I was reading Hendrik Luehrsen’s excellent post “WordPress isn’t WordPress anymore“, and I decided I had to write more about this. I recently spoke at WordCamp NL about “The missing features of WordPress”, and these two things “touch”, in an important way. I love WordPress. I love WordPress plugins. I don’t love some of the recent developments in WordPress and that’s not just talking about the recent Automattic – WP Engine drama.
When I started writing this, there had been no decision on the injunction yet. Now there is, and as a community we have to figure out how we deal with all this. I don’t think we can separate that entirely from looking at what we all want from (and for) WordPress. Good discussions about where WordPress should be headed have been lacking from the community for a while.
So: let me share my thoughts, first by looking at WordPress’s market share and the trends in the wider market, then by discussing what WordPress is, or should be.
Looking at WordPress’s market share
If we look at WordPress’s market share, we have to admit that it’s stable at best. WordPress, which has been growing for the longest time since it was created, is now stagnant. The current market share (as of November 2024) is slightly below what we started the year at:
This means there’s also hardly any growth since early 2023, which should be (and probably is to everyone who’s aware) concerning to everyone in the WordPress industry.
To make things worse, all other open source, self-hosted technologies are losing:
If you’re now thinking “but then other projects must be winning”, you’re 100% right: SaaS tools are winning. Just look at the numbers:
I’ll give my sharper analysis: if WordPress hadn’t had Elementor, it would have shrunk (something I’ve been saying for a while). This is definitely also true, though with less big of an impact, for WooCommerce, the other driver for WordPress growth. Which shows that it’s not that people don’t want Open Source. Let me show this with a graph too:
So, it’s not that people dislike or don’t want Open Source. I honestly think people don’t care that much about this at all. I think we have other problems. Let’s dive deeper.
Why is this happening?
The question we have to ask ourselves of course is: why is this happening? What are people and companies looking for that WordPress is not providing? I think there are several important factors to discuss around this. Many of the arguments that people use against WordPress, we are very often quick to dismiss: WordPress is old, not modern. WordPress is insecure. WordPress is slow. There are certainly more pain points, but let’s try to cover these for now.
WordPress is slow & technically stagnant
WordPress is often ridiculed by developers in the wider web ecosystem. It’s considered backward, slow and bloated. And the truth is harsh but simple: this is true. We have failed to modernize large parts of our codebase over the last decade, especially on the PHP side. Many projects and potential improvements have been discussed, but we focused instead on building features (like the duotone filters Hendrik mentioned). Fantastic work has been done to keep WordPress working with modern versions of PHP, but we’ve failed to truly deal with our technical debt and with that also failed to make WordPress that much faster.
On the JavaScript side, we’re doing the opposite: we’re churning through code harder than we should. The speed of development is outpacing what even some of the best developers can follow if they’re not on the project full time. We bring new JavaScript APIs very often, developing our own instead of integrating existing projects.
WordPress is also at the absolute forefront of some of the web’s development though. The Performance team includes some of the world’s very best frontend performance engineers, who jointly made WordPress much, much faster over the last few years. The problem is: our competitors are doing the same. Shopify improved more than we did. Part of the problem here is that we’re simply not merging the fantastic work the Performance team is doing into WordPress core fast enough.
WordPress is insecure
WordPress is very often dubbed insecure by its competitors. The standard response to this is “WordPress core has hardly had any security issues the last few years”. We have to be honest here: WordPress is often considered insecure because it gets hacked a lot. It doesn’t get hacked through WordPress core, it gets hacked through plugins most of the time. But we don’t have the luxury of saying that those plugins are not WordPress. They very much are. Plugins are, as Hendrik so poignantly discussed in his post too, the reason for WordPress’s popularity, this easy extensibility is also its challenge. You can’t look away from their shortcomings yet benefit from what they do.
Recently PatchStack (disclosure: we’re an investor as Emilia Capital and I’m on their board) held a special event in their Bug Bounty program for Cyber Security Month and they closed 1,000 plugins on WordPress.org as a result. Now that is a herculean effort, but it also leaves everyone wondering: what would happen if we looked at all plugins a bit more deeply?
It also makes you wonder: why can’t we see in our WordPress admins that those plugins are closed on WordPress.org? Why is WordPress, the software, not telling us that plugins we’re using that we installed from WordPress.org, have been closed for security reasons? The ticket requesting this is now officially over a decade old.
A step further would even be to say: this set of plugins hasn’t been updated in the last year. The plugin directory talks about 59,000 plugins. The XML sitemap for the plugin directory tells me that only 14,648 of them have been updated in the last year or so.
