For two decades, open source content management systems coded in the scripting language PHP have dominated the website landscape and set the pace for innovation and must-have functionality. Although PHP continues to power half of the internet, there are signs that the market is beginning to look to other frameworks and providers to support the use cases that are most important to businesses today. What are the CMS alternatives that are gaining momentum and how should digital marketers consider these trends in future plans for your website?
This year I attended my first WordCamp US, the largest community gathering for WordPress, the CMS that is used by nearly half of the world’s websites. As someone who has been producing websites since the earliest days of the industry in the mid-90s, I am amazed that a single platform dominates at this level:
The overwhelming market share of WordPress makes it difficult to ascertain trends among the CMS competitors far below. However, in the past year, I believe we are beginning to see shifts in the CMS landscape that will in particular impact the nonprofit industry, where open source predominates today (powering over 62% of sites per Nonprofit Tech for Good) but may not to the same degree in the future. If you are a decision-maker for your organization, I hope this article will prove useful in understanding what is happening in the CMS market that might inform the technology direction for your website.
Is peak PHP behind us?
Let’s start with looking at the background of how WordPress emerged from the pack of competing systems it surpassed to reach the heights we see today.
The earliest content management systems that could support large-scale websites that emerged in the mid-1990s – Vignette, Interwoven, Documentum, FatWire, FutureTense, Inso, and EPiServer – were commercial products, closely tied to commercial web servers licensed by Microsoft, IBM and Oracle. These providers were expensive and ultimately could not meet or control the world’s appetite to build out the internet. By the end of the 90s, developers were creating bespoke solutions coded in the programming languages Perl, Python, ASP.NET, Java and PHP.
Among these, PHP was by far the most accessible, easiest to learn, and proved most adept at web use cases. Used in tandem with three other open source tools – Linux, Apache, MySQL – it was easy, and generally free, to spin up a PHP web server and launch a site. Many web developers in the early 2000s gravitated to this stack (known as LAMP), and soon a plethora of CMS platforms built in PHP emerged from the open source community: OpenCMS, PHP-Nuke, Mambo, WordPress, Drupal, Plone, Joomla, Alfresco and TYPO3 among dozens of others.
From this group, WordPress, Drupal and Joomla emerged as the pacesetters, and in the 2010s rushed ahead of the proprietary CMS platforms to lead the industry. But now, after a 20 year run, could we be entering the end of this golden era of PHP-based platforms? For Joomla and Drupal, the trendlines are now pointing firmly in the wrong direction:
Does that slight dip on the far right for market share among the top 1M websites signal the beginning of a similar decline for WordPress? What might cause these platforms to fade? Below are some forces I believe are in play.
Communities eventually run out of steam
Open source software relies on a passionate community of developers – usually working for free – to plan the roadmap, create new features, fix bugs, evangelize, and recruit new contributors. But as projects lose momentum, the developers who are the lifeblood of the software look elsewhere to build their careers.
It’s the same challenge that many nonprofits face when their donor base ages – if younger donors don’t join the movement, revenue dries up. The big open source projects of the 2000s might have hit this moment.
Joomla appears nearest to the end. From 2015 to 2019 the Joomla community was wracked with mounting dissension (detailed on Wikipedia). The pandemic took a further heavy toll on the social fabric that inspires participation. Joomla 4 was released in 2021, but 82.4% of Joomla sites remain stuck on Joomla 3, which was originally released in 2012. It seems likely that the great majority of those sites will move on to another CMS rather than upgrade. The customer base has lost confidence, evidenced by the steep downward slope in the chart above. This matches our experience at Blue State – since 2013, we have not seen a single RFP requesting a build in Joomla, but have seen a few eager to move off of Joomla.
Existential anxiety is beginning to creep into the Drupal community as well. While Drupal bowed out of competition several years ago for the small and mid-market tiers (where WordPress particularly dominates) it remains a strong player at the enterprise level – Drupal serves 7% of the 10,000 highest traffic websites, #4 on the list of CMS platforms. But sustaining that share will rely on replenishing the ranks of developers to contribute code and carry the project forward. As has been stressed regularly in the annual Drupal Business Survey, Drupal is a development-intensive platform, and competition for qualified talent is an ongoing challenge that limits growth. Many Drupalists who joined the movement in the 2000s are now aging out of their coding days. Attendance at Drupalcon, Drupal’s annual U.S. conference for contribution and social connection, has fallen from 2,500-3,000 attendees pre-pandemic to 1200-1500 attendees in 2022 and 2023. Is that an indicator of a shrinking developer community?
At WordCamp US this year, in the keynote Q&A with head honchos Matt Mullenweg and Josepha Haden Chomphosy, the question was asked: what might pose an existential threat to WordPress? Their answer was simply “Ourselves.” If the community doesn’t maintain its cadre of contributors to keep pushing the product forward, WordPress will decline. Although the 2023 national WordCamp was the highest attended ever with around 2,000 attendees, there is, in fact, some indication that in the U.S. a decline in participation has started: around half of the network of local WordPress meetups that previously existed in the US have not restarted following the pandemic. (However, I also heard from Wordcamp organizers that lower participation in the US has been offset by a boom in Europe and Latin America.)
WordPress remains spectacularly dominant in 2023, but its resilience relies on the continued enthusiasm of its developer community. What has happened with Joomla can happen to WordPress as well. If the community splits into camps with opposing views on the direction of the project (here’s an example of such rumblings) expect developers to seek out alternatives that suit their visions.
Younger developers are just not excited about websites
Related to the above, I have a sense that younger developers are drifting away from CMS work in general. In the early 2000s, joining a movement seeking to innovate systems to make website creation easier, scalable and flexible to meet new uses – while at the same time innovating systems to allow large communities of developers to collaborate together on complex projects – was heady and exciting stuff. These communities felt that they were building the future.
