The Headless Revolution part 2 - where does many heads leave the body?
In the first part of my musings on Headless 360 I mentioned that this approach wasn’t all a promise of sunlit uplands:
I see some risk though, given that they are presenting themselves as infrastructure rather than a complete solution, but that’s a discussion for the next post.
Simply put, going headless risks playing straight into the hands of those that have long claimed that Salesforce is just a database and triggers. One of the traditional differentiators of Salesforce, and definitely one that appealed to me when I was starting out on the platform, was the metadata-driven User Interface. All I had to do was create an object and fields, drag the fields around in the page layout editor, and I had a set of pages for users to manage the record instances, they were available in reports, and I even had access through a mobile application. Does the privot to custom front ends and agentic access reduce the value of Salesforce to data storage and workflow?
Salesforce has always been a lot more than a database and automation, with sophisticated security and sharing, identity management, event management, extensibility, and a host of partners creating solutions on the app exchange. Rather than a database, I view it as a business execution engine. Salesforce isn’t looking to turn themselves into a modern version of Database.com for AI tools, it wants to pivot from the CRM your users live in to the enterprise execution layer underpinning both humans and AI.
For existing customers I don’t think it matters too much what Salesforce are now claiming they are. These customers have invested years into customising Salesforce to their exact needs and there won’t be any appetite to do that again on different technology.
Prospects are a different kettle of fish though. Salesforce has typically pitched a unified proposition of database, applications, User Interface, automation, security, and ecosystem, which allowed them to charge premium prices. If the User Interface is composable, and indeed optional, why not the rest of the layers? Once prospects start thinking in those terms, the pricing comes under pressure. Competitors can also attack individual layers, rather than having to compete against the whole behemoth. If this pivot is successful and Salesforce becomes the invisible infrastructure beneath agents, existing customers might start questioning why they are paying so much for a system that nobody logs in to any more.
In my opinion we’ll see the following segments in the future:
Large or regulated organisations value the governance, compliance, and operating maturity of Salesforce
Smaller, more technical, organisations opt for the self-assembly route, building a custom CRM from cloud-native components
AI native organisations bypass the whole legacy CRM approach entirely. Users living in systems of engagement are replaced by AI tools coordinating customer operations across many systems
Although when “the future” actually is, how large these segments are, and the total addressable market of everything else, I couldn’t even begin to guess at. It’s fun to think about though!