Why Customers Switch: How to Find the Real Reasons and Reduce Churn

Imagine a standard business case: a company loses one of its customers and tries to identify the causes of switching. Based on its churn data, the reason was price. The account manager says the customer had a budget freeze. The product team has its own version - the customer left because of a missing integration.

All three explanations come from within the company, and all three may sound convincing. Yet, there’s one major “BUT”: none of them came from the customer.

To figure out why a customer really left, companies need to look beyond internal data and assumptions and, when possible, ask the customer directly.

In 1996, Frederick Reichheld, author and business strategist, wrote in The Harvard Business Review that every five years, the typical U.S. company loses 50% of its clientele. He discovered that most CEOs had little understanding of why customers left.

Even though companies have spent decades developing sophisticated customer analytics, many retention teams can measure exactly how many customers they lose and even predict this dynamic for future periods, but still cannot explain why a specific customer left.

Let’s try to analyze the actual reasons for customer churn, not the generic ones.

What is Customer Churn?

Customer churn describes the rate at which existing clients stop doing business with a firm during a specific time period, for whatever cause.

Churn rate is usually shown as a percentage of the customer base lost:

Customer churn rate = (Lost customers / Total customers at the start of the time period) x 100.

For example, if a firm begins its year with 1,000 clients and 50 leave during that year, then the churn rate equals (50 / 1,000) x 100 = 5%.

There are two main ways to measure churn:

1. By number of customers (accounts): How many customers left?
2. By revenue: How much revenue did those lost customers represent?

Each measure tells a different story. For instance, a firm could lose only 5% of its customers, but if these are the company’s biggest clients, the revenue loss could hit 20% or even more.

When you measure churn, it’s crucial to look for both the number of clients that left and how valuable those clients were.

Companies can track churn weekly, monthly, or annually – it all comes down to the type of business.

It is also useful to compare your churn rate with your previous results or industry benchmarks. Thus, a sudden increase can signal certain issues with the product, service, pricing, customer experience, or sales process.

Example:

A company that tracked churn rate each month lost 300 of their 75,000 customers. That means their monthly churn rate would be 0.4%.

Voluntary vs. Involuntary Churn

Not every client that a firm lost made the decision to leave. Sometimes people’s cards expire and they cannot make a payment. This is called involuntary churn.

Failed payments and billing failures account for 20 to 40% of total churn in subscription businesses, according to ProfitWell via GoCardless.

Reasons For Customer Churn: The List Everyone Gives You

If you type "reduce customer churn" into a search bar, you'll likely come across some version of the same eight reasons:

1. Wrong-fit customers acquired in the first place
2. Poor onboarding and slow time-to-value (a client signs up but doesn't experience the payoff quickly enough and loses patience)
3. Unresolved support failures
4. A competitor offering something you don't
5. Product gaps or unreliability
6. Price
7. Champion turnover (the person who backed you internally left)
8. Involuntary churn from failed payments.

For instance, 44% of churned customers say they weren't achieving the outcomes they expected (note that this does not mean "the product was missing the feature X," but rather "I bought this to reach a certain goal, and it never got me there").

The list's flaw is that it applies to all businesses and actually doesn't tell you much about yours.

Why Customers Leave: Trigger vs. Root Cause

Reasons that customers provide may not always be the ones explaining why they left.

When you lose a customer, your Customer Relationship Management (CRM) records provide reasons that read “too expensive,” “missing feature,” “chose a competitor," or “budget freeze.” And though it sounds useful, these reasons may only describe the final trigger but do not offer details on the things that led to the decision.

A 2023 study emphasizing customer churn in retail banking found that lost customers gave an average of 2.2 reasons for leaving, which demonstrates that churn is more complex than a single complaint or event.

Another thing worth noting is that companies and clients do not always see the same story.

PwC's research has found a gap between what executives believe regarding why clients leave and the reasons the customers themselves provide.

Executives were asked about the major reasons why clients left their company in the past year, while consumers were asked about the reasons they stopped using/purchasing from the firm/brand.

Bain & Company has spent years examining why customers leave and found that when you ask executives why a customer decided to leave, they typically point to the final event before departure. However, that event is rarely the actual cause; it is merely the trigger—the final straw.

Experts at Bain found that customer churn often builds up over time:

1. Pre-churn: Something triggers an issue (ex., an unresolved product issue).
2. Root causes: The client’s experiences repeatedly led to frustrations.
3. Tipping point: One or more issues make the customer consider leaving.
4. Final trigger: The client made up their mind to leave after a certain event (price increase, outage).
5. Churn: The customer leaves.
6. Post-churn: What happens afterward can affect whether the customer might return.

That is why a useful churn analysis should not stop at “Why did the customer leave?” but instead should ask the question “What happened before that made leaving the logical choice?” which is where the root cause usually becomes visible.