WordPress is too hard to use
The last, most painful argument that I want to cover is that WordPress is too hard to use. Currently, WordPress forces users to learn several different admin UIs, with different design languages. WordPress has for years tried to get people to use the Block Editor and more recently the Full Site Editor, and the harsh truth is that the editor that has grown the most in that time is a third: Elementor. See for yourself:
Note that the Block Editor was released December 6th, 2018, now 6 years ago, but wasn’t tracked in this data until earlier this year, and that the drop in Elementor’s growth is the result of a measurement change.
One of the best additions to WordPress’s marketing recently has been the flurry of activity on the WordPress Youtube channel mostly by Jamie Marsland. I have to admit to have been utterly dismayed when I saw Matt Mullenweg battle Jessica Lyschik recently, because it showed so clearly what many people have been saying for a while now: the block editor is too hard to use. Matt resorted to use AI to help him create patterns but that didn’t help him all that much. If Matt, who leads this project and should be pretty well versed in it, can’t figure this stuff out faster, shouldn’t we take that as a sign?
So what to do?
Instead of dismissing this feedback, we should take it head on. If the WordPress performance team has shown us one thing, it’s that we can improve. Now, we must. But to do that, we must solve our existential crisis first:
What is WordPress?
Hendrik in his post, underneath a heading asking “Who is WordPress for” says this:
Historically, WordPress didn’t try to be everything for everyone. Its lean core and robust ecosystem allowed it to meet a wide range of needs without dictating how it should be used. Developers could tailor it to clients. Hobbyists could start small and scale up. Agencies could rely on its flexibility to create bespoke solutions. WordPress provided the tools, but the user always defined the experience.
The “user” over the years have ranged from individuals, small & large businesses up to large enterprises. The power of WordPress is exactly that: it is what you want (and dare to dream) it to be. The problem is that the last few years, new features that went into core didn’t always seem to fit that “lean core” idea.
Of course there are features the enterprise would want, like personalization, logging, A/B testing, etc. There are features that not just enterprise but more sites would like to see, like multilingual support, custom fields, better admin UIs and many, many things. Most of those things do not need to be in core. Plugins have always been a source of innovation and they should continue to be just that.
WordPress core needs to provide the underlying APIs that allow for extension and force some regularity in how we do these things. I was a big proponent of the REST API when it was developed. In the same way, I’m now a big proponent of merging the fantastic work Felix has done on the AI services plugin into core, or at least make that a canonical plugin. I think multilingual support should definitely be built into core, at the very least the underlying APIs.
We also need a new admin UI and a design language that we force on plugins (and themes) a bit harder. Plugins simply should not have (to develop) their own admin UI, a point I’ve been stressing for years. More broadly speaking, we need to work on UX everywhere and make more logical decisions. For example: “Edit site” does not need to be a button in my admin bar. Nobody edits their site’s templates every day, and nor should they.
I’d love to have more of these thoughtful conversations, where we all openly and freely discuss where WordPress should go. So: let’s see your blog posts, outlining your wishes for the project!
Update Dec 20th: I followed up this post with another one with a vision for a new WordPress era.
Here are my wishes, as penned in September.
As a Marketer… my frustrations extend to the Marketing side of things.
Main points:
– WordPress is for developers (and they make it for everyone else)
– Reposition WordPress as the ‘Operating System for the free web’
– Marketing WordPress should take a leaf out of political campaigns
– 0.75% for the Future <- Marketing should get more in terms of contributions / resources
https://wpbiz.dev/community/marketing-wordpress-future/
Great commentary once again.
Sadly, probably one of over a hundred great summaries pointing at the elephants in the room.
If it’s obvious to half the community, leadership failing to recognize any of that is a clear signal we live in completely different realms.
I was surprised to see Matt take on the challenge in the first place. You don’t have to know what he does on a day to day to say 100 percent that he’s only used Gutenberg as a replacement of TinyMCE.
Spending an hour with a random site builder like Webflow, Framework, Wix, Squarespace, even Shopify, and starting with a fast, beautiful template, not bothering with updates or hosting, should be obvious when it comes to comparing the experience today.
At some point we should also try and calculate market share in economic value or size or something more nuanced than mere number of installations. One thousand little blogs might be less impactful than one big enterprise portal.
I feel like the idea behind Core Fields could be one pillar of the extensibility you mentioned, which could potentially provide a plugin’s settings page automatically without having to code the UI for it. Unfortunately, since Scott Kingsley Clark stepped away, I’m not sure who would be willing to lead that charge after Scott tried for years.
