The Dirty Secret About AI in CS.

Everyone talks a big game on LinkedIn about how AI is changing CS.

But as much as the algorithm wants to make us all feel like we're constantly behind, here's what AI adoption looks like in most CS organisations right now:

CSMs are using generic LLMs and mini point solutions. Summarising calls. Drafting follow-ups. Dumping customer data into a prompt to generate talking points before a meeting. Maybe they automated a few small tasks like calendar management.

Leadership looks at the usage dashboards, sees over 90% WAU and calls it a win.

My hot take: I don't think that's a win. 

Maybe we're getting a bit more productive. But I haven't seen a real transformation in how we deliver value to customers yet. 

My hypothesis, from dozens of conversations with CS leaders, is that most organisations are plastering generic AI on top of a broken workflow and messy data and hoping individual contributors will figure it out. 

But with a wobbly foundation, all they're doing is getting slightly faster at the same old stuff, while becoming no more effective at the job. 

The 2 questions that cut through all of it:

1. Are your customers getting more value from you than they were last year?

2. Is your company getting better ROI from CS than they were 12 months ago?

If you can't answer positively and with confidence, AI is just the latest shiny object. 

It doesn't have to be, though. 

The Core Value Problem Hasn't Gone Away

I've been surveying CS leaders for years. Every year, I ask about the top challenges. Every year, defining and measuring customer value ranks number one (Here's the 2025 research).

Value is not a new problem. It's a problem we're stuck with.

Most CS teams still can't clearly define and measure the value each given customer is trying to achieve. They know the use cases. They know the features adopted. They can show activity, seats filled, and consumption. But value? The actual business outcome the customer bought the product to get?

Still fuzzy.

CSMs continue to run plays against goals they can't see. 

And here's what that fuzziness costs you in practice: 

1. Every renewal conversation becomes a negotiation without evidence.

2. Churn is still a surprise that you catch way too late to turn around.

3. Every expansion pitch is a gamble versus a tailored conversation.

Traditional CS Tools Won't Close This Gap

The platforms most CS teams run on were built to manage activity. Tasks. Touchpoints. Made up health scores that tell you what a customer did, not whether any of it mattered.

They weren't built to:

• Track customer value.

• Measure progress against it.

• Integrate dynamic intelligence into the CSM's workflow.
  

So no matter how much AI you layer on top of this current stack, the workflow is broken from the start.  

The AI inherits the same blind spots that your CSMs already have, or worse, because it lacks context that lives only in the rep's head. It might be able to summarise activity and draft communications. But it cannot tell your CSM whether this customer is on track to achieve what they came for and what to do about it.

That's the gap.

You can't close a value gap you can't see.

What CS teams need is an intelligent system of action:

• Continuous value definition and alignment

• Deep understanding of personas' interests 

• Proactive tracking of key milestones

• Measurement of real business outcomes

• Insight into the customer's perception of value.

No "quick" shortcut or randomly giving your CSMs LLMs will fix this problem.

We need a re-build, not a redecoration. 

Closing The Gap With AI

I won't lie, I've been pretty underwhelmed by the solutions in the market. I tried so many different tools in search of one that would give my teams an edge here (our CS Ops team ended up building it internally at HubSpot).

But last week, I happened to be on a call with Nigel, CEO of Foresight, and he showed me what he has been focused on over the last year, and I was genuinely impressed. They rebuilt Foresight from the ground up to surface customer value in a scalable, repeatable, workflow-embedded way, using AI.

Not as a separate analytics exercise you do once a year.

Not as a slide that your CSM builds once a quarter.

As live context, in the rep's workflow, connected to what the customer said they needed, and tracking whether they're getting there. 

I was impressed.

The good news is that if you're struggling to define, measure and proactively drive value using AI, Foresight is hosting a free event on the 2nd of June that I think you should attend: Closing the Value Gap: The New Frontier for Customer Success.

What This Means for CS Leaders

As a leader, it's not enough to bring generic LLMs to your team and expect AI to transform your business from the bottom up. 

1. Stop measuring AI adoption by licence utilisation. That metric tells you how many people used it, but it tells you nothing about business impact.

2. Audit the foundation before you add more on top. If your CSMs are the glue between your systems or if your data is not trustworthy, no amount of AI bolted on top will fix that. The foundation has to be solid.

3. Design for the median CSM, not the exceptional one. The transformation doesn't come from what your best people figure out on their own. It comes from what your whole team does consistently, with less effort and better context than before.

Want to learn more about how to deliver value in the age of AI?

Come to this event on June 2nd!  

Daphne Costa Lopes

Linkedin