The Wrong Customer Problem: Why Your Highest-Impact Churn Fix Isn't in Post-Sales

Every CS leader I know has a version of the same story. It starts with a customer who looked great on paper. Decent ARR, signed a multi-year deal, showed up enthusiastic to kickoff. Fast forward six months: adoption is flat, the champion has ghosted, and the health score is doing that slow drift from green to yellow that everyone pretends not to notice.

Then they churn. And in the post-mortem, the CS team gets the questions.

“What happened here? Did we miss something in onboarding? Was the CSM not proactive enough?”

I’ve sat through enough of these conversations to know what usually doesn’t get said: this customer was never going to succeed. Not because the CSM dropped the ball. Not because onboarding was slow. Because they were the wrong customer from the start. This is what I’ve started calling inherited churn — churn that CS inherits from a targeting or qualification decision made months before the CSM ever touched the account.

The Inconvenient Math

Here’s the thing about churn: we treat it like a post-sales problem because that’s where it becomes visible. The customer cancels on a CSM’s watch, so it lands on a CSM’s scorecard. But the root cause often traces back months, sometimes a full year, to a sales conversation where someone decided this account was worth closing.

The data backs this up. Churn is fundamentally a company-wide problem, not a CS-specific one. Misaligned expectations set during the sales process emerged as a major churn driver in recent industry surveys, and one that no amount of post-sale heroics can reliably fix. The conclusion was blunt: when sales targets customers who aren’t the right fit, churn becomes nearly inevitable regardless of CS effort.

This isn’t a fringe take. As one practitioner put it: “Sales closes a product that isn’t a good fit, and customer success gets blamed when those customers leave.”

Meanwhile, the benchmark data tells its own story. Agile Growth Labs’ 2025 analysis of over 1,200 SaaS companies found that the average SaaS churn rate sits at 4.1%, split into 3.0% voluntary and 1.1% involuntary. But those averages mask enormous variation. Companies pursuing growth-at-all-costs strategies, chasing volume over fit, attract users who were never likely to stay in the first place. Contract structure matters too: annual contracts show 8.5% churn versus 16% for month-to-month, which tells you something about the correlation between buyer commitment and retention.

The ICP Misalignment Tax

I think of bad-fit customers as an invisible tax on the entire post-sales organization. The ICP misalignment tax is the cumulative cost that wrong-fit customers impose across CS, support, and product teams — not just the direct revenue loss from early churn, but the compounding drain on bandwidth, morale, and expansion capacity that nobody tracks on a single line item.

Every hour a CSM spends trying to save an account that was never going to work is an hour they’re not spending on expansion opportunities with customers who actually love the product. Every support ticket from a customer who’s stretching your tool beyond its design center is a ticket that drags down response times for everyone else.

The cost to serve a bad-fit customer — the fully loaded cost including CS time, support overhead, engineering escalations, and opportunity cost — is staggering if you work through the math. If your average customer lifetime value is $50,000 but poor-fit customers only generate $18,000 before they leave, you’re losing $32,000 per misaligned account. Multiply that across a couple dozen wrong-fit customers per quarter, and you’re looking at hundreds of thousands in lost revenue. That’s not counting the opportunity cost of all that CS bandwidth.

There’s a metric distinction here that matters. Most teams track net revenue retention (NRR), but gross revenue retention (GRR) is actually the purer signal of ICP fit. GRR strips out expansion revenue and shows you how much of your existing base you’re keeping. If your NRR looks healthy at 108% but your GRR is sitting at 82%, you’re papering over a fit problem with expansion from your best accounts. The bad-fit customers are dragging down your base while your good customers are compensating through upsells. That works until it doesn’t.

And then there’s the morale tax, which is harder to quantify but just as real. Nothing burns out a CSM faster than the feeling that they’re being set up to fail. You can only attend so many kickoff calls where you immediately see the red flags — wrong use case, wrong buyer persona, wrong expectations — before you start questioning whether anyone upstream is paying attention. I’ve watched talented CSMs leave the profession entirely over this. Not because they couldn’t do the job, but because the job as constructed was impossible.

