The Case for Auditing Old Processes Before Building New Ones ft. Kellie Snyder (LinkSquares)

When your team supports thousands of customers, churn can stay hidden in plain sight. This week on [Un]Churned, Kellie shares how AI helped bring them to light.

Kellie Snyder, Chief Customer Officer at LinkSquares, sat down with Josh Schahter on this week’s episode of [Un]Churned, “191. The Retention Killer Hiding in Your Product Demos,” to share what it really takes to rebuild a Customer Success (CS) approach with AI, how a platform relaunch can drive change in ways internal efforts often can’t, and what the data showed about why customers were leaving quietly.

---

Listen on YouTube, Spotify, Apple Podcasts, and Gainsight.com.

Subscribe now and get the [Un]Churned take straight to your inbox.

---

🎯 The [Un]Churned Take: Your Best Processes Deserve a Second Look

These days, most CS teams are wondering, “What new thing should we build with AI?” We’re thinking about new workflows and playbooks. The idea behind this question is that the missing piece is something we haven’t built yet.

But Kellie Snyder’s experience suggests a different question: What have you already built that’s worth revisiting?

Retention matters more than ever. Gainsight’s 2025 CS Index found that 76% of companies now see customer retention as a main revenue metric. It’s getting harder to win new customers, and renewals face more scrutiny. In this climate, not knowing about a broken process can be more costly than before.

What Kellie Found When She Looked Back

LinkSquares is an agentic Contract Lifecycle Management platform serving legal, sales, and procurement teams. When Kellie became CCO eight months ago, she took over a carefully built onboarding process. The existing process was thoughtful and thorough, and made by people who really understood the product. There were detailed demos, full feature walkthroughs, and a real effort to show customers everything the platform offered.

The team who built the process wanted customers to see the full scope of LinkSquares’ capabilities.

What they couldn’t see was how it felt for customers at scale. When Kellie’s team pulled data from nine systems and analyzed it with Claude, the AI revealed a previously hidden pattern: customers were leaving after being shown too much, too quickly.

“We scared some of those customers,” Kellie said. “That was part of our retention problem. It was just too much for them. It was more than what they needed.”

No one set out to design a bad process. It made sense from the inside, but the outside perspective only became clear when AI could review hundreds of accounts at once.

What’s Slipping Through the Cracks?

This story matters for more than just onboarding because it raises questions about every process you’ve built and stopped examining.

CS teams build up processes over the years, like QBR cadences, health score weightings, escalation triggers, and renewal timelines. Each was created by someone with good judgment and real context. Most have never been tested against large-scale data, because until recently, that kind of analysis wasn’t practical. It used to take a data team, a project, and months of work.

Now, it just takes an afternoon and a well-crafted prompt.

CS leaders are rushing to build new AI-native processes, but the smarter move might be to use AI to review what they already have. Not to automate, but to audit. To ask the question the original designers couldn’t answer: does this actually work the way we think it does?

Kellie’s answer changed how LinkSquares onboards every new customer from then on. The real question is what you might discover about your own processes.

---

🎧 Listen for These Moments

If you're inheriting an org or just trying to keep up with the AI wave, you'll want to hear from Kellie. Here are four moments to listen for in this week’s episode:

1. Why Kellie compares AI to A/B testing — and what her days at Adobe taught her about the gap between what humans think they know and what the data actually shows.

2. How LinkSquares thinks about value realization — why they start identifying customer outcomes in pre-sales, and why they don’t assume those outcomes stay the same over time.

3. The digital-first over-rotation — why Kellie thinks CS went too far, how the new platform gave her cover to pull back, and what rebuilding the human engagement model actually looks like.

---

🔎 Where to Find Kellie

• LinkedIn: linkedin.com/in/kelliesnyder

📎 Referenced in This Episode

• LinkSquares — AI-native contract lifecycle management platform

• Claude by Anthropic — the AI tool Kellie’s team uses for customer data analysis and internal enablement

• Gong — referenced as a tool for capturing call data and reducing manual CSM data entry

• Snowflake — referenced as the data infrastructure layer for Link Squares’ customer 360 consolidation

---

Wrapping Up

The processes you trust the most are the ones worth interrogating first. Don’t forget to make time to pull up processes you haven't questioned in the past couple of years. That’s your new starting point.

See you next week 🧠