# How Brex Catches Churn Risk and Competitor Moves | Rippit

> Brex reads every support ticket on the Rippit Growth Data Platform, flags churn risk in 2 to 3% of tickets a week and reaches customers within hours.

Canonical: https://www.rippit.com/customer-story/brex

Customer story · Brex

# How Brex uses Rippit to catch churn risk and competitor moves

Each week, 2 to 3% of customer cases show signs a customer might leave. Brex now reaches them within hours and sees competitor offers as they land.

By Rippit Team · Last updated Oct 29, 2026

With Rippit, Brex

-   ### Spots churn risk every week
    
    Finds churn risk in 2 to 3% of tickets each week
    
-   ### Gets to unhappy customers proactively
    
    Reaches at-risk customers within hours of a flagged case, often before they could answer a survey
    
-   ### Caught a competitor’s play in one week
    
    Confirmed a competitor’s offer to its customers by spotting ten in one week, and confirmed it for the whole company
    

On this page

Overview

Industry

Fintech

Company size

About 1,500 employees

Support

In-house CX, a VIP team and BPO partners, including a team in Greece

Brex, the financial platform for fast-growing companies, uses Rippit, the Growth Data Platform, to read every support ticket for churn risk and satisfaction. At-risk customers now hear from Brex within hours, competitor offers surface as they happen, and a churn risk report reaches everyone from the CEO down every two weeks. It went from pilot to a company-wide report in about four months.

The stakes

## Why does catching churn risk early matter to Brex?

Brex’s competitors don’t wait for a renewal date. Some were reaching out to Brex customers directly with offers to switch. Every customer who leaves is revenue a competitor wins.

Customers often say they are thinking about leaving long before they go. Some say it outright: close my account. Others drop subtler hints in a support ticket. Brex needed to hear both while there was still time to keep the customer.

The problem

## Why didn’t Brex hear these customers before?

The only process reading Brex’s support tickets closely was QA, and it was built to grade agents, not to listen for customers at risk. It was a traditional program: a yes-or-no scorecard, about three reviewed tickets per agent, and 20 to 25 minutes per review. Long phone calls rarely got picked. Surveys came back on 13 or 14 cases a week.

Scores barely moved for a couple of years, and the weekly business review showed the same trends every time. Agents and coaches spent their energy debating points instead of improving the customer experience, and QA was seen as the quality police.

“Two to three percent of our cases every week have some churn risk indication. I cannot remember one time in a QA session where somebody flagged churn risk,” said Brian Klaja, Sr Director of Operations Strategy & Enablement at Brex.

The shift

## How did Brex start hearing every customer?

Brex had wanted deeper conversation analysis for a while, but it stayed on a roadmap that never arrived. So the team went looking and found Rippit. The week it went live, they showed the COO a demo.

“After a quick preview, our COO stopped us and said, ‘No more manual QA. Go deep into this platform. This is going to be incredible for us.’ That was the catalyst to burn our QA playbook and rebuild it from the ground up,” said Brian.

The team didn’t automate the old scorecard. It replaced it with analysis built around customers and outcomes. Brex piloted the new approach with its smallest BPO team, in Greece, refining it together over about six weeks. Nothing got worse, and Brex’s VIP support team asked to join next. The remaining BPO teams and internal CX are following.

“When we stopped QA, it wasn’t, let’s automate the scoring. It was, let’s stop doing what we’re doing and rebuild the whole thing,” said Medha Mayawala, Business Development at Brex.

How Brex uses Rippit

## Retention, competition, product and coaching from the same customer data

Use case 1: Retention

### How Brex reaches at-risk customers within hours

Every support ticket is checked for churn risk: a customer who is not just unhappy but actively looking at alternatives, whether because of support, the product or a better offer elsewhere. Each week, 2 to 3% of tickets are flagged.

Brex already reached out to customers after a bad survey score. Now it does the same for churn risk. A team lead or manager checks in with every customer at direct risk, usually within hours of the case closing.

> With CSAT, you have to wait for something to go wrong and for somebody to say something before you reach out. With these cases, we’re getting ahead of it. Within hours, somebody is checking in with the customer, maybe before they’ve even had a chance to submit a survey.

**Medha Mayawala** Business Development, Brex

Related use case: [Renewal Risk Watch](https://www.rippit.com/use-cases/renewal-risk-watch-b2b) →

Use case 2: Competition

### How Brex spots competitor moves as they happen

When a customer mentions a competitor, Brex wants to know why, not just which one. The churn risk analysis records the competitor, the reason the customer is looking, and the customer’s own words.

Brex’s team watches its internal competitive intel channel too. When a competitor started emailing Brex customers with an offer to switch, the same offer started showing up in support tickets, and the team could confirm in that channel how often it was happening.

> We’re able to surface those back into our competitive intel channel and say, yeah, we saw ten of these come through last week. This is a real thing.

**Brian Klaja** Sr Director, Operations Strategy & Enablement, Brex

Related use case: [Competitor Threat Alert](https://www.rippit.com/use-cases/competitor-threat-alert-b2b) →

Use case 3: Leadership and product

### How Brex puts customer evidence in front of the COO and product

Every two weeks, a churn risk report goes to the COO and everyone from the CEO down. It covers what happened, why it matters and what to do about it. Each case is tagged by theme, like onboarding or application issues, with the reason it was flagged and the customer’s exact quote.

