every rule in its robots.txt. We follow links only to its own pages, within the limits below.
How often, for a business site
Live sites: at least 1 day apart, typically every 6 days. Silent sites: again after 2, 4 and 8 days, then out to once every 90 days. While the census re-reads every site with a new version of our engine, the program that reads sites, a site that once answered may get one early try.
Time limit
Eight seconds for the first robots.txt, twelve for the page, six for anything else. We give up rather than hold a connection open.
Size limit
Three megabytes for the page, half a megabyte for the first robots.txt. We stop reading there and close the connection.
What we run
Nothing. We read what your server sends and close. We do not run your scripts, fill your forms or sign in to anything. On a business website we follow no links.
How to slow us down
Put a crawl delay in your robots.txt. We read it and we obey it.
User-agent: TrueMeasure-Inspector
Crawl-delay: 10We wait that long between the starts of two visits to your website, up to 60 seconds. A visit is the requests listed above, sent together in one short burst. Ask for longer and we take it as a no.
How to block us
The ordinary way, in robots.txt. We check it before we read anything.
We read robots.txt at the address we were given, and a rule that shuts out the whole site stops us.
We do not check narrower rules, such as one folder, against the small files we fetch with the page. At an address your page forwards us to, we read its robots.txt for the report, not for permission. At an address it borrows a file from, we do not read robots.txt.
User-agent: TrueMeasure-Inspector
Disallow: /That takes effect the next time we look, and it costs you one small file read. If you would rather we never touch your server again, including that file, use the next option instead.
How to be left alone completely
Write to hello@truemeasurewebdesign.com with the address of your site. We add it to a list read before any connection opens, so neither the census nor the free inspection starts reading it again, robots.txt included.
A different site that forwards to yours, or borrows your image, icon or sitemap, can still send a request your way.
We do not ask you to prove you own the site. If somebody who did not own it asked us to stop, we lose one measurement, and the real owner can write to us to have it read again. Turning away a real owner would cost us far more than that.
What we keep, and what we never publish
We keep our own reading of your page: what it scored, which checks passed, what kind of business the page looks like, and the town it publishes. That is a record of our measurement. The only copies of a page we keep are the study copies below.
To improve how we recognize kinds of business, we sometimes keep a copy of a business's front page for one named study, as our crawler received it. A copy holds whatever that page shows, which can include a name or email address the business printed on it.
Study copies are kept privately and are never published. Our rule is to delete them when the study ends. Copies from one finished study are still on our computer and are waiting to be deleted.
We also keep whether a business shows a phone number, with a fingerprint of each number: a code made with a secret key kept only on our census machine. It can be matched against a public register, an official list of businesses, and cannot be turned back into the number without that key.
We keep its license numbers as printed, for the same check.
Readings taken before this change reached our census still hold the phone number itself.
If your business is listed in a public business directory we read, we may keep part of its listing, even if it shows no website. That part is its name, categories, city, state, ZIP code and listing number, the year it joined and its website.
We also note the county we place it in. We never keep a listing's phone number or street address. When a listed website address points to a street address or what looks like a person's own page, we keep only the bare site.
If a number is a person's, tell us: that business leaves the census for good, with its readings, our map copy and any directory listing deleted. Copies of those readings stay in our backups, logs and file history.
A copy of the listing stays in a backup kept only on our own computer.
Separately, if someone asked a free inspection to remember that site, we keep its address and its last result.
Apart from what is described above, we store nothing about a person: no visitor records, and nothing about whoever happens to be using your site.
We never publish your name. When we report what we have found across the websites we have read, every figure is a count and no business is identified. A measurement is the point. Naming anybody would make it something else.
If we got something wrong
Our tool can be wrong. When it was wrong, we said so in public and fixed it. Our corrections page lists each sentence we fix.
If it read your site badly, tell us what it missed. A reading we cannot defend is worth more as a correction than as a number.
If this is your site and you want the report, you can run the same inspection yourself and read everything it found, at no cost and without giving us anything.