Joost de Valk sparked alarm yesterday with a provocative blog post: WordPress market share is shrinking. Joost highlighted that W3Techs’ CMS market share data shows WordPress losing 0.4% market share, since February.
We noted this at the top of our post about WordPress search volume trends last week, and didn’t focus on it because we didn’t think it was a big deal. I have a couple of issues with focusing on market share.
At some point we decided that WordPress powering every website on the internet was unequivocally a good thing, and nobody is asking why this is a good idea. We’re unquestioningly taking “more” to equal “better”, with literally no nuance.
The W3Techs methodology is one way of calculating market share, but it’s one of many. It is incredibly hard to understand what CMS every single website on the internet is using. As we’ll discuss, the two mainstream methods have a 10% variance in WordPress’ market share – given that, I don’t see a 0.4% decrease in market share as something to panic about.
Since when have we agreed on market share anyway?
The exact market share % has always been a number which changed according to who was presenting it, depending on how it was calculated. W3Techs takes the top 10 million sites from Alexa and uses that as its basis: the current drop being discussed is based on the same calculations it’s always been, so at face value, irrelevant of the %, the drop is real.
Note, the 0.4% drop is since February 2022. Year-on-year, W3Techs shows WordPress’ market share as up 1.7%.
The HTTP Archive project’s annual Web Almanac however gives an alternative method for calculating market share. I wrote the CMS chapter for 2020, and the 2021 chapter was written by a Software Engineer at Wix. As the author, you’re given the raw data to analyse, so there’s scope for interpretation of how you interpret the data, but you do not influence the data collection.
The HTTP Archive data takes a set of 8,198,531 websites extracted from the Chrome UX report. That’s fewer websites than W3Techs, but both are still huge numbers of websites. HTTP Archive excludes Shopify from its CMS analysis, and I’m unclear on why. This reinforces my point that even with objective stats, subjective decisions get made that change the data.
When we look at the individual CMS market share from HTTP Archive, which is available here, we see WordPress’ CMS market share increasing 0.4% year-on-year:
2019 | 2020 | 2021 | |
WordPress | 74.2% | 75.0% | 75.4% |
Joomla | 5.9% | 5.1% | 4.4% |
Drupal | 5.6% | 4.8% | 3.9% |
Wix | 2.4% | 3.0% | 3.7% |
Squarespace | 1.9% | 2.0% | 2.3% |
When we chart this, you see WordPress’ dominance:
This is annual and we don’t have 2022 yet. The W3Techs data is more up-to-date; it’s updated monthly. But Web Almanac gives WordPress a 10% higher market share than W3Techs.
Yet another source, BuiltWith, gives a completely different set of numbers for the top CMS’ market share relative to each other:
2019 | 2020 | 2021 | 2022 | |
WordPress | 74.6% | 73.1% | 72.0% | 68.2% |
Shopify | 2.1% | 4.0% | 6.6% | 6.2% |
Wix | 10.1% | 9.9% | 10.7% | 15.8% |
Squarespace | 6.0% | 6.9% | 6.3% | 5.6% |
Drupal | 1.8% | 1.6% | 1.3% | 1.2% |
Joomla | 5.4% | 4.4% | 3.1% | 2.9% |
These three methodologies give us 3 different pictures of WordPress’ market share year on year:
- W3Techs: up 1.7% (note Joost looks at the most recent quarter, where there is a decline – but it’s still up YoY)
- Web Almanac: up 0.4%
- BuiltWith: down 3.8%
Which is right? We have no idea. These are comparative figures to all websites with a CMS, so the market share could have shrunk whilst the absolute numbers of sites went up.
This is the case in the BuiltWith data, which lists 30,200,000 sites running WordPress for 2020 and 2021, but more sites using Wix. This is what causes the % decline in WordPress’ market share – but the number of sites has not decreased.
We’ll see what the latest HTTP Archive data is in a couple of months. If WordPress’ market share is down 20%, sure, by all means panic, but I very much doubt that’s going to happen.
