Ask anyone who has run Customer Success what the job is and you'll get some version of this. Get the customer to do the things they need to do to get value from what they bought.
That sentence has been the operating assumption for fifteen years. It's worth sitting with, because it contains two asks that don't survive being said out loud.
The first ask
The customer has to do the thing.
Not intend to. Not agree it's a good idea on the call. Actually do it, on their own time, inside their own company, against everything else competing for that week. And plenty of them won't. Some can't. Some quietly decide not to and never mention it.
That isn't news and it isn't a character flaw. Everybody who has done this work has a list of accounts that failed for exactly one reason, which is that the customer didn't do the part that was theirs.
Here's the part that gets skipped though. For the ones who do try, trying isn't enough.
They have to be good at it. In most categories they have to be genuinely excellent at it, which means becoming world-class at something that isn't their job, that they'll do occasionally, and that you do every day. Buy a cold email platform and you now have to be good at cold email. Buy an analytics product and you have to be good at analysis. The product doesn't produce the outcome. Skilled use of the product produces the outcome, and the skill is the customer's problem.
So the model requires two things from them: do the work, and be excellent at work they've never done before.
The second ask
Now look at your own side of it.
For a CSM to help that customer, the CSM has to be world-class too. Not only at the product, which is the easy part and the part everyone trains for. At the customer's domain, the one they're failing in. And on top of that, at getting a reluctant adult to do something they've already demonstrated they'd rather not do.
Every one of them has to be that. The person who started Monday. The one two years into their first job. The one who's wonderful with people and has never run a campaign in their life.
Some of your team will be that good. Not all of them, and it usually isn't their fault. They're new, or they're new to this, or the domain takes years. But the model needs all of them to be excellent, immediately, or the accounts they touch don't work.
What that produces, entirely predictably
Two impossible asks, stacked, and then a lot of surprise about the results.
Burnout gets explained as workload. Eighteen-month tenures get explained as ambition, or the market, or a bad fit. Leaders leave before anyone works out they're bad at the job, except they were never bad at the job.
Nobody is bad at this job. The job as specified cannot be done, and it's been specified that way from the beginning.
I'd know. I named the thing and I've been arguing about what it should be for fifteen years, and it still took me an embarrassingly long time to notice that the central mechanic was an impossibility rather than a difficulty. Everything downstream of it is people being asked to compensate for that, which is why the compensating gets admired instead of fixed.
The move nobody would make
If the customer can't or won't do the work, and your team can't reliably teach them to, there's an option that everybody has thought of and almost nobody takes seriously.
Do it for them.
It sounds expensive and usually isn't, and the reason it isn't will be uncomfortable if you go and measure it. Count what you currently spend re-managing expectations, re-explaining what they bought, chasing work that hasn't happened and rebuilding after it's done badly. In most companies the time spent persuading a customer to do the work is several times the time it would take to do the work.
You already pay for it. You just pay for it in a form that doesn't appear on any line item, and that produces a worse outcome than the thing you were avoiding.
What it changes
The reason this matters here isn't cost, it's what becomes possible afterwards.
A relationship where the recurring conversation is whether they did their homework has nowhere to go. You can't name a next thing for someone who's failing the current one. Every milestone is theoretical, because the milestone depends on work that isn't happening. There's nothing to orchestrate toward.
Take the work off them and the conversation changes shape. It becomes here's what happened, here's what it produced, here's what's next. That's the conversation expansion actually runs on, and it isn't available while both sides are pretending the homework is going to get done this time.
The job was never possible. It's possible now, and the thing that changed is that doing the work stopped being the expensive option.
Lincoln Murphy formally named and popularized Customer Success starting in 2010 and has spent 15 years connecting it to expansion revenue and commercial outcomes. Read The Premise.