The “Sales Will Be Sales” Problem

Here’s where this gets uncomfortable: the incentive structure in most SaaS companies actively encourages selling to the wrong customers. Quota attainment is measured on bookings, not retention. A rep who closes 30 deals, 10 of which churn within a year, often gets promoted faster than a rep who closes 20 deals that all renew and expand.

I’m not blaming individual reps. They’re playing the game the way it’s designed. If the comp plan rewards logos and first-year ACV with no clawback for early churn, you’re going to get logos and first-year ACV. The system is working exactly as intended. It’s just that the intention is wrong.

Kyle Poyar’s research at Growth Unhinged, analyzing 3,500 software companies, illustrates the long-term consequences clearly. There’s a strong correlation between net revenue retention and sustainable growth. Companies with low retention are three times more likely to be shrinking than growing quickly. The companies that make it to $5 million ARR and beyond consistently have substantially better retention rates than their early-stage peers. You can’t outrun a bad fit with new logos forever.

OpenView nailed this years ago in a piece that should be required reading for every CRO: one portfolio company saw annual retention rates vary from 50% to 90% across different customer types for the same product. Same onboarding, same CSM team, same product. The variable wasn’t customer success execution. It was who they were selling to.

Five Red Flags You’re Selling to the Wrong Customers

If you’re a CS leader, you probably already know this intuitively. But intuition doesn’t change pipeline strategy. Here’s how to make the case with evidence your leadership team can’t ignore.

First-year churn is disproportionately high. If customers who churn are overwhelmingly in their first 12 months, that’s not an adoption problem. It’s a fit problem. Customers who are the right fit tend to survive the messy early period because the underlying value proposition is real. We looked at our own data last year and found that 71% of our churned accounts had been customers for less than 14 months. The accounts that made it past that threshold renewed at a rate that was almost unrecognizable compared to the first-year cohort.

Churned customers cluster around specific characteristics. Run a cohort analysis on your last 20 churns. I’d bet good money they share two or three attributes: a particular industry vertical, company size, use case, or buyer persona. That’s not a pattern CS can fix. That’s an ICP problem.

When we ran this analysis, we found three segments that accounted for over half our churn but less than a quarter of our ARR. Those segments had been growing quarter over quarter because sales was hitting them hard. Nobody had connected the pipeline composition to the retention numbers until I put both datasets on the same slide. The reaction from our CRO was telling: not anger, but genuine surprise. The data had always been there. It just lived in two different dashboards that nobody looked at together.

Support volume is wildly uneven. This one is easy to check and hard to argue with. If 20% of your customers are generating 60% of your support tickets, look at who those customers are. Odds are they’re using your product for something it wasn’t designed for. They’re filing tickets that reveal a fundamental mismatch between what they expected and what they got. When I dug into our Zendesk data, the correlation between high-touch support accounts and eventual churn was 0.73.

Your NPS is bimodal. A healthy chunk of 9s and 10s alongside a troubling number of 3s and 4s means you likely have two customer populations: those for whom your product is a great fit, and those for whom it isn’t. The detractors aren’t giving you feedback about your product. They’re telling you they shouldn’t be your customer.

CSMs can call churns at kickoff. This one stings, and it’s the signal I avoided for too long because it implies a problem nobody wants to own. If your experienced CSMs can look at a new account in the first two weeks and predict with reasonable accuracy whether it’s going to make it, you have a targeting problem, not a delivery problem. I asked my team to informally score confidence on 40 accounts that came in over two quarters. Their predictions correlated with actual outcomes at a rate that made me uncomfortable. That institutional knowledge, the pattern recognition that lives in your most experienced CSMs’ heads, needs to flow back upstream. Most of the time it dies in a Slack DM or a hallway conversation.

What CS Leaders Can Actually Do About This

Identifying the problem is the easy part. The harder question is what to do when you’re a CS leader who doesn’t control the sales pipeline.

Build the churn attribution model your company doesn’t have

Most organizations track churn by reason code: “budget constraints,” “competitor,” “product gaps.” These are symptoms, not causes. A churn attribution model traces churn back to pre-sales characteristics — how the deal was sourced, what was promised in the sales cycle, how closely the customer matched your ICP, and how long the sales cycle was.