Now, 20 years later, we are living in that future and all of these systems for collaborating to create websites exist. Young developers want to build their careers working on new tech that will shape the next 20 years. They don’t want to sign on to work on the margins of legacy products – even if there is paying CMS-related work available.
This may be a provocative assertion, but it is based on conversations we’re having with colleagues in the field. For example, we are hearing from our partners that their junior devs view working on WordPress websites as training work; they are willing to do it for a year or so, but then they want to be assigned to meatier, more modern projects. But is there any data to back up this impression?
This year Stackoverflow added a new section to their annual developer survey to forecast technology trends. By measuring growth in admiration for certain technologies Stackoverflow seeks to provide insight into what languages and frameworks will have staying power and which will struggle to generate coveted evangelists to convert new users that will stick around. Looking specifically at web frameworks and technologies, the visualization below shows the distance between the proportion of respondents who want to use a technology (“desired”) and the proportion of users that have used the same technology in the past year and want to continue using it (“admired”). Wide distances means that a language or framework generates enthusiasm once you get to know it, and shorter distances means that people using the tool are less excited about it.
This chart has sobering implications for WordPress and Drupal. Both rank under 10% for “desired” and the short bars to “admired” suggest that developers currently working in the frameworks are not hyping them up and getting other developers excited to get involved.
The stability and evolution of these tools currently relies almost entirely on a volunteer workforce. If that volunteer force ebbs away, these projects will be forced to shift to commercial models. We see indicators of this in Drupal today, where for several years, large organizations with vested interests such as Acquia and Pfizer have subsidized developers to contribute. At Drupalcon in Pittsburgh this year, the Drupal Association piloted a grant competition with $98,000 in funding for Drupal innovation projects. WordPress also has a pledge program to encourage business users to subsidize contributors. The question is, are these incentives bringing in new contributors, or simply keeping current contributors in the fold?
From a customer standpoint, if the roadmap for WordPress and Drupal is becoming more dependent on direct financial investment, it seems likely that in the future organizations that use the software will be asked to shoulder the ongoing costs in some form.
The effort – and cost – of ongoing maintenance is not negligible
While open source software remains “free” in the sense that everyone can download, install and run it without charge, it soon becomes apparent to organizations that operate a site running on WordPress or Drupal that the health and stability of the CMS requires ongoing developer attention that drives up the cost of ownership and can limit business flexibility. The challenges to navigate include:
- Security updates. Precisely because open source software is open, bad actors relentlessly seek out vulnerabilities and exploits. WordPress may be the most hacked software in the world, and it requires regular – sometimes weekly – security updates to keep running safely.
- Major version updates. An open source CMS lives in an ecosystem of other open source software, each of which evolves over time. Every few years, software on which the CMS depends has changed to a degree that the CMS software needs to be significantly updated as well to remain compatible. To keep pace with these major updates, many of the contributed plugins and modules commonly used by websites also need to be updated to work with the new CMS version, and in the worst case custom code that was created based on those plugins and modules for a specific website will need to be refactored to continue working. This can add up to a lot of effort – sometimes equal to creating a new website from scratch – which an organization may not have planned for in advance.
- Lack of market support for site maintenance. It is unfortunately common for organizations to invest in a WordPress or Drupal site that they later discover has limitations, is fragile, or is time consuming to maintain. They may not have forecast – or budgeted – for the level of ongoing upkeep. It is also often the case that agencies that design and build sites are often not set up well to provide reliable maintenance, because this work is sporadic and difficult to service profitably.
So, an open source CMS can be more affordable at the start but you need a plan and budget for ongoing maintenance, enhancements and upgrades. Clients don’t enjoy paying for this work, and it’s also not fun for developers, who prefer to be creating new sites than fixing older ones. Managed hosting providers like Pantheon, WP Engine and Acquia offer to automate some of the busywork aspects of code maintenance and security. Managed hosting also standardizes the development process and makes it easier to switch developers in the future. But again, this is a cost that may not have been anticipated.
It’s too much effort to create the functions that are most important to organizations today
The great promise of an open source platform is that it is not a black box and you are not entirely dependent on a commercial provider’s APIs and the pace of their feature roadmap (we see you Blackbaud customers!). Open source frameworks are transparent and fully accessible to developers to build upon in whatever direction you need to take it. That freedom ushered the massive wave of adoption a decade ago, as the proprietary tools struggled to provide support at the pace demanded by the market for new, must-have best practices such as responsive layouts for mobile, accessibility and component-based page building.
But now there is greater parity across CMS platforms for all of these table stakes publishing features. The features that organizations most want right now are ones that provide interoperability of the website with other platforms in the martech stack. This is where the open source platforms are acutely lagging behind. Creating those connections is possible, but it takes work, which means investment, time, debugging and maintenance. Proprietary platforms that offer readymade, turnkey features for A/B testing, localization, personalization, first party data collection, and data sharing with other systems are getting a second look. At Blue State, we are now starting to see RFPs with an appetite for proprietary content management systems that were entirely off the radar five years ago.
As we’ve written about recently, in the current cycle, nonprofit revenues are down, and organizations are making tough choices between platforms in the marketing stack when planning budgets. It takes custom work to integrate WordPress and Drupal websites with the CRM and reporting systems (and then to maintain those integrations) and we are not seeing that type of work getting funded.
So despite the unabated dominance of WordPress across the web today, we think there are some concerning signals about both the future strength of the contributor community and changing business needs that are opening doors to other frameworks. In Part 2, we’ll take a look at a few of those.