> It’s great on a dashboard. But if you put it in a doc and send it to the COO, all of a sudden you see comments directly from the customer, in their exact words. You just can’t ignore that.

**Medha Mayawala** Business Development, Brex

Rippit also feeds Brex’s monthly Customer Pulse report, which ends with fifteen fixes for product and engineering to lower contact rates and raise satisfaction. Building it used to take two weeks a month. Now it takes a week or less.

Related use case: [Churn-Risk Bug Watch](https://www.rippit.com/use-cases/churn-risk-bug-watch-b2b) →

Use case 4: Coaching

### How Brex coaches every agent on everything they handle

Team leads used to coach from a QA score built on three tickets. Now they run a weekly coaching session in Rippit that reads every ticket the agent handled that week, about 30 instead of 3, in 25 to 30 minutes. A short menu of prompts asks what the agent did well, where they can improve, and which cases need attention.

Every ticket also gets a predicted satisfaction score from 1 to 5, and churn risk cases load into each session automatically. Two checks stay manual and yes-or-no: identity and personal data handling, and following Brex’s processes.

“Agents are excited because they feel more fairly evaluated. They’re not getting a small sample of their cases, they’re getting all of them, and we’re looking at behaviors instead of very specific checkboxes,” said Brian.

Related use cases: [Agent Coaching Brief](https://www.rippit.com/use-cases/agent-coaching-brief-b2b), [Predictive CSAT](https://www.rippit.com/use-cases/predictive-csat-b2b) →

Working with AI

## Where do ChatGPT, Gemini and Claude fit at Brex?

Brex uses chat assistants to write better prompts. Medha Mayawala, who leads the work, drafts each prompt and runs it through Gemini, ChatGPT and Claude, having them improve on each other’s versions. Her predicted satisfaction prompt went through 18 or 19 rounds.

Then the prompt goes into Rippit, which runs it on every ticket, every week, and keeps the results in tables that feed coaching sessions, dashboards and the churn risk report. The chat assistant helps write the question. Rippit answers it across all of Brex’s customers.

“She’ll put the prompts across Gemini, ChatGPT and Claude, ask them to improve on each other, put it back in, test it and calibrate. There’s a lot of testing and iteration, but once it’s good, it’s really good,” said Brian.

Across the company

## How did customer insight reach the whole company?

Medha first sent the churn risk report to CX leaders as training material. Word spread. Sales and go-to-market teams asked for it, and product ops started asking whether issues needed to go to engineering. Now it reaches everyone from the CEO down.

Brex didn’t grow its quality team to do this. Rippit took over the manual reviews, which freed Medha to work as an advisor to the business instead of a reporter.

“We’re four months into this journey, and ten times more people know who Medha is and are reaching out to her for information, insights and support than in the past couple of years in this role,” said Brian.

Results

## What results has Brex seen with Rippit?

Brex now hears when a customer is about to leave or a competitor is making a move, and acts while there is still time to keep the customer. Sales, product and leadership get the evidence in customers’ own words.

“In the years we’ve been doing QA, nothing actually came out of it after we presented the scores. People would say, ninety-four, cool, see you next week. The churn risk report became a process within a month and a half,” said Medha.

Retention

### 2–3%Churn risk found every week

2 to 3% of tickets each week show a customer may leave. Each one is a chance to keep a customer before a competitor wins them.

Speed

### HoursOutreach before the customer leaves

A team lead or manager checks in within hours of a flagged case, often before the customer could answer a survey.

Product

### 2 weeks → 1Faster fixes list for product

The monthly Customer Pulse report and its fifteen fixes for product and engineering take a week or less to build, down from two.

Competition

### 10 in a weekCompetitor moves confirmed with evidence

When a competitor offered Brex customers a package to switch, support tickets showed ten in a single week, enough to confirm it was real.

FAQ

## Frequently asked questions

### What is Brex?

Brex is the financial platform for fast-growing companies. From cards to spend management, it serves startups and enterprises at every stage of growth.

### How does Brex find churn risk?

Rippit checks every support ticket for customers who are unhappy and actively looking at alternatives, whether because of support, the product or a competitor. About 2 to 3% of tickets are flagged each week.

### How fast does Brex act on churn risk?

Within hours. A team lead or manager checks in with each customer at direct risk, often before the customer could answer a satisfaction survey.

### How does Brex coach agents without QA scores?

Team leads run a weekly coaching session in Rippit that reads about 30 of each agent’s tickets instead of 3. It surfaces what the agent did well, where to improve, and any churn risk or low predicted satisfaction cases.

### Does Brex still review anything by hand?

Yes, two yes-or-no checks: identity and personal data handling, and following Brex’s processes. Rippit summarizes those notes for managers.

### How does Brex use competitor mentions?

Its churn risk analysis records which competitor a customer mentions, why they are looking, and their own words. Brex shares what it finds with sales and its competitive intel channel.

## Find your next edge

Connect a source, set a growth goal and launch an agent that keeps working toward it. Your competitors are already copying. Start building what they can’t.

### Start in Rippit

Create an account, write a goal like “Tell me which accounts are at risk, and why,” and Rippit guides you the rest of the way.

[Try Rippit free →](https://app.rippit.com/signup)

### Start from your coding agent

Copy this line into ChatGPT, Claude, Gemini or your coding agent. It takes it from there.

`Set me up: rippit.com/setup.md`

Setup steps for AI agents: https://www.rippit.com/setup.md