The Limits of Customer Churn Analysis

Analytics tools can provide a lot of useful data that you can use, allowing you to learn about who is using your product, when they use it, and much more. They can also warn you that an account may be at risk before a person notices the problem.

But analytics cannot reliably provide you the exact reason why something is happening.

For example, if a customer's log-ins drop by 40%, the data alone won’t provide you any concrete reason. It could be that the client found a better tool, or maybe they are still happy with your product, and it’s just that they simply do not need to use it as often.

The Exit Survey Response Rate Problem

Survey responses represent just a tiny fraction of customers and that’s the main reason they should be treated with caution. Say 10 out of 100 clients who left decided to answer your exit survey, and seven out of 10 mentioned that "the price was too high." The sample consisting of seven people out of 100 is insufficient for solid conclusions. .

Email exit surveys typically return 6-15% responses; NPS surveys* run 5-15%.

*NPS (Net Promoter Score) is a survey that asks clients how inclined they are to suggest a business or product to others.

Early Warning Signs Worth Tracking Anyway

Falling usage, an increase in support escalations, champion exit, payment failures, and a silence during renewal outreach - treat all these as flags that should trigger a conversation (a flag tells you when to look, not what you'll find).

Predictive models (machine learning/statistical scores based on historical account data) can be really useful if you wish to forecast clients at elevated risk. But these are not designed to explain why but rather to spot statistical patterns and correlations in observable data (usage, support, etc).

Suggested reading: Types of Survey Questions: A Guide for Market Researchers

How to Find The Real Reasons Customers Switch

The person who made the decision to leave is the only one who can provide true reasons regarding their decision, and the only way to learn about it is to engage in a conversation with that person.

Whom to Interview (And It's Not Just Your Former Champion)

If a person clicked "cancel," that doesn't necessarily mean that it was their sole decision.

In B2B, when a company decides not to use one product and switch to the other, the decision is usually not made by just one person but through a collective agreement often encompassing multiple reasons. Thus, the list of users who might have taken part in this decision-making process can include:

1. The economic buyer (the person controlling the budget).
2. The day-to-day user.
3. The procurement or IT specialist (people who may evaluate the price, contract, security, technical requirements, etc.)
4. The champion (the person inside the firm who supported your product).

For example:

Your company loses a client, and you ask your main contact for the reasons they left. The answer is, "The product didn't have the integration we needed." But your main contact may not be aware of the full picture. What if:

1. The CFO thought the product was too expensive for the value it provided.
2. The IT dept. rejected it because of security concerns.
3. The CEO preferred a competitor.
4. Procurement negotiated a better price elsewhere.

How Many Interviews and Over What Window

Eight to 12 interviews is a good starting point for a single, well-defined churn segment, and 15 to 25 interviews if you're spanning multiple segments or geographies.

Guest, Bunce and Johnson's 2006 study found that in homogeneous groups, most patterns were starting to get noticed after the first dozen interviews.

Two Cohorts And What Each Can And Can't Tell You

A Worked Example, End-to-End

Note: This is an illustrative representation based on common B2B retention patterns - no figure in this example should be used as a benchmark.

1. Situation:

A mid-market SaaS team spotted churn concentrated in accounts between months nine and 14 of their contract, which is a rather large cohort to focus on and quite distinct from the average, so it's worth investigating on its own.

2. Hypothesis:

The company's CRM said that the main reason customers left is the price, but the team doubted it for two reasons:

• Pricing didn’t undergo any change in that window.
• Many customers were leaving after (more or less) the same amount of time.

3. Recruit (Interview candidates):

The firm interviewed 12 people from recently churned customers, including both decision-makers and everyday users rather than speaking only to the main contact. Each person took part in a paid 30-minute interview conducted by an independent party (so customers would be more likely to speak honestly).

4. Interview (Questions to ask):

The team asked, "What could we have done differently, and what would have made you stay?”

Several clients gave a similar answer - they had planned to expand their use of the product in the second year, but they couldn’t get internal budget approval. The reason is that the vendor hadn’t provided enough evidence or materials to help them convince their management that the additional spending was worthwhile.

5. Code (Interview analysis):

Two people reviewed the interview answers separately and grouped the reasons for churn into several categories. But not every churned customer has one single reason for leaving, sometimes several problems contribute to the decision.

6. Assign (Identifying the responsible unit):

The root cause turned out to sit with sales enablement, not customer success. The issue was that customers' internal champions were not offered the materials needed to convince their managers to renew the product.

7. Outcome (Changes in business):

The company created a renewal-readiness kit - a set of materials designed to help the customer's internal champion justify renewing the firm’s product.

• They gave it to customers before the renewal risk became serious.
• They tracked whether the next group of customers was still retained at the same point in their customer journey.

Instead of simply looking at overall churn, they asked, "Are customers still staying around month 10 after we introduced this change?" This way, they could see whether the new approach actually worked.

