Blogs

Beware of CMS Fads

John Michael Thomas

As the resident dotCurmudgeon™, I feel it’s my duty to bring up instances where hype and promises are at risk of failing to meet (possibly over-inflated) expectations. And as someone who's been working deeply with technology for (cough, cough - transcript unintelligible) decades, I've seen a lot of tech fads come and go.

Sometimes, these tech fads are global, like (dare I say it) AI. And sometimes, they’re specific to industries, like (yes, I will say it) headless CMS.

But to be clear, I’m not saying that these technologies are bad or wrong in any way. Both AI and headless CMS are powerful technologies that solve certain problems very well. In fact, dotCMS has great (IMnsHO) implementations of both.

What I’m saying is that AI and headless have both become technologies that organizations rush headlong into adopting without thinking it through, often at their peril.

Let me explain.

Key Takeaways:

  • When you’re shopping for any tech (including a CMS), look for solutions, not ideology.

  • Don’t get too attached to the specific technology.

    • If you’re not a developer, ask both the vendor and your own technical team how the proposed solution will solve your problems without creating new ones.

    • If you are a developer, don’t get starry-eyed. All technologies have trade-offs, so make sure you know what they are before you fall in love with one.

  • Beware of hype and fads

    • Even for us developers, it’s hard to separate the hype from the results until after we’ve implemented it (and maybe failed) once or twice.

    • So, newness and “hotness” should be reasons for caution, not enthusiasm.

What is a technology fad?

What makes a technology fad isn’t that no one uses the new technology or trend. In fact, I’ve never seen a tech fad that didn’t start with a solution that actually worked very well for some use cases.

What turns a good solution into a fad is when everyone tries to use it for everything - even things it was never intended for. A technology’s status as a fad becomes firmly entrenched when lots of organizations throw gobs of time and money at it before they even try to figure out if it makes sense for them, let alone if it actually lives up to the utopian promises.

That's time and money that could be better spent understanding the problem and finding a real solution or even just making some targeted tweaks and improvements to what already mostly works.

Let’s look at a few examples of CMS fads in various stages of the hype cycle.

A “Retired” CMS fad: Hyper-personalization

In the days of yore, there was a huge push in the CMS space for personalization. Personalize everything, and your engagement (and revenue) will skyrocket!

dotCMS implemented several powerful personalization features, such as Personas, customizable Rules, and automatic personalized content pulls, to great fanfare. I personally gave a presentation on them at the 2016 dotCMS User’s Conference, and customers were excited about it.

But then…only a fraction of our customers actually implemented any of it.

The problem wasn’t that our personalization features didn’t work; they did, and they still do (some of our customers use them heavily). And it wasn’t that most companies didn’t want to use personalization; in fact, since it was the buzzword of the day, many organizations spent plenty of time and energy trying to implement it before they even asked the question of whether or not they should.

The problem was that once companies got into the trenches, they realized that, even with great personalization tools, implementing and maintaining personalization is a lot of work. And for many organizations - especially those that don’t run online stores - the benefit of personalization wasn’t (and still isn’t) clear. As a result, many organizations found that the ROI just wasn’t there.

So, the fad came, time and energy were spent on scuttled implementations, and the fad went. What remains is a tool—personalization—that level-headed organizations can now implement where it’s needed, when it’s needed, and ignore when it doesn’t help.

In other words, it’s graduated from a shiny fad to a solid business tool that has both benefits and trade-offs. More of these, please.

A Dying CMS fad: “Pure” Headless

Several years ago, as headless continued to gain steam, I said to myself, "I don't see how this is going to work for 90% of organizations.” I was one of the dotCMS trainers for years, and the number of customers I trained who said, “We want to have to ask developers to make changes for us,” was precisely zero.

I said as much out loud, but since headless was the buzzword of the day, it didn’t really matter who was right or wrong - we had to tell the story that our prospects and customers wanted to hear. And since dotCMS was built with headless APIs and technology agnosticism from day one (before headless was cool), that was a story we could tell with the best of them.

But my opinion hardened as we had customers come to us very explicitly saying they want to enable more content editor self-help. And let’s be honest with ourselves: Editor self-help is the direct antithesis of pure headless; how can you allow editors to help themselves when it takes a code change to update the layout of a page?

