Every CS team has had this conversation at least once.
"Can the tool do this?"
"No. But what we recommend is..."
And then someone explains, very professionally, why the way you work with your customers is wrong, why their way is right, and why adapting your entire operating model to fit their product is actually called implementing best practices.
You call it every week.
This is structural failure number three in customer success, and it's the one that makes the first two worse. We built a profession on a broken premise. We staffed it with exceptional people and gave them the wrong work. And then we handed them a tech stack that locked them into doing both of those things the way the software wanted them done.
The tools aren't the problem in isolation. Some CS platforms are genuinely good at what they do. The problem is the assumption baked into all of them: that customer success is standard enough to fit inside someone else's schema.
It is not.
The way you onboard your specific customers. The signals that matter in your specific context. The moments that make or break a relationship in your specific market. The irreplaceable way your team delivers value to your specific ICP. None of that fits cleanly into a workflow engine a product team designed for the average CS org.
So you adapt. You compromise. You figure out the workaround. You build your operation inside the box the software gave you and call the constraints "best practices." Over time you stop seeing the box. It just becomes how things work.
The compound effect is that you're now optimizing a broken premise with the wrong people doing the wrong work inside a system that can't flex to match how you actually operate. Every structural failure amplifies the others. The tool lock-in is the last layer, but it seals everything underneath it in place.
What would it look like without the box?
That's not a rhetorical question. It's the one worth sitting with. If the tools bent to your operation instead of the other way around, what would you build? What complexity would you absorb for the customer that you currently hand back because the software can't handle it? What signals would you monitor that you currently miss because nobody built a detector for the thing that matters in your specific context?
For most CS teams, the honest answer is: a lot. The box has been there long enough that the imagination has conformed to it.
The infrastructure that breaks this constraint exists now. Agentic workflows built around your specific operation, your specific customers, your specific signals. Not a platform you configure — a system you design. We go deep on what that looks like and how to build it at Extensible Agents.
But the first step isn't technical. It's recognizing that the box was always optional.
This is the third in a three-part series on the structural failures of Customer Success. Part 1: The Broken Premise | Part 2: The Best People Are Doing the Wrong Work
