What is the term used to describe the collection of user usage data?

Question

We want to implement the collection of user usage data, for example crash reports, errors reported in logs, and normal usage patterns & statistics (how often the application is launched, how long certain tasks take to execute, etc.) I'm trying to determine whether there are third-party solutions for doing this (via some library you link against and a service they provide for collecting the data and uploading it to a service they provide, and we can query), or whether it makes sense to roll our own.

The fact that I don't know the industry term for this practice is preventing me from finding companies that may provide this service. I've always used the term "software telemetry" for this, but searching for that seems to be giving me just results related to traditional telemetry (tracking animals & such).

We have around 100,000 users, and our app is written in C++ in a unix environment, although it would not be strictly necessary to have system that has C/C++ libraries, as we could always create an intermediate file from our code that is uploaded to a remote service via a separate process.

I'm not sure if there's something more accurate, but [software analytics](http://en.wikipedia.org/wiki/Software_analytics) comes to mind first. — Drew Dormann, Apr 03 '15 at 14:31
The term "telemetry" is quite common, but as you noticed it has conflicting uses. — molbdnilo, Apr 03 '15 at 14:54

score 0 · Answer 1 · answered Apr 06 '15 at 12:06

I have seen the term Process Mining used in this context, from the paper Is Your Upgrade Worth It? Process Mining Can Tell published in IEEE Software (paywalled).

Abstract:

Software vendors typically release updates and upgrades of their software once or twice a year. Users are then faced with the question of whether the upgrade is worth the price and the trouble. The software industry doesn't provide much evidence that it's worthwhile to upgrade to new releases. The authors propose the use of process mining to prove that upgrading to the next release provides quantifiable benefits to the end user. Process mining capitalizes on the fact that event logs capture information about processes. These events can be used to make processes visible and show the benefits of using a software product's next release. Three groups benefits from this process: end users, software suppliers, and researchers. The authors applied process mining to a medical software product and captured empirical data from 1,400 cases. The data shows that the new version was 11 percent more efficient than the old release.

score 0 · Accepted Answer · answered Apr 20 '15 at 13:05

0

The terms I was looking for are "Application Performance Monitoring" and "Software Analytics".

answered Apr 20 '15 at 13:05

Steve Broberg

4,255
3
28
40

What is the term used to describe the collection of user usage data?

2 Answers2