Download websites that use Microsoft Word

About Microsoft Word

Finding Microsoft Word on a website is not a statement about the server. Word does not run on the web, and when our crawler records it the evidence is a fingerprint left behind by an export: markup saved out of the application with its generator metadata intact, or an office document published straight to a page. What the detection really says is that somebody in the organisation authored the file on a desktop and put it online without passing it through a publishing workflow.

Presence is therefore a signal about process rather than about stack. Sites that publish Word output tend to be institutions with more documents than developers: local government, schools, law practices, clinics, standards bodies, membership associations. A pipeline that ends in a desktop export usually means no editorial system in the middle, an intranet culture pushed onto the public web, and pages written by subject specialists rather than by anyone who works on the site.

Microsoft Word has been a Microsoft product under a Proprietary license since 1983, which makes it older than the web it now leaves traces on. The generator strings we collect reflect that long history. We record 16 distinct versions and the one appearing most often is 15, but the version that produced a file says nothing about when the page went live, only about which desktop it came off. Treat the version column here as weak evidence about the organisation and no evidence at all about the age of the page.

The country picture behaves the same way. United States supplies 10,113 of the detections and .com leads the domain side, which tracks where office software licensing is concentrated rather than any property of the websites themselves. Reading it as a statement about national web development practice would be a mistake.

As of October 2026 we see traces of the application on 31,741 sites, filed under Editors because there is no better home for a desktop program that keeps surfacing in web output. A per-domain check tells you one page came out of Word. The downloadable list tells you which segments of the web still publish that way, which is the question that matters if you sell document conversion, accessibility remediation, PDF tooling, records management or a redesign. Accessibility work in particular starts here, because exported markup is reliably poor at heading structure, alternative text and table semantics, so a list of sites doing it is a list of remediation candidates with a defensible reason to be contacted. The limit is worth repeating: we can only see exports that kept their generator metadata, so the count is a floor, and the organisations whose tooling strips that metadata are invisible to us even when their process is identical.