A version of this article previously ran in GigaOm.
In my career, I’ve worked with a number of SaaS startups to help them refine the efficiency of their operations. One of the things that is often neglected is customer churn. Early focus on churn helps build discipline that becomes even more important as a company grows to $100M or more in revenues and can be a huge factor in driving growth.
In the bad old days of on-premise Enterprise software, a startup was considered to have traction when it got its first dozen paying customers. These were typically large six figure deals, and required a lot of heavy lifting to get the customers into production. Not surprisingly, every employee knew who the early lighthouse customers were and what they needed. While the old school field sales model was not particularly efficient, it had the virtue of driving an organization towards making early customers successful.
The irony of high-volume SaaS sales is that when you have hundreds or thousands of customers, things can become rather anonymous. The more customers you have, the harder it is to keep a pulse on what people are really doing with your product. As a result, it’s harder to make sure customers are using your product correctly. And if customers aren’t successful in using your product, they churn.
Having churn is like rowing a leaky boat. After a while, you spend more time bailing water than rowing and you’re not going to make much forward progress. By contrast, organizations that focus on reducing churn will find that their revenue growth gets that much stronger every quarter.
Break it Down
To accurately measure what’s going on, you should begin by breaking out churn (customer cancelation) from contraction (a downgrade in spending). Some customers may downgrade in accordance with seasonal business, such as is common in the retail sector. You should measure downgrade and churn separate from uprades and expansion; otherwise your net growth numbers will mask problems that are bubbling below the surface. If your SaaS product has different editions (e.g. Basic, Pro, Enterprise) you should watch for downgrades that suggest customers are not seeing the value in the higher-end features.
It’s also worth paying attention to churn and contraction by customer segment. In most SaaS businesses, churn is highest in low-end customers. Some of those smaller customers will inevitably be acquired or go out of business. Over time, if you move upmarket to larger SMB or Enterprise customers, low-end churn becomes less significant. Customers who spend a lot of money usually have greater commitment and resources for working through any speed bumps during implementation. They’re also far less likely to switch to another vendor with newer features.
Understand the Reasons
The most important thing is to understand why customers are downgrading or churning. While it’s easy to speculate, the best approach is simply to ask your customers through a combination of surveys, emails and ideally phone calls or face-to-face meetings. Remember, your purpose in doing this is to understand the customer’s experience, not to second guess them.
While sometimes founders get defensive with customers, recognize the value of the feedback you’re getting, even if its negative. And make sure that this information is shared within the company to help come up with ways to improve your offering and provide a better experience.
Classify your findings to determine whether it was a customer issue, a sales issue, a product issue or some external factor. Did the customer not understand what they were buying? Did they lack the skills to implement your product? Was it missing key features? Did users find it too hard to use? Were there quality or reliability issues?
Your first line of defense in reducing churn is to make sure your product lives up to its marketing claims and that it quickly delivers value to the customers. The more people a customer uses your product, the less likely they are to churn.
Depending on your product, it may be worth adding some introspection to help you determine usage patterns that suggest a customer is at risk of churning. Some, like less frequent log-ins, are obvious. If you’ve got thousands of customers, it may be worth doing some analysis to come up with less-obvious predictive signals.
Human Touch Goes a Long Way
At Zendesk and MySQL, we implemented several customer programs to reduce churn and benefited from below-industry churn rates. One of the key principles was that I wanted to avoid being the kind of company that only calls up customers when seeking a renewal. Instead, we built a Customer Account Management (CAM) team that worked with customers over the entire lifecycle.
We defined a twelve-month program with regular check-ins, emails and phone calls to make sure customers were successful in using the product. The Customer Account Manager’s role was to be proactive with customers, identify any issues that were getting in their way and build a relationship of value and trust. Through this relationship, we earned the right to additional expansion and upgrade opportunities within the customer and through referrals.
The Customer Account Managers also kept their eyes peeled for issues that might suggest the account was at risk of churn, such as if a new manager was hired, or if certain features weren’t being used. They kept accounts up to date with information about best-practices for customizing their setup, implementing new features or learning about training that might be helpful to them, and so on. When necessary, the Customer Account Manager could call on resources in Support, Engineering or elsewhere to solve a customer’s issues. The account managers were also a great first line of communications on those rare occasions when there was downtime.
Occasionally, fast growing customers would outgrow their initial setup. Maybe they went from a dozen users to several hundred without putting in place the proper controls or reporting. Or in some cases the person who set things up never read the documentation and did a lousy job. Regardless, we would work proactively to help customers through these situations. While this meant undertaking some modest amount of services for free, the alternative was far worse: an unhappy customer who would inevitably blame our software. In general, by being proactive, we could turn these situations around and win the ongoing loyalty of what would otherwise have become a problem customer.
Towards Negative Churn
In a rapidly growing business, you may experience “negative churn” meaning that your upgrades and expansion more than make up for contraction and churn. But even in this kind of scenario, it’s vital to pay attention to churn. Failing to do so leaves you blind to trends of customer dissatisfaction that may indicate fundamental problems with your product or your business model. Negative churn can easily mask an eroding customer base that leaves your business vulnerable. In a fast moving boat, even small leaks are worth patching.
While some modest amount of churn is inevitable, you should strive to get your churn rate as low as possible and then continue to monitor any changes. If your churn rate is more than 10% annually, you have work to do.