How Much of Magento 1 Is Still Around?

Magento 1 still has thousands of detected live websites. But our data does not support a precise total or a reliable percentage of the Magento market.

That may sound cautious. We arrived at it after finding a tempting shortcut in the data—and then testing whether it worked.

As part of our Magento market research, we bought BuiltWith data. We had two exports dated August 24, 2026: one for all live Magento sites and another for Magento 2.

The first contained 151,428 unique root domains. The second contained 109,428.

Subtract one from the other, and the answer is exactly 42,000.

It would make a neat headline: “42,000 websites still run Magento 1.”

But it would not be a finding we could defend.

The two lists do not fit neatly inside each other

We compared the actual domains rather than just the totals.

The Magento 2 list was not a complete subset of the general Magento list. It included 2,289 domains that the broader export did not contain.

That means subtracting the headline totals does not even tell us how many general Magento domains were missing from the Magento 2 list. The actual unmatched count was 44,289, or 29.2% of the general export.

This is a useful research group. It is not a verified Magento 1 group.

Differences in detection, indexing, report scope, or activity across different parts of a website could help explain the mismatch. The exports alone do not tell us which explanation applies to each domain.

A Hyvä signal exposed the problem

We then looked at the technology labels attached to those 44,289 unmatched domains.

2,491 had a Hyvä Themes label.

Hyvä describes its products as built for Magento 2 and Adobe Commerce. A Hyvä signal is therefore a strong reason to question an automatic Magento 1 classification. Hyvä documentation

These signals could reflect incomplete coverage in the Magento 2 export, multiple storefronts under one root domain, or a detection error. We have not manually checked all of them.

But the finding is enough to reject the shortcut:

Missing from a Magento 2 list does not mean confirmed Magento 1.

We also cannot simply remove the Hyvä domains and call everything left Magento 1. The same detection gaps may affect other sites without that label.

Direct version signals give us a firmer starting point

To find positive evidence of Magento 1, we checked BuiltWith's public reports for specific versions.

At the time of this research, on September 3, 2026, the retrieved pages reported:

These are version-specific detections. They are stronger evidence of a surviving Magento 1 footprint than the absence of a Magento 2 label.

They are not a complete, deduplicated count of Magento 1. A domain may appear in more than one version report, and these three versions do not cover every Magento 1 edition or fork. The public pages also represent a different observation date from our purchased exports.

For those reasons, we do not add the figures together or divide them by the August export total to claim a market share.

The supported conclusion is still meaningful: BuiltWith continues to detect thousands of live sites with Magento 1 version signals.

Be careful with redirects and historical sites

The Magento 1.9 report illustrates another easy counting mistake.

It lists 7,734 live websites and 8,705 additional redirect websites, producing a larger list of 16,439. It also separately lists 112,742 historical sites. BuiltWith Magento 1.9 report

Those figures answer different questions. Redirect domains can point to an existing store, while historical detections are not current usage.

Using the larger list total as the number of active Magento 1.9 stores would overstate what the report says. Historical totals also do not tell us how many merchants migrated during a particular year or which platform they chose.

OpenMage changes what “still on Magento 1” means

There is another group to consider: stores that remain in the Magento 1 family while using a maintained community fork.

OpenMage is a community-driven fork of Magento Community Edition. It provides a continuation path for Magento 1-based stores. That is different from simply leaving the original software untouched. OpenMage

Our general Magento export contained 1,493 domains with an OpenMage label.

Of those:

• 1,461 were absent from the Magento 2 export.
• 32 appeared in both the OpenMage-labeled group and the Magento 2 export.

That small overlap is another reminder that root domains and software installations are not the same unit. A root domain may contain different storefronts, or the recorded signals may need checking.

We should not add the OpenMage count to the public Magento 1 version counts: we have not checked their domain-level overlap, and they come from different snapshots.

For market research, it is more useful to distinguish between original Magento 1 installations, Magento 1-based forks, and unresolved detections. These groups can have different maintenance needs and migration plans.

End of support was not the same as disappearance

Adobe's June 2020 security bulletin stated that support for Magento Commerce 1.14 and Magento Open Source 1 was ending that month. More than six years later, version-specific detections remain visible. Adobe security bulletin

The data shows that continued presence. It does not explain why individual merchants stayed.

Possible reasons include the cost of rebuilding integrations, reliance on custom workflows, limited resources, or a decision to use a maintained fork. These are questions for merchant interviews, not conclusions we can extract from a technology label.

A detected Magento 1 site does not automatically tell us whether the business is active, whether its software is maintained, or whether it has approved a migration budget.

For a service provider, those differences matter more than the raw domain count.

How we would measure the remaining market more accurately

The next step is to build a domain-level list from direct Magento 1 version detections, including relevant Community and Enterprise versions, and record OpenMage separately.

We would then deduplicate domains and compare them with Magento 2 detections from the same period. Mixed or conflicting signals should remain unresolved until checked.

A sample review should establish whether the website is a working store, a redirect, a parked domain, or a test environment. Merchant or agency confirmation would help resolve version and maintenance questions that public detection cannot answer reliably.

Only after that should we compare traffic, catalog size, and business activity to estimate the part of the market that may need maintenance or migration services.

Repeated snapshots would answer a different question: how quickly is this group changing? One export cannot tell us whether the remaining base is shrinking slowly or rapidly.

So, how much of Magento 1 is still around?

There is enough evidence to say that Magento 1 has a surviving footprint measured in thousands of detected websites. OpenMage also has a visible presence in our purchased data.

There is not enough evidence here to say that 42,000—or 44,289—active stores still run Magento 1. Nor can we turn the unmatched 29.2% of the general export into a Magento 1 market-share figure.

The useful story is that Magento 1 remains visible well after official support ended, but its remaining market needs careful classification. Some sites may need a new platform. Some may use a maintained fork. Others may no longer represent active merchants.

Finding out which is which is the next stage of the research.

Data and methodology

We compared two local BuiltWith exports named All Live Magento Sites — 2026-08-24 and All Live Magento 2 Sites — 2026-08-24. The second CSV was read directly from its ZIP archive. We matched the Root Domain field after trimming whitespace and converting it to lowercase. Both exports had unique root domains within their respective lists.

The match reconciles as follows: 107,139 shared domains plus 44,289 general-only domains equals 151,428. The same 107,139 shared domains plus 2,289 Magento-2-only domains equals 109,428.

Hyvä Themes and OpenMage counts use the corresponding exact labels in the Ecommerce Platform field of the general export. These labels are detection signals, not a manual assessment of the software currently running.

Public version counts were retrieved on September 3, 2026. They are reported separately from the August exports and can change as BuiltWith updates its data. We did not inspect every website or obtain a complete set of deduplicated Magento 1 version exports. No precise global Magento 1 total or market share is claimed.