“The era of classic CMS is over” - In conversation with Preston So

Recently we had the chance to have a conversation with Preston So on voice content. Preston has been involved with CMS technology for almost 20 years now, and among many different things, has authored a book on voice content. These days, Preston is mostly thinking about what we are to expect in the future of digital experiences, being a “digital experience futurist” as he calls it.

Our conversation started with some of the key issues facing content management today, then moved on to headless in particular, and finally arrived at voice content. This article is a short summary of our almost-an-hour long call.

Section titled After all these years, why do we still need to talk about “developers vs. business users”? After all these years, why do we still need to talk about “developers vs. business users”?

This seemed to be a valid question as an entry-point to other issues. After all, at this point CMS technology is decades old, and yet the assumption still seems to be that we have to sacrifice one group, just to cater to the needs of the other one. One could even argue that perhaps the constant need for rebranding the technology (web CMS, agile CMS, decoupled CMS, DXP, etc.) at least partially stems from the fact that at any given moment, one user-group feels disenfranchised with the solution.

Preston believes that perhaps the biggest reason why is that CMSs are inherently different from most softwares in one key aspect. While most softwares (Microsoft Word, Photoshop, Adobe Audition, ...) have a specific target persona (writers, photographers/editors, sound engineers, ...), a CMS by design does not focus on a specific job description.

Looking back at the first generations of CMSs (e.g. Wordpress, Drupal), they aimed to empower non-devs by giving them the ability to manipulate the way their content appears on its own page, and also how it manifests across the website. This is what initially led to what Preston calls “a grand compromise”, a double-edged sword still influencing CMS technology.

Preston continued by mentioning that the grand compromise was perfectly fine at the beginning, however 2 major factors shook things up:

1. The “channel explosion”: The number of available channels has exponentially grown in the last decade or two. This obviously means that today we no longer think of only one place for our content to go, and have to think about voice interfaces, AR/VR devices, digital signage, smart TVS, and an array of other channels in addition to the web.
2. The CMS world was almost exclusively focused on the web until pretty recently.

Mix these two issues together, and you can see why we are where we are today.

Section titled Headless systems and “the grand compromise” Headless systems and “the grand compromise”

The traditional monolithic CMSs were usually tied to a site builder that allowed for this grand compromise to take place. With the advent of headless systems, the situation started to change, leading to many marketers and other business users feeling rather left out. Preston added:

“Many marketers in different organizations are feeling like they are losing a battle. They are being disenfranchised in this move towards a headless CMS. One of the things that I fear is that the move towards headless systems that skyrocketed by developer adoptions, has really started to flatline. Because content creators and marketing organizations don’t see how they can get involved in the same rich way that they were used to”

The conversation naturally moved towards finding a possible solution for this imbalance, specifically concerning headless systems.

Preston believes that one of the earliest problems was the way headless systems initially positioned themselves. During the first marketing campaigns of early headless systems, the emphasis was solely on developers, leaving out the business users almost completely. As Preston put it “They didn’t have the chance to understand why they would need to move to a headless system”.

A look at the early articles on headless systems re-enforces what Preston pointed out. In the initial discussions surrounding headless systems, the emphasis on implementation benefits and better performance were so great that they completely eclipsed any discussion on anything more content creator focused. However, Preston does not see this as merely a positioning issue, but a more fundamental one: “A Headless CMS cuts off many aspects of content management that have been integral to the notion of what content management is.”

However, Preston is pretty optimistic about the solution:

“We have to stop thinking in terms of ‘experience manager’ and move towards ‘experience orchestrator’. [...] The era of classic CMS is over. We need to accept that, and educate everyone. We have to stop thinking of CMS as a web-only-aspect of how we manage content, and think about the fact that we cannot think about managing content on new channels, such as voice, in the same way we used to manage content on websites.”

For Preston, the solution to the disenfranchised marketers in the headless scene, is not reverting back to “web CMSs”, but to rethink our understanding of content and its distribution. Preston continued:

“We as CMS vendors have to do a much better job of how to connect the lines. How to surgically bring back the things we cut out”

Of course, “Things that we cut out” can be slightly different in the case of each system, but in general the main problems are usually shared. In hindsight this is probably a good thing, as everyone is trying to solve (more-or-less) the same problem. For us at Storyblok, empowering the business users comes in the form of a real-time visual editor, or a component-based approach to content.

Section titled How do we go from a visual-centric approach to managing voice content? How do we go from a visual-centric approach to managing voice content?

Thinking about the way we usually interact with content (both creation and consumption), visual clues are almost always key in the overall experience. Even in the simplest articles without any images, we still rely on hyperlinks to direct readers to certain points, offer them a visual menu somewhere in the page which they can easily use to navigate, or simply giving them a top-down view over the whole content.

With voice content, we immediately lose the ability to take advantage of all these clues. Yet, Preston believes there is something else that can make up for these shortcomings - the fact that voice content is by definition conversational, and thus more “instinctive” in nature:

“The notion of voice content is very different from web content [...] and it demands a very different approach from the visual manifestation of a website or a mobile application. When we work with web content, we usually base it on skills that we artificially acquire through time - using a mouse or a keyboard, a gamepad, or a touchscreen. However, conversational content is primordial. You don’t have to learn the means to have a conversation. We should consider this when thinking about spoken content and its relationship with traditional web content”

When talking about the more practical aspects of starting with voice content and voice interfaces, Preston briefly mentioned his previous experience with the Ask GeorgiaGov project, the first voice interface for residents of Georgia. Preston believes that the first thing most organizations should do when thinking about voice content, is to conduct an “omnichannel content audit”. In other words, analyzing every piece of content and asking where you want it delivered to, and how are you going to adapt it for these different areas?

Additionally, Preston’s practical advice for those at the early stages of voice content, is to look for cases of “voice-friendly” content. Simply meaning content that is more “dialogical” in nature. Some prime examples of such content are FAQs, or submission forms.

Section titled The particularities of voice content The particularities of voice content

There are too many differences between voice and other forms of content to be mentioned here. Some, such as the lack of visual clues and navigation were barely scratched during our conversation. As our time was running out, we decided to simply touch upon another critical voice-specific issue:

How do voice interfaces deal with the nuances of human culture and language? Should we simply accept the fact that many (or more precisely, most) languages, dialects, and accents are going to be underrepresented? Preston is concerned with what he believes to be a “problematic notion of anglophone privilege”.

Not only many non-english speakers are forced to interact with voice content in a foreign language, but in some cases where a language is represented there is still much work to be done. While dialects, accents, and vernaculars are obvious examples, the problem can be sometimes more subtle. One main example that Preston brought up was languages where the spoken and the written forms have clear differences (or in general, Diglossia).

There is much more to talk about when thinking about voice content. You can check out Preston’s blog and his recent book to learn more about his view on the topic.

If you are interested in practical guides of implementing voice content, you should check out How to create an Alexa Skill with a Headless CMS.