After the announcement of Salesforce Ventures investing an astonishing $300 million in Automattic last week, I wrote extensively on the topic and in particular the impact on potential strategic partnership between Salesforce and Automattic, given that WCM is still a gap in their portfolio. I concluded (and other smart folks in the industry agreed) that:
I don’t see WordPress replacing [Salesforce’s] internal CMS ambitions; it’s not architected as something that would slot in easily and does not fit the publishing model and use cases that Salesforce needs to enable better customer experiences.
The CEO of Automattic, Matt Mullenweg, also replied:
If this were Elon Musk, I would assume he was trolling, but those of us who follow the industry and vendors understand Matt to be a pretty thoughtful individual, so I’m taking him at face value and writing my thoughts on the topic. Admittedly, it’s not new ground – Automattic has appeared in both The Forrester Wave™: Web Content Management Systems, Q4 2018 and Gartner Magic Quadrant for Web Content Management, but I’d like to try to approach this from a different perspective.
Without spoiling the fun of reading the analyst reports yourself, I will simply suffice to say that Automattic is ranked a fair bit lower down in the classification of both reports. This is not to say this is an accurate reflection of the company and prospects – indeed at this point it is very likely that Automattic is the most valuable WCM vendor in the entire evaluation list. But it is merely a reflection of the fact is that the criteria is set against “enterprise requirements”, which is largely what I based my statement around “architecture” and use cases not fitting Salesforce needs. To be clear, I use and like WordPress (indeed, this is a WordPress site – when I was in a hurry to get blogging again, this was by far the quickest and most suited tool to my publishing needs). But that key use case reveals the kernel of truth in the philosophy of their product and users – it has a very strong bias to publishing and publishers – to the detriment of other use cases or generalized patterns which can be extended. In fact, if you look at the WordPress VIP case studies, you will see that they are about 90% publishing properties. Similarly, the list of integrations is largely around media needs (video, stock photography, advertising, social, etc.)
And here’s the big issue for an enterprise WCM buyer – most publishing is largely fire-and-forget. In contrast, digital marketing is a constant agile iterative process, that is governed by many different constraints; unique content type requirements, ever changing product and service offerings, campaigns, buyer persona, customer journey, etc.
Every interface and metaphor within WordPress is aligned around making that simple publishing model faster and more effective – right down to the big push around Gutenberg (which is an obvious nod to this bias). This model has a number of knock-on effects;
- WordPress is aligned around web pages, and not structured, componentized, content. Again, this metaphor is great for a publisher, but not if you are an enterprise that needs to manage the content process for a new product launch campaign that will end up with dozens of structured products, features and benefits in ten different languages, dozens of locales and re-used in media ranging from email campaigns, to mobile apps, to kiosks. This includes the ability to have visibility into all of those facets and relationships. Nothing about WordPress is architected to handle this particularly well – not the content model (which is basically two types; posts and pages), not content sharing between properties, transformation, or headless APIs.
- Publishing new content, and not findability/curation. This means that elements such as taxonomy and search are very fixed and insufficient for enterprise needs, especially where these are usually being synced with other ECM or MDM systems. As an example, an enterprise WCM would be able to enable things like distinguishing between “active” content and “historical” content and taxonomies – so anyone searching for a product would get to your new information unless the specifically navigate in the context of searching for historical/documentation information. This would include things like URL redirect management. Similarly if you had a campaign with licensed content (such as a celebrity endorsement) being able to track down, remove, and redirect would be a common and important task. The methods of search, workflow and content management are all related to this “curation” function and user task role which largely does not exist in WordPress.
- Content marketing is not the same as publishing (though publishing can be good content marketing). Creating content to a proper content strategy requires understanding exactly why you are creating a bit of content; this means being able to see in context who you are creating the content for. What are they trying to accomplish? How is you content helping? What other content already exists for this? Content strategy requires having to update, replace or even remove out of date content on a regular basis. Having the tools to understand all of this context easily is key – and most leading WCM systems have the structure in place to either do this natively, or make it comparatively easy to do so. In contrast, if you start fighting the inherent publishing bias in WordPress, you get into trouble very quickly.
Now, to be fair, a number of very capable organizations doing content marketing use WordPress as their WCM in their marketing technology stack. But this potentially represents some selection bias in that most organizations who care enough about martech to submit Stackie award submissions probably also have a very capable team who are able to understand and work around the inherent biases and models in WordPress and compliment with other tools as required. Yes, you can add custom types in WordPress – but it’s not particularly easy or friendly. For an Enterprise CMS and those use cases, you basically have to start from the perspective of turn the metaphor completely on its head – you often start with nothing but the content model and custom types (which are typically very easy to create) and then layer on the functions you would need (usually by configuration or customization). It’s a lot of work, and probably a pretty poor choice for a publication-type site (especially when WordPress does so much out of the box!) but the corollary is that it does mean that literally any use case can be accommodated without breaking the inherent metaphor of the system.
So going back to the Salesforce example, as a platform, they literally have dozens of different use cases and requirements for content (customer relationship management, commerce products, support cases, emails, etc.) They are all very different, but interlinked. None of these are particularly aligned to the “publisher” model.
To conclude, WordPress is a great publishing CMS, but for enterprise requirements it often falls short. The analyst reports and rankings in this specific context are accurate – but it does reveal a truth – don’t just pick the “top” vendor in those reports; do your research, because your requirements may lend itself to a “lesser” vendor with a specific focus. If you have a publishing model, it’s very likely that WordPress is one of your best WCM options, it just may not be most companies best “Enterprise WCM” option.
I’m not sure I see it as an apples to apples comparison. As per my previous post, I see Automattic as less of an “enterprise WCM/DXP” and more like a potential hybrid social network. If you look at many of the enabling technologies in WordPress, they are centered around creators and communications; Jetpack enabled comments, pingbacks, sending emails automatically on new posts. Similarly, Tumblr was actually a “proper” social network and still boasts a decent audience (especially for what Automattic paid). If you read what Mullenweg talks about on a regular basis and his interest in the platform, publishing, and the “open web” it’s clear where his interests lie. And I think it’s a better business play, compared to WCM, which is a very red ocean. For the last few years, the Social space is a garbage dump – between the racism, wholesale trafficking of customer data and the focus on “engagement” as the only metric that matters, and it is ripe for disruption. Everything around Automattic seems far more interested in self-publishing and actual enablement of two way communication (rather than “engagement” – which is a passive, emotional consumption – basically like the High Fructose Corn Syrup of the web; cheap, lazy and largely bad for you).
All that to say, I assume the $3 billion valuation is predecated on investments and growth with the social/media space and not particularly around enterprise WCM. And if I’m wrong, Matt, and you do want to try to compete more fully in that space, feel free to reach out – I just might know some marginally clever folks who love to work remotely and have been doing enterprise WCM for a long time (spoiler alert: I’m referring to myself). My first recommendation for shopping with the new Salesforce Ventures credit card might be a good Content Marketing Platform – they are ripe for purchase and it fits with your “creator” view and will help to give a good grounding in content marketing (vs. publishing).