How many US manufacturers use HubSpot? We checked 1,000 websites

About 3.5% of US manufacturers listed on Google Maps run HubSpot on their website. In October 2026 we checked a random sample of 1,000 manufacturer websites from Google Maps: 29 of the 817 sites we could read had a confirmed HubSpot portal (95% confidence interval 2.5% to 5.1%), and 2 more were probable.

This is Scout's own measurement. Below is how we ran it, what we found, where detectors go wrong, and what the number does not tell you.

How we measured it

We started from 107,696 unique websites of US businesses listed on Google Maps under custom manufacturing and machinery manufacturing. Social profiles, marketplaces and site-builder pages were removed first. From that list we drew 1,000 websites at random with a fixed seed, so the sample can be redrawn.

We did not try to get around bot protection. The 183 sites we could not read are left out of every rate on this page.

For each readable site we fetched the homepage and looked for HubSpot's script hosts and portal ID. When a site used Google Tag Manager and showed no HubSpot code directly, we also fetched the Tag Manager container. Every hit was then loaded in a real Chromium browser to see whether HubSpot actually made requests. Hits where the browser showed nothing were reviewed by hand. As a check on misses, we also loaded 150 random sites with no hit in the browser.

What we found

Of the 150 random sites with no hit, none loaded HubSpot in the browser. Hidden installs that a careful scan misses are rare.

Why almost a third of the hits were invisible in the page source

10 of the 35 raw detections were visible only inside the Google Tag Manager container, not in the homepage HTML. A detector that reads only the page source would have missed them. If you build or buy technographic data, ask whether it reads tag manager containers.

Three ways a HubSpot detector gets it wrong

All four false positives came from three patterns, and each one can be filtered with a simple rule:

1. Cookie consent plugins. Some consent tools ship a list of known tracking scripts, HubSpot included. The pattern is on the page, but nothing loads.
2. Resource hints only. A dns-prefetch or preconnect link to a HubSpot host, with no script and no portal.
3. Agency portals. Two unrelated businesses shared the same Tag Manager container and the same HubSpot portal ID. The portal belongs to the agency or website platform that built both sites, not to either business.

The rule that removes all three: require a portal ID or a real request to HubSpot, and reject a portal ID that appears on unrelated domains.

A portal is not proof of CRM use

A HubSpot script proves the company has a HubSpot account. It does not show whether the sales team runs its deals in HubSpot. Some signs point closer to the CRM. Of the 24 portals that loaded in our browser check, 7 ran HubSpot chat, which writes conversations into the CRM. None linked to a HubSpot meetings page, and none ran Pardot, Marketo, Zoho or Pipedrive next to HubSpot.

Job postings did not help find HubSpot users. We searched the past three months of postings for the 31 confirmed or probable companies. 6 were hiring for sales roles, but not one posting asked for HubSpot by name. The website was the only reliable source.

Limits of this study

• The population skews small and local. Google Maps manufacturing listings include machine shops, print shops and fabricators alongside larger industrial firms. B2B companies with dedicated sales teams likely use HubSpot more often. We did not measure them here.
• 18% of the sample could not be read. If HubSpot users block bots more or less often than other sites, the true rate differs.
• One point in time. The scan ran on October 21, 2026.
• A portal is an account, not usage. See the section above.

What this means if you sell to HubSpot users

If your agency or product serves companies on HubSpot, the website is a cheap and accurate filter, as long as the detector reads tag managers and applies the three rules above. In this market roughly 1 in 28 readable sites qualifies, so the filter alone gives you a short list, not a reason to call. Pair it with a timing signal such as a sales hire, then verify the event at its source. Our guide to finding companies that need your services covers that second step, and you can request five free leads on your own ICP.