Great article. One point of significant contention though – the Performance Team (who, indeed, have done great work) are also mired in bureaucracy and complacency. Many substantive proposals have been completely ignored, or even outright dismissed.
The real issue is that there’s really no one steering the ship at WordPress. Matt is quite obviously not competent or well-intentioned, and then in the (well-intentioned, but ultimately short-sighted and destructive) name of “inclusivity”, there’s just too many unempowered people trying to design things by ad hoc committee. There needs to be competent people in charge of each area, whose Yes and No carry authority.
And then, of course, there needs to be infinitely (because dividing by 0 = infinite) more accountability throughout the project – something that can only be achieved through real governance.
None of this will ever happen with Matt and his Audrey Capital cronies in complete control. So, any discussion of improving WordPress needs to center on that. Anything else is just a waste of time.
t’s great to finally see some big names in the WordPress world admitting what’s been obvious to most users outside the WordPress bubble. Having a modern post editor is definitely a step forward (I moved three of my sites to Ghost, and it’s been awesome!), but using the same editor for creating pages? That’s a hard no.
What we really need isn’t just more features—we need a fresh starting point. Here’s what I mean:
Consistency in the Look and Feel: The user interface should be clear and consistent, so we’re not hunting around trying to figure out what’s what.
Better User Experience (UX): Make it easier and smoother to use, plain and simple.
Fewer Blocks in the Core: Get rid of all those blocks most of us only use once. Keep the ones we actually need for writing and creating content.
Speed and Security: Make WordPress faster and safer—those should be top priorities.
Also, there are a few things that really should be built into WordPress itself, not left to plugins:
Native Multilanguage Support: We shouldn’t need a plugin to create multilingual sites—it’s 2024, people!
Built-in Privacy Tools: WordPress should help us comply with GDPR and COPPA without adding extra hassle.
Custom Post Types and Fields: These are essential for making WordPress truly flexible and powerful.
The bottom line? We don’t need more bells and whistles. What we need is a WordPress that works better, feels better, and sets us up for success.
I think you’ll have heard me admit to many of these things over the years, I just haven’t blogged as much 🙂
WordPress has many limiting factors that are hard to change and it’s too late now like the limited database structure, lacking native multi language, react hype introduced in core with gutenberg that added complexity and scared beginners away etc
Now with many bloated freemium plugins and adware makes the WordPress experience slow and annoying.
I think that the only hope in the open source cms arena are new projects like Vvveb cms, old projects like Drupal and Joomla while less flexible than Wordpres are becoming even more complex to develop for medium and beginner developers, and the myriad of flat file, half cooked laravel cmses, headless js or static generators are toys for hobbyists compared with the power WordPress gives to the user and developers.
I couldn’t agree more, Joost. New WordPress might need a group of developers who are trained in and use a modern codebase. We’re in a world of Teslas, and we have a classic car that we keep repairing with obsolete parts. Time for a fresh start.
My other wishlist is founded on marketing principles. Websites haven’t been our business cards since the late 1990s. They are primarily used for marketing, yet very few marketing capabilities (aside from Permalinks and RSS) are baked in. At a minimum, we need HTML titles and meta descriptions, if not crawling and indexing options.
Thanks for putting this together. I’m definitely going to share this with our team and clients!
This is awesome Joost. I’m the developer the WLP (WhitelabelPress) fork at the moment, and lots of what you say resonates.
Here are some solutions that are already part of the upcoming V2 release that connect with what you talk about:
– For the admin UI, we default to dark mode, total black background with more padding, giving less need for developing complicating interfaces.
– For the leaner core and AI services, in WLP V2 the main features like widgets, rest API are now a set of (optional) core plugins, that are loaded in wp-settings and can be enabled/disables or simply removed.
– For the plugin security a new (open-source) repository is being made, decentralized, where the current idea is to use a few hand selected plugins in “stable” and put the rest in a “hatchery” directory where users will be made aware of the risks, and if new packages prove to be valueable they can be transfered to the stable.
I also like your idea of security warning about removed unsafe plugins, and I’d love to see Yoast in the first Decentralized repo, if not compatible I can try to fork Yoast SEO for WLP and then you can take over later. ✨
I’m not in charge at or even “related” to Yoast anymore, so I can’t help you on that last bit 😉
Ah ok, thanks for letting me know! I read your about + other blogs and understand it’s now owned by Newfold.
Perhaps it’s a good time for a Web3-DIDs alternative anyway in the long-term 🙂 (the trend that has been fueled by BlueSky + supported by the upcoming WLP release)
Great points.
For me think this recent debacle has just made it more obvious how WordPress’s direction has been driven by Matt for the benefit of Matt over the last few years.
