Finding the Right Customer Story in Seconds: Building the Customer Reference GPT

Finding the Right Customer Story in Seconds: Building the Customer Reference GPT

In sales, timing matters. When a prospect asks, "Do you have a customer that's dealt with this exact problem?" being able to quickly surface a relevant example can be the difference between moving a deal forward and losing momentum.

The challenge isn't a lack of customer proof points. Quite the opposite. Most go-to-market teams accumulate a wealth of knowledge across call recordings, Slack channels, case studies, business reviews, and internal documentation. The problem is finding the right story when it matters most.

Why Finding the Right Story is a GTM Bottleneck

It's a pattern that shows up in most go-to-market teams: customer knowledge exists, but finding the right story at the right moment means searching across multiple systems, digging through call notes, or relying on whoever happens to have the institutional knowledge.

The larger the team, the more pronounced this challenge becomes. Valuable knowledge fragments across tools and conversations. Even strong references get overlooked simply because they can't be found fast enough—a hidden cost measured in time spent searching instead of selling.

I wanted information to find people, not the other way around.

What Does the Customer Reference GPT Do?

The Customer Reference GPT acts as a centralized intelligence layer for customer knowledge. Instead of navigating multiple repositories, users ask questions in natural language and get relevant customer examples, proof points, and source material—fast.

A few example queries:

• Which customers have addressed phishing attacks targeting executives?
• Do we have financial services customers using this workflow?
• Can I find customers similar to this prospect's size, industry, or use case?

The GPT pulls from call notes, Slack channels, case studies, business reviews, and other internal resources, synthesizing across sources rather than returning a simple keyword match. Customer reference discovery becomes a conversation, not a search.

Early Results and What's Ahead

We're still in the early stages, but the initial signal is encouraging.

The most immediate impact has been time saved. Previously, teams needed weeks to reach out to teammates, dig through systems, and piece together context. Now, everything can happen in seconds. Teams are surfacing proof points they didn't know existed: strong customer stories that weren't missing, just hard to find.

There's also been a meaningful reduction in dependency on a small number of people who historically held all the knowledge. Teams can now access what they need independently, when they need it.

As we add more sources, refine how the GPT reasons over customer data, and gather feedback from the field, we expect both the quality and coverage of results to keep improving. The goal is for the collective experience of our customer base to become a true shared organizational asset, not something that lives in disconnected systems or in a handful of people's heads.

The Path to Frictionless Discovery

Customer knowledge is one of the most valuable assets an organization has. The challenge isn't creating more of it—it's making it accessible at the right moment.

The Customer Reference GPT is an early but meaningful step toward that. When the right story is only a question away, teams spend less time searching and more time building trust and moving conversations forward.

The best customer reference isn't the one sitting in a document somewhere. It's the one you can actually find when you need it, and we're building toward that.