The Customer Exit Interview Problem

Even though companies try customer exit interview programs, these fail for several reasons:

Failure 1: Former clients have no reason to talk to you.

They've already left and a call from your company will likely look like a disguised save attempt. So pair the call with compensation for their time - typical rates are $25-$50 for consumer participants versus $75-$150 for B2B experts, and higher for senior roles.

Failure 2: Your own team is the worst possible interviewer.

A CSM asking a customer why they left is the same as asking a person to criticize you to your face with nothing to gain from honesty. So, use third-party interviewers.

Failure 3: Asking for a reason instead of a reconstruction.

Don't just ask, “Why did you leave?" It's better if customers walk you through how the decision happened: when did they first start considering alternatives, what options did they compare, and what ultimately led them to switch?

What interviews still can't tell you

Even though customer interviews are useful, they have their limitations.

For instance, people don’t always remember when they decided to leave, and the reasons they provide may be different from the actual ones (and they might not even realize that).

Those who agree to an interview cannot represent every person who left, which is why interview findings should be compared with your actual customer data, such as product usage and billing history.

If what a customer says doesn't match the data, don't simply dismiss one of them because the difference may provide some insights worth looking into.

Suggested reading: How to Conduct an Expert Interview: A Complete Guide For Insight-driven Research

Customer Exit Interview Questions That Get Real Answers

Timeline: “When did you first consider looking into alternatives, and what was going on at the time?”

Instigator: “Who brought up the subject of switching initially, and what were they reacting to?”

Trigger: “Was the decision finalized at a particular point in time?”

Expectation gap: “When you signed up, what did you anticipate that didn't work out?”

Alternative: “What do you currently use, and does it perform better or differently than we did?”

Decision structure: “Who had to sign off on the change, and what mattered most to them?”

Friction: “What made leaving worth the hassle of migrating?”

Counterfactual: “What would we have had to do, and by when, for you to have stayed?”

Blind spot: “What do you think are our reasons why you left that are not quite right?”

Strategies to Reduce Customer Churn Once You Know The Motives

Here's how to turn the insights you gained into something you can work on:

Sort your findings into clear categories:

1. Create categories based on what customers actually said.
2. Give each account (client) one main cause, plus extra factors if needed.
3. Have two people sort the same interviews on their own, then compare notes. If they disagree on more than one in five (or agree on 80%), your categories need work.
4. Mark how confident you are in each label, because if confidence is low, you need to interview more people before reporting a finding.
5. Once your categories hold up, use them to replace the free-text reason in your CRM's churn field.

Decide what to fix first

• The first to fix are the things that affect the most clients and revenue. Each issue should be fixed by a specific team - sales problems need a sales fix, for instance.
• Start with failed payments, since they account for 20-40% of all churn.
• Most Customer Success (CS) teams don't have a clear plan for reducing customer churn, but your interviews already reveal which customers might have been saved.

Check if it worked

Decide how you'll measure success before you make any change. Choose one group of customers and one number that should move if the real cause was fixed, not your overall churn rate (which moves too slowly and unevenly to prove anything).

It's worth the effort: according to Bain, a 5% gain in retention can raise profit by 25-95%.

Frequently Asked Questions (FAQ)

What does a 20% churn rate mean?

It means that 20 out of every 100 (or twenty percent of revenue, if measured by revenue) either stopped using a service or product during the period measured. Whether that's alarming depends on the time period - losing 20% of customers every month is a serious problem, but losing 20% over a year is much less severe, although whether it is acceptable depends on the industry.

What is a good churn rate?

Recurly's data (as of July 2026) puts median annual churn for B2B SaaS at around 3.2%. But churn varies by customer type, contract length, and price. There is no single "good" churn rate. What matters is how your rate compares with similar companies.

How is this different from win/loss analysis?

Customer churn rate estimates the number of existing customers who leave, answering the question, “How many customers did we lose?" Win/loss analysis looks at why clients chose you or a competitor when making a purchase decision, answering the question, “Why did we win or lose the deal?”

Example:

A SaaS company loses 10% of its customers this year. Churn analysis shows how much business was lost. Win/loss analysis can investigate why its customers chose a competitor instead or why new prospects chose the company.

Suggested reading: Win/Loss Analysis: The Tool That Reveals Why You Really Win (or Lose)

Final word

The eight reasons listed in virtually every churn article are nearly the same because they describe churn in general. Yours are specific and specific answers come from people, not dashboards. The latter tell you who left and when, but only lost clients can provide you the exact reason why. Remember that people are much more sincere when talking to someone they don’t work for you.

ENC connects you with multiple Expert Networks through one platform, making it easier to source the right experts for churn interviews and wider market research.

The 22 networks on the platform provide combined access to 2M+ experts, including people who can offer perspectives beyond your own customer base. You can work with multiple networks as a single ecosystem, with no subscription or prepayment, while keeping your research workflow in one place.

Launch your project