Companies Using Microsoft Azure

This page was rebuilt from the ground up in August 2026. It had been running on a detection identity our crawler no longer uses, which meant every number on it was more than a year stale and the category it sat in was the wrong one. Since the underlying analysis changed completely, so did the conclusions. Here is what the current data supports.

What the detection proves

Three signals put a domain on this list. The first is DNS: the domain's nameservers are delegated to Azure DNS. The second is a session-affinity cookie, ARRAffinity or TiPMix, which Azure App Service sets so that a user keeps hitting the same instance. The third is an Azure service header on the response.

TechnologyChecker.io resolved a random sample of 260 detected domains by hand. 59.2% matched on nameservers, 38.8% on the App Service cookie, and 91.9% matched on at least one of the two. Only 21 domains matched nothing we could confirm, and six of those were blocking our request rather than failing the test, so the genuinely unexplained share is 5.8%.

That is the number to hold onto, because the same test on AWS returned 67.7%. The difference is structural rather than lucky: AWS's rule includes a clause that fires when the string amazonaws.com appears anywhere in a page's JavaScript, which a third-party widget can satisfy on a site with no Amazon account at all. Azure's rule has no such clause. Every signal in it belongs to the domain owner.

The header clause is worth a footnote because it fired zero times in 260 domains. That is not a broken test; we verified the check against known-good headers. It is that a CDN in front of the origin replaces the server header, while a session-affinity cookie has to pass through untouched or the feature breaks. The cookie survives the proxy and the header does not.

Sales Signal: A list this clean can be worked without a qualification step. Where you would filter an AWS list down to the domains that prove an account, roughly nine in ten of these already do.

Growth trajectory

Our earliest record is 349 domains in May 2013 and the current figure is 175,651. Do not turn that into a growth multiple, because three months in between are ours rather than Azure's: April 2014, January 2018 and February 2019 each show active and total domains jumping together with the ratio flat, the signature of coverage widening. The previous version of this page reported the February 2019 jump as evidence of an "enterprise cloud migration push". It was us starting to look harder.

The consistently measured stretch runs from 2019, over which Azure moved from 40,431 to 125,613 domains by the end of 2024. Just over 3x in five years, for a platform that was already mature at the start of it.

"The old page turned our own coverage change into a story about Azure's sales motion. That is the most common way a long chart lies, and the tell is always the same: real growth moves active domains against a stable total, while a coverage change moves both at once." - David Thomson, CTO at TechnologyChecker.io

Sales Signal: Ignore month-to-month movement in the headline while coverage is still widening. The migration table is where an individual domain changing hands is a real event you can act on.

Customer profile

Azure's base is materially more enterprise than AWS's, and because both are detected the same way in the same category in the same month, the comparison holds. 7.49% of Azure's matched companies employ more than 1,000 people against 2.56% of AWS's, and 19.05% employ more than 200 against 8.14%. On age the split is just as wide: 44.44% of Azure's base was founded before 2000, against 21.34% of AWS's. AWS's base is 63% companies founded since 2010; Azure's is 37.6%.

One honest caveat that runs the other way. Azure's rule catches App Service, a product small teams use, and AWS's platform rule has no cookie clause at all because its load-balancer cookies are split into separate technologies. If anything that asymmetry should pull Azure's numbers down-market, so the enterprise skew survives it rather than depending on it.

Sales Signal: If you sell into procurement cycles, the ratio between these two lists is the useful fact. Per thousand detected domains, Azure yields roughly three times as many 1,000-plus-employee companies as AWS does.

Industry and geographic concentration

IT Services leads at 6.8% and Software Development follows at 6.32%, which is unremarkable. Third place is not: Insurance at 4.55%, a vertical that does not reach AWS's top ten. With Financial Services at 4.33% and Hospitals and Health Care at 2.65%, regulated industries account for 11.53% of Azure's matched base.

Geography tells the same story from another angle. The United States takes 29.99% here against 41.76% on AWS, and the shortfall reappears in Europe: the Netherlands, Germany, Sweden, Switzerland, Norway, Belgium and Denmark together reach 16.87%, with four of those seven absent from AWS's top fifteen entirely. Brazil at 6.78% is the largest non-Western market, ahead of Canada.

Sales Signal: The Nordics and Benelux are where Azure is over-represented relative to its overall size. For compliance, data-residency or sovereignty tooling, that concentration lines up with the regulated-industry tilt rather than cutting across it.

Migration patterns

The biggest corridor is 1,740 domains that dropped Azure and now run Cloudflare, against 474 inbound. That number needs the same discipline as the equivalent on the AWS page. Cloudflare onboarding replaces a domain's nameservers, and nameservers are what 59.2% of these detections rest on, so a company can move DNS and keep every workload where it is. We measured it: 19.94% of that cohort still runs another Azure service, and that is a floor, since it only catches the ones running a second Azure product we detect separately.

Against AWS the flow is as close to level as this data gets: 329 domains arrived from AWS and 311 went the other way. Both figures appear on the AWS profile with the directions reversed, which is a useful check that the two pages are reading the same events. Azure is clearly ahead against Amazon S3 (510 in, 221 out) and Google Cloud CDN (386 in, 84 out), and behind against Vercel (62 in, 147 out).

"Level is the finding. Everybody expects one of these two to be quietly bleeding into the other, and across two months of observed churn neither is." - David Thomson, CTO at TechnologyChecker.io

Sales Signal: The AWS corridor is a poor prospecting list in either direction, because it is balanced and small. The 510 domains that moved off Amazon S3 onto Azure are the better one, since a storage migration usually means somebody has just rewritten how their assets are served.

Technology ecosystem

17.1% of Azure domains carry a second Azure service, led by Front Door at 9.08% and Monitor at 7.09%. We would not compare that with the AWS figure, and it is worth saying why: our ruleset detects roughly eleven AWS infrastructure products against six for Azure, and the two biggest AWS ones serve files that show up on any page. That number is as much a statement about our coverage as about anyone's architecture.

What does compare is the multi-vendor picture. 20.67% of Azure domains also run Cloudflare against 34.26% of AWS domains, and 3.47% run AWS alongside Azure outright. Google Analytics sits on 53.49%, which tells you nothing about Azure. Microsoft Dynamics 365 appears on 1.13%, low enough that the Microsoft-estate story is about identity and productivity rather than anything visible on a public website.

Sales Signal: The 2,481 domains running AWS with no second Azure service are the sharpest multi-cloud list on this page. They are paying two vendors for infrastructure and can see both invoices, which is the precondition for any cost-management or workload-placement conversation.