Whether that be pushing Gutenberg to allow wordpress.com to compete with other page builders, features being spun out into paid services ( Jetpack / VideoPress / Vaultpress ) or long requested functions like custom fields not making it into core without blessing from above.
The UI revamp from Gutenberg should have been a wholistic UI framework for the admin area, made as an api for plugin developers to consistently implement.
The shakedown on WP Engine should have been a plugin marketplace on .org with paid plugins available to license ( and the 100s of freemium admin upsell popups put to bed with it ). With a revenue share back to the foundation to fund development and cover the costs of running .org, not automattic.
Would love WordPress to continue growing, but for that to happen I think there needs to be change at the top. An independent board, clarity on the legal structure of .org and the foundation, actual product leadership ( in terms of design and direction ) coupled with a public roadmap that reflects what users want.
For the moment i’ve taken Kirby CMS for a spin – it’s fast out the box, totally customisable and has been a joy to work with a modern, logical and consistent API.
> To make things worse, all other open source, self-hosted technologies are losing
The issue with that statement is that static site generators are growing in popularity, and there’s no 100% accurate way to tell that a site is generated using one (since it can just be plain HTML with no defining characteristics). For sites that are mostly static content, there’s limited extra value in having a highly dynamic system, and deploying static HTML has a bunch of advantages (no possibility for security holes, faster than dynamically-generated pages, can easily push the files to a CDN, etc)
I think we can work on recognizing those, too… And I think their rise is a thing, but it’s not that big. It’s very much a “developer thing”.
Thanks for this post Joost. It was needed. I’m pleased Daniel mentioned the growth SSGs.
I followed many web standards based developers toward WordPress and again away from it and on to static first (Jamstack) approaches. Having used WP since 2007 I would still like to see it survive and use it, but it gets harder all the time.
WP’s new JS first approach speeded this developer shift away and put it at odds with the W3C. A decision forced on all by one who could have respected all the community and made it an option.
Page builders have hidden WP’s decline by luring a new audience of no-coders (and the false idea that a website is only what YOU see).
WP keeps making a deal out of the fear that if you don’t support it you are advancing proprietary software. Drupal may get a mention, but never the Jamstack open sources projects.
The 2024 Web Almanac shows 67% growth in static and hybrid website architectures among the top 10,000 most-visited sites.
I think WP needs to be honest with itself about the rise of static and focus on the gap it can fill rather than ignore it. Accept that it can not be everything to everyone.
Isn’t it time for a fork of WordPress with Joost as the inspiring leader? That would really be a game changer and very inspiring.
Joost
Everything you say is technically true. Since the project is mostly run by volunteers you get the governance model that you do.
The biggest thing is not a technical thing at all. The project is mostly managed and driven forward by people who get paid to maintain the project. And in terms of governance A8C clearly runs this project and will continue for the foreseeable future.
When the project has a completely new governance model managed by the community entirely with many more contributors who get paid for their work then we will have a new WordPress. Until then nothing will change.
This is actually what Matt wants from WPEngine. And the response has been : take a long walk off a short pier.
Why won’t large companies who generate revenue from WordPress not insist on a new governance model that includes them as part of this project? I still don’t understand why WPEngine can’t or won’t contribute much more to the project. Is the A8C culture that resistant to outside suggestions and contributions ?
I’m afraid the answer to that last question is the basis of a lot of the current problem… Are companies not willing to contribute, or at they not willing to contribute when with that “taxation” comes no “representation”. I touched on that in a recent post.
We are closed in a loop: companies that have the capacity to support a different path may not do so to avoid fragmenting the Community and the project. At the same time they do not want to give money to a direct competitor who completely controls the project route. It seems inescapable that the only possible change is internal, supported by reference players in the Community, but not related to A8C.
As a plugin author for more than a decade now I concur with everything here.
One thing I haven’t seen mentioned, maybe ever, but came to mind reading this and the comments was a unified settings page.
Currently Settings is a huge mess on most active sites with any load of plugins.
Each creating their own pages, many of which are complete UI/UX baked into what should be simple settings panels.
I’d argue WP core should be offering a more general settings panel, tabs and sub tabs, with official ways to hook in custom ones. I know the settings APIs exist but your on the hook for custom UX to handle that too.
It should be a standardized but hookable format to bring all “settings” into one place.
Further just brainstorming but there should be a new area of the admin or grouping of menus in ways that allow less used plugin panels away. This currently requires plugins, none of which are really super great to use themselves.
No offense but Yoast SEO comes to mind. I set it and want to forget it til I don’t. But it’s always on the top menu bar along with dozens of others.