Why and where was the market share growing?
WordPress’ market share has grown more or less without fail since 2003. That is a miracle! This has gotten us used to unending growth, but we never really understood why it was growing, and in which areas:
- Were more websites switching to WordPress because it was better?
- Were more websites switching to WordPress because it’s worse, but free?
- Was it because of Gutenberg?
- Was it because of the latest updates to the ecosystem?
- Were more people putting together blogs for fun, or were more eCommerce websites with large footprints switching to WordPress?
We had no idea why the market share was growing, and we accordingly have no idea why it’s shrinking.
WordPress’ market share shrinking throws up a straw man for everyone to tweet their frustrations with the WordPress project. Too many plugins! Not enough plugins! The block editor! The classic editor! We have no idea. Some of these may be true, but we have no insight.
Without clear answers to those questions, we can’t reliably tell where the growth was coming from. And without that information, we equally can’t tell why it’s now declining. It’s therefore irrelevant to try and pinpoint a specific reason for it, and it’s therefore not an opportunity to air grievances with the ecosystem which may or may not be related. The truth is there are simply too many variables at play for any single reason for the slight dip.
Search volumes give us more nuance
Market share is one mechanism for understanding a CMS’ popularity. Pure market share gives us no insight into how those sites are used, though. WordPress could have 70% market share by having 10 million empty websites.
I like search volume data for getting more nuance and insight into what people are searching for. We can see specific terms people are asking Google, and infer from those specific usage. We can also see what advertisers are willing to pay to reach searchers of each term, which is a decent proxy for commercial value.
In our post last week, we used FALCON AI’s database of 35,000 WordPress keywords as the basis for our analysis. I wanted to understand how Wix, Squarespace, and Shopify’s organic searches compare to WordPress.
For this analysis, we took the top 3,000 keywords for each CMS by search volume. We extracted quarterly search volume trends, CPC ranges, and competitive information. As discussed up top: Shopify and Wix are public companies, so they have to disclose to the SEC their marketing spend. Their investor relation documents show them spending hundreds of millions of dollars each quarter on marketing. What does this get them?
Let’s take a look at aggregated search volumes for each CMS’ top 3000 brand-match keywords:
We’re looking at the same quantity for each keyword here, so this makes the average a good comparative popularity metric. Here we see Shopify solidly popular throughout the year. I’m surprised Wix’s average is decreasing despite Wix overall gaining market share. I suspect this reflects that people build simpler sites with Wix.
Or maybe it’s just Wix is easier to use, so you don’t have to Google to figure things out! This reinforces my point: you need to look at many sources, because all methods are flawed.
I’ll share the average CPC data for those keywords because I have it, although I’m not sure how useful this is. This suggests Shopify sites are massively more valuable, which makes sense as they’re just eCommerce sites:
A quick word on Google Trends: Google Trends can be a good source of data, but it uses a relative “interest” metric, rather than absolute search volumes. A higher number indicates higher interest compared to the same term, previously. It does not indicate higher search volumes. Google has an explainer here.
Regardless, I’m not sure this is especially insightful anyway:
WordPress’ market share is obviously finite, and that’s ok
It is obviously impossible for WordPress’ market share to increase in %, infinitely. I’m not sure anyone thinks 100% of websites being built on WordPress is actually a good idea. WordPress is the dominant CMS by a phenomenal margin.
It was inevitable WordPress’ market share would plateau: the total addressable market will always keep growing, but the % share of it simply cannot. The way we’re looking at it, therefore, is that this level (wherever it is) is probably where WordPress’ share will lie for the foreseeable future, and that’s okay, as that still means growth.
What matters more is how the activity within the market is proceeding. What’s going on with demand for WordPress products, and what’s happening with the industry trends such as pricing, branding, mergers and acquisitions? These are clearer signs of what’s happening, and would be more powerful indicators of things going wrong, including where they’d be doing so.
Co-written with James Baldacchino, Ellipsis’ Head of Strategy.