Glossary

Technology churn

Technology churn is the share of companies that stop using a technology in a period. How to calculate it, read removals from outside without false alarms, and spot a switch.

Updated 5 October 20262 min read

Technology churn is the rate at which companies stop using a technology: the share of those that had it at the start of a period that no longer have it at the end. It can be measured for a product, a category or a vendor, and is read beside the rate of adoption.

Gross, net and switching

Gross churn is the removals in a period divided by the installed base at its start, the companies that had the technology. Net change also counts additions, so a technology can lose users and still grow. Acme Robotics (illustrative) has its tag on 2,000 websites at the start of a quarter. By the end, 150 have removed it and 100 have added it. Gross churn is 150 divided by 2,000, or 7.5 percent, and net change is minus 50 websites, or minus 2.5 percent.

Separate dropping from switching. A company that removes one analytics tool and adds a rival's has switched, and that pair of events is the signal behind competitor displacement plays.

Reading removals from outside

A removal is an absence, and absences are noisy. Tags come and go with consent banners, tests and site rebuilds. Require the technology to be missing on more than one reading, hours apart, before calling it removed, and count a removal only from a reading that covers the view in which the tool used to be seen. Even a clean removal is not a cancelled contract: it says the tool has gone from the site, and only the vendor knows whether the account left.

In Fokals data

Technology Changes records a removal as a dated event with its observation time, once the technology is confirmed missing on a later observation. A commerce or content platform replaced by another is recorded as a platform migration. A first observation of a site sets a baseline and writes no events.

Market Series carries weekly removal series per 100 websites, on same-store cohorts, so a rate is over the sites covered before its window began. The marketing stack dataset holds the datasets, and the guide to churn risk from technology removals shows a vendor watching its own tag.

Churn is measured against an installed base. Removals and additions come from technographic data, often seen through a tag manager, and a switch is a common trigger event.

Frequently asked questions

Does a removed tag mean the customer has cancelled?

Not necessarily. A tag that is no longer detected can reflect a cancellation, a replacement, a redesign, a test, or a change that hides the tool from the page. Treat a removal as a reason to look, not a conclusion. Check whether a competing tool appeared on the same site in the same window, and whether the same tool is still seen through another view, such as a tag-manager container.

How do you find companies that have dropped a competitor's technology?

Filter Technology Changes for the competitor's technology with a change of removed inside a recent window, then look for an addition of another tool in the same category on the same site. In Technology Changes that means events for the technology, a change of removed and an observation time inside your window. A removal with no replacement may be an abandoned tool, and one with a replacement is a switch.

How long should a tool be missing before it counts as removed?

There is no universal rule, but one missed reading is too few, because tags disappear briefly for reasons unrelated to the contract. Confirm a removal when the technology is still missing on a later reading some hours after the first miss, taken over the view in which the tool was seen. Longer waits trade speed for fewer false alarms.