Would not hurt to segment the users in the dashboard by purpose and give them better designed experiences for their purpose. Roles can do this but takes a lot of work and there is no standardized way to register menu item for a specific “user purpose”.
Great article. I stopped working on Elementor and WooCommerce related products development for a long time. It looks like I am missing most growing part of WordPress. Thank you for awesome analysis.
Thank you for your article and your thoughts. Direct to the point as really needed to move forward and succeed. Congratulations. I agree with all questions you have pointed along the article.
On the other side, look at the web frameworks today if you need to break out of the ecosystem that limiting you to what you can do and cannot do. Honestly, the technical debts is in our “mind”, we need to get rid of it before we can get rid in the code.
>We also need a new admin UI and a design language that we force on plugins (and themes) a bit harder. Plugins simply should not have (to develop) their own admin UI, a point I’ve been stressing for years.
This +1000
Excellent points. WordPress has consistently been late to the party in the marketing game. Either not participating, being slow to adopt, or doing things that fly in the face logic. Ultimately WordPress is for people who are not developers:
Thoughts:
We give websites ( a new one every year to WordCamps but not meetups). Image the SEO, not to mention saved labor and ease of contacting new folks and keeping them around if there was a consistent website for each community that used WordPress. Plus there would be a ton of saved Meetup.com fees. The new people coming into meetups have zero interest in the complicated way to join the make Slack, they don’t want a gravatar, and can’t grasp why we don’t have a website and newsletter. It’s not 2005 any more. Why are we acting like it still is?
The Learn WordPress crew has made phenomenal strides in the last two years with learning paths and great LMS type courses.This needs a bigger signal boost.
Theme presentation: https://wordpress.org/themes/ Everything older than 2022 save for Astra is about as exciting and visually appealing as wet paint drying
I completely agree with Joost de Valk’s analysis regarding WordPress’s current state. As a developer who works with WordPress daily, I find his points about the stagnant market share and growing competition from SaaS platforms particularly concerning. His observation about the technical debt in PHP and the challenges with multiple admin UIs resonates strongly with my own experiences.
The security issues he highlights with the plugin ecosystem are something I’ve encountered firsthand, and I believe his suggestion for a standardized admin UI design language is crucial for WordPress’s future. Like de Valk, I believe WordPress needs to maintain its lean core while providing better APIs for extension.
His call for more community discussion about WordPress’s direction is exactly what we need right now. The platform needs to evolve while staying true to its open-source roots, and I’m eager to be part of that conversation.
The article was very good, but it seems that the expectations from WordPress are higher than its real place. WordPress has done well in its segment and even beyond its segment and has offered its services to the public without expectations. But the markets have become very specialized and it is no longer possible to advance the business with the approach of ten years ago. You have to be logical and expect from each product to the extent of your ability. I don’t think business owners understand this and it is natural that they quickly look for a suitable alternative or develop the version they need. The current WordPress is also very suitable for building startups and for small businesses. WordPress up to the NVP version is great, but you need to think about the infrastructure and develop it in accordance with the development of the business. For the current version of WordPress, a decisive administrator is needed to prevent useless expansion and to standardize it to some extent so that the general audience does not get confused and confused at first glance and does not get involved in troublesome maintenance during use.
It took me a few years of rebellion before I started using Gutenberg. Now I don’t use any other page builders. BUT I dislike that Gutenberg is gaining complexity and features. I had a client set font sizes on headings. Those are in PIXELS. I use EMs and REMS on my sites. I should be able disable these “inline style” options so I can maintain a cohesive standards based website based on MY CSS rules. The problem with Elementor (and all other Page Builders) is they make sites inconsistent with themselves. Designers have to set every margin / every row, every element instead of making global changes with CSS and the customizer. Democratizing the web is not a reasonable goal unless they start teaching best practices (including Web Content Accessibility Guidelines) to all students. Never going to happen.
From someone who simply blogs, the block editor is not a big blocker. I mean there is a small learning curve, but it’s comparatively easy to get-over with. But when it comes to site editing, it’s hard, very hard.
This is the same reason why wix and other SAAS are gaining. Rather than block only themes it’s still classic+block themes that has more active daily installs. Kadence, Astra and all still provide that customization feature, which helps end-users launch a site fast & easy.
Now going back isn’t an option, nor will multiple forks of WP gonna help. When a major share of plugin devs are adding more block features, over time, it will phase out major WordPress forks that don’t include Gutenberg.”
I think there is a very important thing it should have happened and done which is Multi Language Support at its core. If it had a built in Multi Language Support (Not translation plugins) i think we would have seen more than 80% of the Internet websites uses WordPress.