I've built a customer success function from nothing: the team, the playbooks, the tooling, with support and professional services under it too.
These days I run an enterprise book of around $37M in ARR across the US, LATAM, and Canada. I also ship the AI my CSMs run on, and I write here about where that pairing actually helps, and where it breaks.
That's usually two different people. Most AI work in customer success gets designed by someone who's never had to defend a renewal, and most CS leaders are waiting for a vendor to solve it for them.
I like talking with founders and CS leaders who are figuring out their first CS motion. If that's you, or something here lands (or doesn't), email me or find me on LinkedIn.
Your first customer success hire won't hand you a playbook in month one, and you shouldn't want one. Here's what the first 90 days should actually produce, what you owe them in return, and how to tell by day 90 whether it's working.
Read it October 202610 min read4 things to tryA retention number isn't one quantity. It's an aggregate of decisions made at different times by different people, most of them before customer success ever met the account. Here's which parts a leader can move inside a year, which parts they can only see first, and how to take the number without taking the lie.
Read it September 202610 min read2 things to tryThe promise is that AI lets one CSM cover three times the accounts. The arithmetic behind that is right about the hours and wrong about what the hours were buying, and the ratio it produces fails in a way nobody connects back to the decision.
Read it August 202610 min read1 thing to tryWhen you build the function from nothing, the first artifact isn't a playbook or a platform. It's a written decision about which customers you're willing to lose, and everything else follows from it. Here's what that cost me, and what I'd do again.
Read it August 20269 min read2 things to tryAI is close to owning the analysis half of customer success. The half it can't own is the one where a finding becomes a decision inside someone else's company, and I think that limit is structural, not sentimental.
Read it August 20268 min read1 thing to tryMost QBRs open with a utilization chart, and the chart answers a question your customer never asked. Here's how I build the review around what they've already committed money to instead, and the few rules that keep the process honest.
Read it