That's not Firefox's all-time low, but it's still miserable. Chrome, as you might guess, is the leader with 54.5%, followed by Safari, thanks to iPhones, with 24%, and Edge with 14.2%. Only Internet Explorer -- yes, some people are still running it -- of the well-known browsers has a smaller number of users, with 1.7%.
Edge is Chromium-based so, already that indicates a poor research. Unless we're not counting it for Bing being default? But then if FF ships with Google as default, what does that measure as? What about browsers like Tor Browser that are FF based? All those "others" in market share count on either side. Anyway, there's no citations, so we don't know how this was measured. Lots of privacy features spoof user agents, too. And Chromium is the most common, so there's that. Anyway my thoughts:
- Mozilla is basing some metrics on new user profiles created [1]. IDK if that's a good measure, many people copy+paste / export+import their profiles or simply never run in anything but private mode (which I would think prevents a profile being created, but I may be wrong on the latter).
- Mozilla's metrics on OS show a massive dip in Win10 users and a near-equal rise in Win11 users: It's natural for the user-ship to drop on older OSes. Win10 is going down, but Win11 is up at a near equal rate [1]. Not sure what that means to overall numbers other than "no change"?
- Because the article doesn't like to primary sources (the research) it is hard to say how these measures are done. While Mozilla does have a measure its userbase per OS, and acknowledges Windows makes up 75% of its userbase [1], neither Mozilla nor the article provides an evenly distributed sample; example 50% Windows, 50% Linux (illustrative oversimplification). Why do we do this in research? Because lets say Group A is 15000 of which 13000 people who prefer pop to jazz and Group B is 5000 of which 2000 prefer pop. For pop, Group A has 86.7% market share whereas for Group B a 40% market share. If we add both groups, ie 20k people, we get a 75% market share for pop (15k / 20k). "But then that makes sense, right?" Wrong. For an accurate sample group, there would need to be equal measures of demographics within that group. So we need 10k Group A and 10k Group B. Let's assume both groups are homogeneous to make this simple. So now Group A is 8667 who prefer pop and Group B is 4000 who prefer pop. Now pop only has a 63% market share. That's a 12% or 2400 person difference when doing research accurately. When working with percentages of anything, there's a stark difference between just numbers as opposed to trends. A light example: in a classroom of 40 girls and 10 boys, where 30 girls were crushing on Justin Timberlake, you would not say 60% of total teens have a crush on Justin Timberlake. Your sample group is in this case biased... At least that's what I was taught in school.
[1]
https://data.firefox.com/dashboard/user-activity
And there's my rant. In closing, I'd like to point out that while correct research methodology does not necessarily mean those stats would change, it does call them into question. Remember not every user in every country has had their primary browser measured. Remember mobile users often use the default browser due to lock-ins and that's often Chromium on Android, which has an excessively high market share in consumer electronics in general. While we can't slice 'n dice everything, we can do better than sweeping statements, or at least state how we reached our numbers.