For two years in 2020 and 2021, I shared Wikipedia’s worldwide browser statistics on Mastodon under #browserstats. They looked a little something like this:

Browser data from 3 May 2021 to 29 May 2021.

Wikipedia.org and sister projects, browserstats for May 2021:

  • 49%: Chrome + Chrome Mobile
  • 24.7%: Safari + Mobile Safari
  • 5.2%: Firefox + Firefox Mobile
  • 2.8%: Edge
  • 2.5%: Samsung Internet
  • […]

100% = 16.4 billion page views (not including bots)

As the data includes the browser’s major version, I wondered whether I could use this to follow the adoption rate through each browser’s release cycle. The short answer is… Yes! Here is what I found as of May 2021:

  • Firefox: 1 week (peaks ~87% every 4 weeks).
  • Edge: 1 week (peaks ~97%, every 6 weeks).
  • Chrome: 2 weeks (peaks ~91%, every 6 weeks).
  • Safari: 1-2 months (peaks ~86%, yearly).
  • Chrome Mobile: 2 weeks (peaks ~80%, every 6 weeks).
  • Mobile Safari: 4 months (peaks ~92%, yearly).

For each browser family I identified the typical adoption “peak”, which is the highest percentage of clients having the same major version of that browser during the last six months. I then measured the time it takes for a given version to reach that peak. To discount noise (such as from early betas and fake user agents) I count from 2% to 90% relative to the browser’s own adoption peak.

Firefox (desktop)

Release cadence: every 4 weeks.
Adoption peak: ~ 87%.
Adoption time: ~ 1 week.

from 1.7% to 78% (2-90% of peak):

  • v85: 26 Jan – 3 Feb.
  • v86: 23 Feb – 2 Mar.
  • v87: 23 Mar – 31 Mar.

Microsoft Edge

Release cadence: every 6 weeks.
Adoption peak: ~ 97%.
Adoption time: ~ 1 week.

from 1.9% to 87% (2-90% of peak):

  • v87: 19 Nov – 29 Nov.
  • v88: 21 Jan – 30 Jan.
  • v89: 4 Mar – 12 Mar.

As of August 2020, Edge aligns its schedule to Chromium releases.

Chrome (desktop)

Release cadence: every 6 weeks.
Adoption peak: ~ 91%.
Adoption time: ~ 2 weeks.

from 1.8% to 82% (2-90% of peak):

  • v86: 7 Oct – 18 Oct.
  • v87: (had a bumpy ride).
  • v88: 20 Jan – Feb 6.
  • v89: 3 Mar – 19 Mar.

Safari (desktop)

Release cadence: every 12 months.
Adoption peak: ~ 86%.
Adoption time: 1-2 months.

from 1.7% to 77% (2-90% of peak):

  • v13: 14 Sep 2019 – 17 Nov 2019.
  • v14: 16 Sep 2020 – 25 Dec 2020.

Chrome Mobile

Release cadence: every 6 weeks.
Adoption peak: ~ 80%.
Adoption time: ~ 2 weeks.

from 1.6% to 72% (2-90% of peak):

  • v86: 7 Oct – 24 Oct.
  • v88: 20 Jan – Feb 3.
  • v89: 3 Mar – 19 Mar.

Mobile Safari (iOS)

Release cadence: every 12 months.
Adoption peak: ~ 92%.
Adoption time: ~ 4 months.

from 1.8% to 82% (2-90% of peak):

  • iOS 13: 9 Sep 2019 – 12 Feb 2020.
  • iOS 14: 16 Sep 2020 – 31 Dec 2020.

See also

You can interact with the adoption graphs on the Navigation Timing by browser dashboard in our public Grafana instance.[1]

Explore the general browser usage and pageview data for yourself, visually:

Or, access the open data in its pure form:


Footnotes:
  1. I used Navigation Timing instead of the dedicated browser usage data, because the browser usage visualisations focussed only on browser family over time, or total usage of a particular browser version during over a date range. While the data is there, there isn’t yet a plot for all of one browser’s major versions over time. In our Grafana dashboard for Navigation Timing data we did have this. The pageview/browser dataset is unsampled, based on aggregate server logs, filtered to only pageviews and non-bots (thus excluding visits to URLs that are not considered pageviews, e.g. when editing articles, or using the account login form, etc.). The Navigation Timing data is randomly 1:1000 sampled, based on any URL where the JS sucessfully loads. ↩︎