From 63% to 92% retention

Hi Everyone,

Segment is a customer data platform. Two years ago, the company was losing about a third of its enterprise revenue every year. They fixed this by narrowing who they sold to. By the end of those two years, they were keeping 92% of enterprise revenue, and total revenue had grown 150%.

When growth slows, the common move is to broaden the customer base. Segment went the other direction.

Today, we'll walk you through how to see this coming and what to do about it, using Segment's playbook and a few practices any team can run.

How Segment did it

Before the refocus, Segment sold broadly into enterprise and gave away free consulting to win deals. About a third of their enterprise revenue churned every year.

The team stopped to figure out which customers got the most value from the product. The answer was large B2C companies with multiple brands and subsidiaries, run by a strong internal data leader. Engineering then built only what those customers needed. Sales stopped chasing inbound technical buyers and started calling on companies that matched the new profile.

Their catalog of pre-built integrations grew from roughly 300 to 3,000 in two years.

Spotting the same problem in your business

The signs are operational at first:

• Sales cycles get longer: M Accelerator looked at deals from growth-stage SaaS companies. Deals with customers outside the original ICP took 47 days to close, compared to 12 days for customers who fit. The looser the ICP gets, the longer every deal takes.
• Acquisition gets more expensive: Customers outside the ICP cost 2.3 times more to acquire. Their lifetime value runs 40% lower.
• More than one in five recent deals comes from outside the ICP: Above that line, the team is closing customers who shouldn't be customers.

If your gross revenue retention is below 90%, the same issue is usually behind it. Bigger contracts from your best customers can hide the problem for a while, while the worst-fit customers leave.

Four practices that keep ICP focused

Reviewing your ICP once a year isn't frequent enough to stop drift. Four practices help fix that.

A monthly review with sales, product, and customer success leaders: Take your last 50 deals and group them by how well each customer matches your ICP.

For each group, check the time to close, the cost to win, how customers use the product, and the support load. Problems show up in these numbers months before they show up in renewals.

A traffic light system for target accounts: Stage 2 Capital tags every prospect as green, yellow, or red.

🟢 Green accounts get 80% of sales time because they're the best fit.

🟡 Yellow accounts are experiments. The sales team can work them but only 20% of the time, and always with a clear goal and deadline.

🔴 Red accounts get turned away.

A list of customers you won't sell to: Write down which segments you'll refuse. Be as specific as you are about the segments you want. Put those rules into your CRM so deals that look tempting but rarely renew get caught early.

A quarterly update: Every three months, rewrite the ICP definition. At the same time, update lead scoring rules, deal routing, and sales compensation. If the doc changes but those systems stay the same, nothing will change.

Try this today

Pick one customer segment that takes more from your team than it gives back. Pull up the last three deals you closed in that segment.

If you wouldn't take those deals again, knowing what you know now, write that segment into the list of customers you won't sell to.

Go deeper

👉 M Accelerator: The ICP Drift Problem – A cohort analysis showing what out-of-ICP customers actually cost, plus a monthly review playbook.

👉 a16z: Getting Ready to Move Upmarket – The full Segment case study, including how they defined the new ICP and rebuilt their sales motion.

👉 Stage 2 Capital: Refine Your ICP to Kick Off 2024 – Templates for the green/yellow/red account system.

👉 a16z: A Framework to Define and Refine Your ICP – Five questions for figuring out who your best customer actually is.

Coming up tomorrow

Tomorrow, we're walking through the 13-week cash flow forecast, the tool that allows you to catch cash problems weeks before they materialize.

Thanks for reading!

P.S. If your ICP hasn't been properly reviewed in over 12 months, forward this to your sales and marketing team.