When working with B2B enterprise software that is often sold in 5+ year contracts, how do you (or do you) effectively measure retention?

The end users are employees of a company that purchased this software and they do not have an alternative readily available. This situation would skew typical retention results.

Or is it best to skip end user retention and simply focus on the larger business's ability to retain the client over the long term?

  • 3
    Work in B2B as well, and as you point out, users are not the decision makers. So a lot of those metrics, such as adoption and engagement are not good UX metrics as they rely more on the client business decisions than individual users. That said, have you tried looking at clients rather than users? The aggregate of all users within a client could give you some insight, although there will be many factors influencing the retention, not just UX.
    – Martyn
    Feb 18, 2020 at 8:40
  • 1
    I'm facing a similar challenge working with a SaaS CRM. I need to find UX Metrics to help improve retention. I've been researching for a few days now, and I'm thinking the best approach is creating specific metrics for specific features of the product.
    – Aline
    Feb 18, 2020 at 12:49
  • You may want to have other metrics for UX quality, possibly very specific to the area of your product. As an example of the relatively universal feedback source, the focus group ('value') group might be a good source - and not even the exact feedback but the attitude of people towards this work. Yes, it's affected by the personalities and business environment but if you have one person running them in different companies the patterns may appear as well as some metrics to track. And, by the way - the content of this feedback is useful anyway, so that's a win-win although a lot of work. Good LUc
    – Nidea
    Jan 28, 2022 at 19:09

2 Answers 2


What is the need to measure retention for enterprise software where the decisions made around the procurement of the software is not often heavily weighed in favour of user requirements?

Also, for most enterprise level software, there is already a very high degree of information available around license usage, which translates directly to the dollar values of the revenue (and hence value generated for the business) as a result of usage.

If you were able to segment and isolate the feedback from end-users on an organisation-by-organisation level, perhaps you can come up with some available data to use and start finding some research questions worth diving deeper into.


As others have said, not all the HEART areas apply to every scenario. I've often found some subtle overlap and have at times chosen to leave one out as it gives less value than others.

That said, as Michael points out, you could look at retention at a high level, contract renewal, but conversely, you could look at a lower level, perhaps by feature.

You could investigate whether users find workarounds for holes in the system, or look for alternatives to perform a subset of features in a way that feels more optimal to them. You could interpret this kind of information in many ways, it could be something that is fed back as feature requests or workflow optimisation within the existing solution. It could be that splinter groups within the org resist using the solution and insist on non digital equivalents etc. It could be that through discussion, whilst using the product, users secretly pine for a competing product they prefer?

If none of this is true, you could still address retention, but treat it more like a goal of the other HEART areas, e.g. Is happiness high enough within the userbase that during the next contract renewal the solution will be extended, or have issues be uncovered which mean competitors would be evaluated.

Depending on your relationship with the incumbent provider, you could use this information to feed into their product roadmap, or cite problems that may result in non-renewal.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.