So, when the industry started “moving to the middle,” as many in the industry have noted recently (and as our VP of Product Preston So predicted many moons ago), my shocked face was suspiciously absent.

Let’s be clear that what turned headless into a fad wasn’t the idea that content should be made available to headless apps. That idea was and still is spot on target. What turned headless into a fad was the idea that customers should throw away their existing systems and use only headless apps. In other words, once the focus became “pure headless,” it became a fad.

Now, I’m not saying pure headless isn't good for anyone. I'm saying it isn't good for everyone. Pure headless makes a ton of sense for a limited number of organizations.

But like most technology fads, that bandwagon got jumped on so hard that oodles of organizations that weren’t a fit for it spent tons of time and resources trying to implement and maintain it. Unfortunately, many of them not only had to backtrack but also alienated their content teams in the process.

So, instead of asking "when can we go headless," maybe what organizations should be asking is, "Should we go headless?" And if the answer is yes, then maybe the next question should be, "Can we go headless without trashing everything we already have, being in limbo for years, and spending as much as the GDP of a small nation-state to get there?"

If the answer is yes, then congratulations; you’re a good fit for pure headless. If the answer is no, then you might want to get off the fad bandwagon and look for a platform that will help solve your real problems instead of chasing a buzzword that promises the world.

An Up-and-Coming CMS fad: AI

I’m calling it now, so let the flames begin: AI in CMS is a technology fad. (I’m the product manager for dotAI, so I might not be completely clueless here.)

I probably shouldn’t have to say this again, but since I’m sure I’ll be lambasted regardless, here we go: I’m not saying AI isn’t great for some customers and use cases because it absolutely is. What I’m saying is that organizations are jumping into AI and using it for things it really isn’t a good fit for, just to keep up with the Joneses.

(For more on this, see my upcoming article on AI Headwinds in the CMS Space).

The first sign of it being a fad was how fast most vendors released AI features. Many of them were basically eye candy, designed more to show you that the vendor wasn’t falling behind than to actually deliver value to their customers.

(Honestly, I wonder what deeper work had to screech to a halt to redirect the developers to whip out eye candy that fast. But I digress.)

The second sign was in how companies all over the world started touting their AI creds, as if that was all that mattered. Forget the fact that some of these companies were decades old (or older) and had billions of dollars in well-entrenched business; it was as if you didn’t utter the magic buzzword, your business was done for.

(This phase is now over. The mention of AI in earnings calls has dropped lower than it was in Q1 2023, and even the New York Times is calling for a pause on the AI hype cycle).

There were other signs, but I’m boring you already.

Yes, AI will make some common CMS tasks much easier and simplify the workflows for many content workers. But it’s still being tasked with solving problems it’s wholly unsuited for—like writing entire articles from scratch without human review or providing authoritative (and legally binding) answers on financially sensitive and liability-breeding issues.

In time, there will be weeping and gnashing of teeth (just ask Air Canada). And after that, we’ll find the balance where AI becomes the best tool there is for some tasks, and is rightly avoided for others. In other words, it will eventually graduate into a solid business tool.

(In a future article, I’ll discuss how the new fad of AI is set to go retro and fulfill the promises of an older fad that didn’t quite live up to its hype at the time. But that can wait for another day.)

Closing

Yes, I know, “It’s about time.” So, I’ll keep this part short.

Technology fads come and go, and the CMS space is no exception. But it’s not healthy to let fads make your business decisions for you.

As CMS vendors, we need to focus on the value technologies like headless and AI can deliver to our users, not on playing buzzword bingo or puffing up our chests to show that we know how to play with the latest shiny thing. We need to build features for all our users— from content editors in the trenches daily to developers building sophisticated apps to business users rightly worried about data privacy and liability.

What, then, should you look for in a CMS?

  • Solutions, not ideology.

  • Tools, not toys.

Keep watching this space to see dotCMS deliver.

John Michael Thomas
Product Manager
June 10, 2024

Recommended Reading

24.04.24 LTS: Upgrade to the Universe

Now available: dotCMS 24.04.24 LTS, the first long-term supported release of 2024

Getting started with the dotCMS Universal Visual Editor with Next.js

In this blog post, I’ll demonstrate how you can get started with the Universal Visual Editor and Next.js in just fifteen minutes.