The Journey: #64 I Carried a Watermelon

Nobody puts Baby in a corner.

And nobody told early-career Kristi that she had no idea what she was actually looking at when it came to her accounts.

If you've seen Dirty Dancing, you know the scene. Baby shows up, nervous, and the only thing she can think to say is: "I carried a watermelon." Just... a watermelon. No context. No awareness of the full picture. Just the thing she was holding.

That was me. For longer than I'd like to admit.

I was carrying watermelon accounts. Green on the outside. Red on the inside. And I didn't know what I didn't know.

The Things I Didn't Know to Look For

Early in my CS career, I thought I was doing a good job managing accounts. I had strong relationships. My customers liked me. I came off calls feeling good.

What I didn't have was the full picture.

I didn't know to look at whether people were actually logging in. I didn't know to connect the dots between open tickets and unresolved friction. I didn't know that the admin who raved about us on every call might also be the one quietly blocking us from the executive team. I didn't know that a drop in usage might mean churn, or might mean a company was about to acquire someone and needed us more than ever.

I was reading the green rind and calling it done.

The problem wasn't effort. I was working hard. The problem was context. I didn't know what I didn't know, so I couldn't ask for what I was missing.

Watermelons Are Everywhere

Now that I'm on the consulting side, coaching CS leaders and practitioners, I see this constantly. CSMs rating customers green who, the second you peel back the layers, are clearly not healthy.

It's not laziness. It's the same thing I experienced: missing context, not knowing what to look or listen for, not having the time or the data access to get the full picture. That's real, and it's happening in a lot of companies right now.

And the stakes are not small. Getting it wrong has a direct impact on revenue.

I've stopped believing that humans alone can consistently uncover the truth. I've also stopped believing that AI alone can do it. Because I've lived both sides of this.

Here's what that looks like in practice:

Customer A drops in usage. The AI flags risk. But on a call, the exec reveals they're supporting an unannounced acquisition, and the company they're buying doesn't have a platform like yours. They want to set up a demo when it closes. Data says churn. Context says growth opportunity.

Customer B has an admin who is your biggest fan. She raves about value, blocks exec access, and asks to shift to a monthly cadence. The CSM leaves the call feeling great. Meanwhile: 50% of users haven't logged in for 30 days, 10 feature requests are open, 6 tickets unresolved. CSM says green. The data says red.

Both of those stories are real. I've lived them.

What I Do Now

I'm supporting a full customer load right now. There is no version of reality where I'm manually pulling the full picture on every single account before every call. That analysis is exactly what I want to outsource.

I've been building this out in Planhat using three AI automation rules that work together: a last conversation summary so nothing falls through the cracks between calls, a business health score that looks at the health of the business outside of our partnership, and a company news flag that surfaces anything happening in their world that could affect the relationship.

Those three automations layer on top of the data fields I've manually added, my own sentiment, plus automated pulls like last engaged date. Together, they give me a holistic picture of the account that goes well beyond what I'm hearing on calls. And I ask a lot of good questions on calls.

If I were on the SaaS side, I'd layer usage and consumption data on top of all of it. That's a pretty powerful setup.

Early-career Kristi didn't have any of this. She had a watermelon and a lot of confidence in her read of the room.

I've upgraded since then. Barely, but meaningfully.

The Thing About What You Don't Know

The hardest part of not knowing what you don't know is that you don't know you're missing it.

I wasn't ignoring signals. I genuinely couldn't see them. It took time, exposure, a few churn surprises I didn't see coming, and eventually the right tools and frameworks to understand what a healthy account actually looked like versus what a friendly account looked like.

Those are not the same thing.

If you're still getting surprised by churn, it might not be a you problem. It might be a context problem. What does your full picture actually look like? What data are you pulling before every call? What are you outsourcing to automation, and what are you relying on instinct to fill in?

The goal isn't to become a data analyst. The goal is to stop carrying watermelons and calling them wins.

Baby figured it out eventually. So did I.