We tried to read what hotels promise. Most of them said no.
We audited 20 major hotel brands against their own published standards. 19 of them carry at least one commitment we could not verify — not because the promise was missing, but because their website would not let an automated reader see it.
We set out to do something unremarkable: check what the biggest hotel brands actually promise business travelers, by reading their own published standards. Not reviews. Not opinions. The brands’ own words, on the brands’ own websites.
Of 20 brands audited, 19 carry at least one commitment we could not verify — because the brand’s own website refused to let an automated reader see the page. One returned a lockout of roughly eight hours.
Two numbers, because the difference matters and rounding them together would overstate this: 18 of those 19 also publish something we could read, so the gap is partial. One brand could not be verified at all. We would rather publish the smaller true number than the larger striking one.
Is it robots.txt, or is it the server?
These are different things and only one of them is a decision. A robots.txt file is a published request — a site saying which automated readers it would like to allow. A 403 from the server is the door itself.
Across every source behind every index on this site, the number disallowed by robots.txt is 0. Not one. Every refusal recorded here happened at the server, and 13 of 19 refused sources sit on hosts whose own robots.txt explicitly permits the crawl their server then refuses.
That gap is the clearest evidence that none of this is policy. These sites are publishing an invitation and enforcing the opposite, which is what an unrevisited bot rule looks like from the outside. It is also, precisely, a promise not matching delivery — the thing this desk exists to measure, arriving in our own inbox.
And our own position, stated plainly, because you should not have to ask. When we verify a source this desk already cites, we fetch it and record what its robots.txt said rather than skipping it — identified user-agent, one request per host, and a person rules on every passage. Sources read against a stated disallow, to date: 0. If that number is ever not zero, it will say so here.
Does the same problem affect software vendors?
No, and we checked rather than assumed. Across the 16 travel-management and expense platforms this desk scores with a recorded source, 16 are readable by an ordinary automated reader and 0 are refused.
The contrast with hotels is stark and worth stating plainly: consumer hotel sites refuse machines; software vendors' product and pricing pages do not. If you are a brand wondering whether blocking is simply the cost of doing business online, the sector selling software to your corporate customers is not paying it.
We say this the same way we would say the reverse. An earlier version of this desk's own reporting claimed four software vendors blocked automated readers. That was our bug — our robots checker was fetching with a user-agent those sites refuse, and treating the refusal as a prohibition. We retracted it before publication. The figures on this page are what survived being checked twice.
Which hotel websites block automated readers?
These are the domains that returned an access refusal to a straightforwardly identified reader, and how many audited brands each one covers:
| Domain | Brands affected |
|---|---|
marriott.com | 8 |
ihg.com | 4 |
hilton.com | 4 |
hyatt.com | 2 |
omnihotels.com | 1 |
Exactly one brand of the twenty was readable end to end. Corporate newsrooms and franchise-development sites were generally open; the consumer sites, where guests and buyers actually look, generally were not.
Why does this matter more than it used to?
Because the reader has changed. When somebody asks an answer engine “does this hotel brand guarantee Wi-Fi?”, that engine goes looking for the same page we did, identifies itself the same way, and gets the same refusal.
A brand that blocks automated readers has not protected its commitments. It has made them unquotable — invisible in the one channel where a growing share of travel decisions now begin. The promise still exists. It simply cannot be found by the thing doing the finding.
Is this deliberate?
Almost certainly not, and it would be unfair to imply otherwise. Blocking is what bot defence does: it stops scrapers, price-crawlers and inventory bots, and those are real problems worth solving. What appears to have happened is that the blanket got wider than anyone intended, and nobody revisited it once answer engines started reading the open web on behalf of customers.
The tell is that the same companies keep their newsrooms wide open. The intent was clearly never secrecy. The effect is simply that the marketing site talks and the standards page does not.
What did this desk do about it?
We recorded it as unverified — never as absence. A page that refuses an automated reader may open perfectly for a person, so scoring a brand down for it would be a false claim about the brand, and scoring it up would be a false claim about us. Nobody lost a point here.
Every affected entry says so in public, on the page where the score appears, naming the domain that refused. That is the same standard we hold vendors to: publish the gap rather than paper over it.
How does a brand fix it?
Two ways, both cheap.
Let the readers in. Allow well-identified automated readers to reach the pages that carry your standards — the room standard, the connectivity commitment, the guarantee terms. Not your booking engine, not your rates. The pages you already want people to find.
Or just send us the document. Any scored brand can email the published standard and we re-score within seven days, free, under published governance that applies to everyone equally. A brand may also submit a response of up to 150 words that we publish verbatim, including where it disagrees with us.
We would rather write next year’s version of this piece about how few brands are still unreadable.
What does this mean for a travel manager today?
Ask for the document. If a brand cannot send you a published standard for in-room work space, connectivity or quiet, then whatever your rep experiences on the night is not something you contracted for — it is something you hoped for. That distinction is the whole of the Revenue Travel standard, and it shows up on the invoice eventually.
The Field Brief — sharp travel-ROI notes for the people who carry the quarter through airports. No signup, no email, nothing to unsubscribe from: read it free in the open archive.
Read the brief →blocked after a live re-fetch confirmed the refusal. Each refusal was recorded against primary sources — the brands’ own domains — during the audit dated July 30, 2026. A blocked page is unverified, not disproven, and no score was changed on that basis. Any brand may correct or update its entry free of charge under the scoring governance. The Sales Traveler is independent and reader-funded — a partnership buys reach, never a rating.