The ICP conversation at Series B is almost always the wrong conversation. Companies either try to tighten an ICP that was already too specific, or they pursue expansion plays before they've fully saturated their core market.
Here's how to tell which problem you have.
The signals that tell you to narrow
You're doing too many things for too many types of customers. Your sales team is running different pitches for different verticals. Your content is trying to speak to everyone. Your onboarding has three different flows because three different customer profiles use the product differently — and none of them feel like they were built for a specific person.
Expansion before saturation looks like this: you're closing new segments, but at longer cycle times, higher CAC, and lower NRR than your original cohort. You're technically growing, but the economics are worse with every new type you add.
The fix here isn't to give up on growth. It's to identify the cohort with the best retention, lowest CAC, and highest expansion revenue — and go much deeper into that segment before opening new ones. There's almost always more supply in the best-fit segment than companies think. Most Series B companies haven't saturated their strongest ICP by half.
The signals that tell you to expand
You've got strong NRR, low churn, and fast sales cycles — but the segment is genuinely small. You can see the addressable market ceiling from where you are. New logos in your core segment are getting harder to find, not because the pitch isn't working, but because you're running out of accounts.
This is real ICP expansion territory. The question is: where's the adjacent segment with the most in common with your best customers?
Don't look at market size. Look at similarity. What other job titles have the same problem? What other industries have the same workflow? What adjacent buying committees have the same pain? The fastest expansions happen into segments that look different on a spreadsheet but feel the same to your salespeople.
A useful test: take your three best customers and describe them without using their industry or company size. If the description fits a different segment, that's worth exploring.
The mistake both groups make
Trying to figure this out from the top of the funnel.
If your ICP diagnosis is based on who's coming in — who's booking demos, who's clicking ads, who's filling out your website form — you're looking at demand, not fit. Demand and fit are related, but they're not the same thing. High-demand segments are often the easiest to reach, not the ones that get the most value.
The question you want to answer is who gets value, not who's interested.
Where to actually look
Your best signal is in your existing customer data. Run cohort analysis on NRR, expansion rate, time-to-value, and support load. Look at which customer segments expand without being pushed. Look at which ones have the lowest support ticket volume. Look at which ones refer other customers.
The ICP worth doubling down on is in there — and it's usually different from the ICP you think you have.
One more thing worth doing: talk to customers who churned. Not to understand what went wrong in the relationship, but to understand whether they were ever the right fit. Churn that comes from wrong-fit customers is a different problem than churn that comes from product or service failure. Knowing which you have changes everything about where you focus next.
What to do once you've decided
If you're narrowing, build a scoring model for your pipeline that weights for best-fit signals — the ones that correlate with low CAC, fast cycle, and high retention — and use it to reprioritise prospecting. Cut content that speaks to segments outside the narrowed ICP. Let your salespeople stop learning new pitches for 12 months.
If you're expanding, start with one adjacent segment. Not two. Not three. One. Run a 90-day experiment with a dedicated motion — dedicated content, dedicated outreach, dedicated success resourcing. Measure whether the economics look anything like your core cohort. If they do, double down. If they don't, learn why before expanding further.
Getting this call right is usually worth more than any channel or campaign optimisation. It changes what you build, who you hire, and how you price — not just who you target.
