What distinguishes a "user" on the chrome web store?

[John]

According to this post: A chrome extension's user count listed on the chrome web store is the number of weekly users for that given extension.

For example, this chrome extension would have a weekly user count of 21,000,000 users:

However, I'm currently puzzled about what actually defines a "user." Does the browser keep some UUID for each chrome profile, or is it simply just the IP address of the user? Essentially, what is the unique identifier that separates users from each other?

Here's the case I'm working with:

I am installing an extension on many computers in a large corporate network - all devices have the same IP.

Will each device be counted as "user" even though they all have the same IP?

[Cuyler Stuwe]

In the absence of an informed and detailed first-party answer, this and other questions can be answered fairly effectively via some slow-paced testing on against a dummy unlisted extension not shared with anyone.

A couple of related questions I’ve pondered myself at times (but haven’t cared quite enough about to bother investigating/testing):

* Are logged-out users identified by machine? If so, are logged-in users identified as the one same “user” across multiple machines?

* Chromium users across many browsers outside of Chrome visit CWS to install extensions. Are these users counted in metrics? And if the default answer happens to be “yes”, what about for the Chromium browsers that make a stronger effort to divest themselves from Google services (e.g., Brave and Edge)?

Ultimately, my attitude on the CWS user count number is that it’s a “blurry/fuzzy” number. There are so many factors that make it usually irrelevant without context (e.g., it includes disabled extensions, people who have your extension installed but just aren’t using it at all, people from overseas countries who have installed your extension despite it being useful/usable only within your region, and so on). Because of all of this noise, it’s generally not worth stressing about these details.

--
You received this message because you are subscribed to the Google Groups "Chromium Extensions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-extens...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-extensions/97f86fdd-2089-4068-b117-a5939d657da1n%40chromium.org.

[John]

How would you test and debug to answer above questions? Really hoping someone from the CWS team can provide a first-party answer.

[Cuyler Stuwe]

With an unlisted throwaway extension you haven’t shared with anyone, you can reasonably assume all installs are your own.

You know that the number at a high level represents essentially the number of unique checks for an update over a 7 day period.

So if you make a change and you wait 2 weeks to test the results, you can come to some reasoned conclusions about how this works.

[John]

That may work, but as you said earlier, it would be rather "slow-paced." Again, hoping someone from CWS with even a shred of information about this can shed some light, as they have done in similar situations.

[Cuyler Stuwe]

It seems like the type of number that a company would be intentionally vague about, to avoid having people “game” it.

I’d venture a guess that therefore nobody is going to come in and offer a system diagram describing the details of how it works.

It would be cool if they did, but I wouldn’t hold my breath. If you browse previous posts on this forum, there are periods where it feels like about 5-10% of the posts are people asking why their own numbers don’t seem to align with CWS numbers.

[John]

Yeah that makes sense.

If any team members do feel comfortable sharing, please feel free to privately email me the details.