August 26, 2021

Long Live Content Cloud

How we’re not going down the same road as Sitecore

“Content management” is a misnomer, and it has been for years.

It made sense back in the day when all we did was actually manage content. Decoupled systems meant that your content management system and your content delivery system were two different things. You managed content, then you kind of threw it over the wall to some rudimentary delivery framework.

We didn’t see the advent of the modern, coupled CMS until the early 2000s. They were more accurately “content management and delivery” systems, but that’s ungainly, so here we are.

What’s the dividing line? It’s vague, but I’ve tended to split it at the publish button – anything you do to the “left” of publish is management, and anything to the “right” of publish is delivery.

It’s fair to say that most of the recent changes in CMS have happened in delivery. This is natural since we had to solve management challenges first. And management is less volatile than delivery. While management is a process of refinement, evolution, and adjustment; the delivery side is in constant upheaval – it keeps reinventing itself.

This means that systems which are poorly architected on the management side will suffer when trying to adapt on the delivery side. Systems without a solid data management foundation find it hard to stay on the edge of the delivery curve. The level of delivery-side change magnifies architectural shortcomings, and several vendors have had to scrap their systems and write new ones.

The most recent vendor to capitulate is Sitecore.

In a video last month, they heavily implied they were abandoning their current CMS architecture in favor of a SaaS approach, and this new approach would be based on their StyleLabs acquisition of 2018. They took the underlying object architecture of the StyleLabs DAM and are using that as their go-to CMS going forward. Apparently, this is the only way they feel they can compete in a headless-centric market environment.

To make that more clear: effectively, Sitecore is deprecating and walking away from the much-vaunted CMS they’ve been developing for 20-some-odd years. From someone who has been in this industry for 20 years, that’s pretty wild to see.

It will take some time to see how the market reacts, but an early reaction from Tony Byrne might be prescient.

Right off the bat a Sitecore marketer pretty much throws in the towel […] it’s rare to see a vendor provide a conceptual demolition of their own, flagship platform

[…] the biggest piece of [the new, proposed] stack – a foundational, SaaS-based CMS component – does not yet exist. Building one to match the complex use cases that Sitecore has targeted will be a journey measured in years, not quarters.

I very much agree with his last statement.

StyleLabs was an appropriate acquisition for them , but content management is a toolset and practice that’s been refined over decades. Resetting and bootstrapping it from scratch is not a trivial endeavor. I suspect Sitecore will be re-solving fundamental problems and features for a long time before they can push forward into innovation again.

To that point, let us state something for the record –

Content Cloud is both Optimizely’s present and future. And the reason is simple: it doesn’t need to be scrapped and re-architected.

At the risk of being snarky, I’m sorry that Sitecore felt their CMS wasn’t good enough for the future. Rest assured we don’t feel that way about ours.

I’ve been working with our product for 13 years, both as a service partner and now a member of the product team, and it never fails to impress me with its underlying elegance. Content in Content Cloud is stored and manipulated as pure data. The only reason content objects appear as web pages or embeddable elements is because their design implementation and the resulting UI makes them function that way. Our foundations are incredibly solid, presentation-agnostic, and they don’t need to change.

We are continuing to invest heavily in Content Cloud. By the end of this year, we’ll be releasing a fully compatible .NET 5 version, with some astonishing performance improvements. This has been a mammoth project for us, but one that positions the platform for an amazing future.

(Some proof: Quan Mai just posted screencaps of B2C Commerce running on a Linux VM.)

Our next major release will be our GraphQL engine, which is now in beta. And we’re increasing the scale of our headless APIs with a new content modeling API (and UI, coming with the .NET 5 release), more features in our content modeling UI, and a new content-focused editing experience that will effectively embed the same featureset that pure-play headless vendors offer inside a more fully-feature CMS platform.

I’ve written before about how Content Cloud already does headless well (and has always done this):

Improving on this on top of our current product means no major re-implementation for our customers. Even as we continue to move forward into cloud delivery, our commitment is that customers will be able to move forward at a pace and path that makes sense for their business.

Let’s be honest for a second –

Given the direction of the market, did we take a long look at Content Cloud and ask ourselves if we should tear it down and start over? Sure, we did.

But we came to the simple conclusion that if we spent years re-doing all this, the underlying framework of the resulting product would probably be awfully similar to what we have today. We trust our system to adapt to whatever is next.

Content Cloud is a strong, stable platform that our customers can count on now and long into the future.