In Partnership With
Hey {{first_name | CS Pro}},
Can AI actually replicate your best CSM? That was the big question at our live podcast recording in Amsterdam, hosted by Channable.
I had 3 brilliant CS leaders on stage with me: Sareena Hirani from Amplitude, Elena Cocco from AlisQI and Io Michels from Channable.
Their answer was pretty clear…AI won't clone your superstar, but it can make the rest of your team a whole lot better. Here's what they're building, what totally flopped, and where to start on Monday.
youtube - https://youtu.be/J66R1We5Plc
But first, today’s sponsor
↓
Onboarding calls shouldn’t rehash product basics
Onboarding each new customer takes up a lot of time teaching essential product knowledge. Business Warrior moved their basic onboarding step into a self-serve training hub with interactive demos. Now they spend 50% less time on each onboarded customer.
Read HowdyGo’s playbook detailing a practical guide to using interactive demos for onboarding, choosing the first interactive demo to build, and tracking whether it’s improving activation.
The playbook covers:
How interactive demos can help in early onboarding and in-app friction points
A framework for deciding which demo to build first
Data-backed tips for structuring demos that get customers to outcomes
Where to put demos: help docs, onboarding hubs, and certification
Amplify your best CSM, don't clone them
Not one panelist wants an "AI CSM." Io said he wouldn't call it a person just yet, but his team does build the traits of their best CSMs into their AI.
His goal is to amplify what great CSMs do best. With 200 to 300 accounts per person, his CSMs are great at making decisions, they just don't have time to scour their whole book. So AI points them to the right signals at the right moment.
Sareena's team built a customer intelligence dashboard based on quarter over quarter churn analysis. The catch? Look at almost any customer and it's all red. Her top CSMs see through that noise, ask the right questions, and know where to spend their limited time.
"AI is not going to tell you what to do."
Elena took a different approach. Instead of prompting AI, she started prompting her CSMs, asking why certain decisions were made at a certain time. Her take? Replicate the decision making and the repeatable playbooks… and let the rock star shine.
Use AI to hear what customers aren't saying
My favorite story of the night came from Sareena. After an EBR with Kahoot, her CSM came back saying it went great. The AE said it was terrible and they weren't going to renew.
Small signals and words from the buyer showed they weren't happy, but you'd miss them listening with happy ears. So her team turned their call notes into a skill that scans critical meetings for certain tones and keywords, then escalates risk to her and the sales leader. It helped them get ahead of a few big risks in time (they still lost one).
This is SO important. "Fine" can mean something totally different if your customer is from the Nordics, the Netherlands, the UK or the US.
Io's team runs month by month subscriptions, so every quarter has 3 churn moments. AI lets them act the day something happens, instead of waiting for the next scheduled QBR.
But AI catching the signal is only step 1. Your CSMs still need to handle the objection and speak the language of the exec who signs the renewal. If that's where your team gets stuck, I run a workshop on ROI storytelling and objection handling for renewals, grab a time to chat about it here.
What failed (because nobody talks about this)
I loved how honest everyone was here.
At Amplitude, leadership believed AI would replace 1:1s and formal QBRs. Those meetings came back within a few weeks, because Claude couldn't generate the right level of data and teams weren't planning ahead like they used to.
Elena swapped a slide deck for a long Claude generated report. Nobody read it, and her CEO flagged information overload. Back to the slide deck.
"If the process was broken before, putting AI on top is not gonna make it broken less."
Io's team spent about a year chasing the perfect AI churn score and never got there. They also overfocused on AI literacy, and some customer conversations suffered. His advice for leaders? Don't over AI.
The skills that make a CSM amazing aren't note taking or updating the CRM. It's objection handling, discovery questions, executive language and business acumen (understanding what makes your customer's business profitable and how your product helps). If you want your team forecasting renewals and upsells with that kind of confidence, book a call to talk about a team workshop.
Weekly challenge
I asked each panelist what to start on Monday.
Io said write down all the small things your best CSM does. Sareena said build an AI brain with the knowledge you want your best CSM to have, so the whole team can lean on it. Elena said start with the most annoying thing.
So here's your challenge. Pick 1 annoying task, ideally a question your team gets asked 3 or more times, because Elena says that's the moment to build a skill. Then sit down with your best CSM, write down how they handle it and why, and turn that into your first AI first playbook.
That's it for this week! If you want to hear the full panel (including the audience Q&A on agents and AI governance), have a listen to the latest podcast on Spotify, Apple Podcasts or YouTube and let me know what you think!
I hope you enjoyed this week’s newsletter.
If you have any questions or suggestions, please feel free to contact us.
Cheers to your CS success,
Anika