When you can show the CFO that customers outside your ICP churn at 3x the rate of those inside it, the conversation shifts from “how do we fix CS?” to “how do we fix revenue quality?” — the concept that not all ARR is equally durable or valuable. I built this model about a year ago using nothing more complicated than a pivot table matching our churn list against the original opportunity data in Salesforce. We also started assigning a closed-won quality score to every deal at the 90-day mark, rating ICP fit based on actual onboarding behavior rather than what the sales cycle predicted. The output was a single slide: ICP-fit customers had an average lifetime of 38 months and an LTV of $127K. Non-fit customers averaged 14 months and $43K. Same product. Same onboarding. Same CSM team structure. The CFO asked me to present it at the next board meeting. That had never happened before.

Get a seat at the deal desk

If CS doesn’t have visibility into what’s being sold and to whom, you’re always going to be cleaning up messes you didn’t make. Push for CS involvement in deal review for accounts above a certain threshold. Not to veto deals. That’s political suicide. But to flag risk and ensure handoff notes are honest about what was promised.

I’ve tried this three times. Failed twice, succeeded once. The first time, I positioned it as “CS quality gate on new deals” and got shut down in one meeting. The second time, I tried an end-run by going directly to the CEO, which succeeded at getting attention and failed at everything else. The time it worked, the CS leader framed it as “protecting the customer experience,” not “policing sales.” The distinction matters more than it should. Nobody wants to be the department that slows down revenue. But a five-minute CS review that flags a deal as high-risk — and routes it to a CSM who’s equipped for that profile — is very different from a veto gate.

One thing I got wrong early on: I tried to make this about individual deals. That doesn’t work. Sales leaders will fight you deal by deal because they have context you don’t about each one. What works is aggregate data. “Deals sourced from this channel churn at 2.1x the rate of deals sourced from that channel.” That’s a pattern, not an accusation.

Make the economic case, not the emotional one

“Sales is closing bad deals” doesn’t land in a boardroom. “Customers outside our ICP have a 14-month average lifetime versus 38 months for ICP-fit customers, representing $2.4M in preventable annual churn” does. Tie every argument to LTV, CAC payback, and net revenue retention. Those are the metrics your CEO and board actually optimize for.

The emotional case feels righteous. I’ve made it. It doesn’t work. The economic case feels cold, but it moves budgets and changes comp plans.

Create the feedback loop that doesn’t exist

Christina Kosmowski, CEO of LogicMonitor and former COO at Slack, has talked about the importance of embedding customer success thinking into the cultural DNA of a company rather than siloing it in a department. At its best, this looks like CS insights flowing back into ICP definition, sales enablement, and product strategy. The CS team sees which customers succeed. That intelligence shapes who gets targeted next.

We’re about eight months into building this at our company. It’s slower than I expected. The sales team was initially defensive, which I understand. But the quarterly ICP review, where CS presents retention data segmented by customer characteristics, has started to shift the conversation. Last quarter, our VP of Sales voluntarily adjusted targeting criteria for two segments based on our data. That felt like a turning point.

The Uncomfortable Truth

I know what some of you are thinking: “This sounds great, but my CEO just wants growth.” Fair. And I’m not suggesting CS leaders walk into the next all-hands and declare that sales is broken.

What I am suggesting is that the most impactful thing many CS leaders can do for retention isn’t building a better QBR template or implementing a new health score algorithm. It’s building an airtight, data-backed case that churn is a GTM targeting problem disguised as a post-sales execution problem, and then making that case to the people who can do something about it.

The benchmark data, the industry research, and frankly the lived experience of every CSM who’s ever inherited a doomed account all point in the same direction: you can’t onboard your way out of a bad fit. The most effective churn prevention doesn’t happen in the first 90 days after close. It happens in the qualification calls before the deal is even in the pipeline.

CS leaders who make this case effectively don’t just reduce churn. They earn a strategic seat at the table because they’re the ones who can prove, with data, which customers the company should be pursuing and which ones it should be walking away from. That’s not a support function. That’s a growth function.

And if your company isn’t ready to hear that yet? At least you know where the real problem is. That’s the first step toward fixing it.