How big of a problem is customer attrition?
Customer churn is a reality all companies have to face. But managing and keeping that as low as possible makes a massive difference in how a company has to focus its time and effort.
Even if a firm has a fairly benign-seeming churn rate of around 5-10% as is the average for SaaS companies, it puts constant pressure on companies to recover before they can think about growth.
Consider two companies, one with an annual churn rate of 5%, another with a 10% churn rate. Both businesses have a modest growth goal of 7%. The company with 10% annual churn has to make over 30% more new sales than the company with 5% annual churn in order to reach the same growth rate of 7%.
That’s 30% more resources for lead generation, sales, and sales support. On the other hand, the company with 5% churn can allocate those resources to product enhancements, customer support, or even pursuing more sales. The most profitable sale comes from existing customers.
Why don’t customers repurchase?
At its core, customer attrition is an experience problem. It is easy to assume that, for example, customers simply found a better price from a competitor. But in this case, the central issue is not that the customer didn’t pay the right price, the issue is that the customer didn’t get the experience he was expecting when he paid that price. There are many ways that companies can mis-frame the experience that customers should expect at each stage of the customer lifecycle:
Companies can only address the problems that they are aware of. Major issues, like product failures or a poor service reputation are easy to realize. But the types of problems outlined above can be very difficult to detect.
Often these problems are hiding just below the surface. These experience problems probably wouldn’t turn up in customer surveys. But companies have a massive source of passive customer intelligence (hyperlink to previous article) within every email, social media post, chat session, and phone interaction. Customer contacts to the call center are the best source of intelligence on what is going wrong with customers’ experiences. Today, some businesses have been able to mine the text of written feedback from emails or social media posts. But the largest source of customer intelligence- speech from phone calls - remains largely untapped.
How speech analytics can reduce customer churn
Using speech analytics, companies are able to use calls to the contact center - typically viewed as a cost to the organization - as a resource for fueling growth by reducing customer churn. Cutting-edge technology like automatic speech recognition (ASR) and natural language understanding (NLU) use machine learning and artificial intelligence to transform audio of speech from calls into structured data that’s ready for text analytics. As examples, consider how speech analytics can solve customer defection at each stage of the customer lifecycle:
Marketing - The marketing message communicates the wrong value proposition
A cell network provider’s marketing campaign for a rate plan targeted at business travelers implies - by using the imagery of a business person traveling on an international airline - that the rate plan is the best value for international travelers, when in fact the rate plan offers in-plane perks. Mining speech data, the company finds that customers that are cancelling mention key-phrases like “high international rates” or frequently mention neighboring countries’ names. The marketer can adjust the marketing campaign to include an illustration of in-flight connectivity with a focus on productivity, and track to see how customers respond by analyzing keyword frequencies.
Sales - The salesperson over promised on the product or service’s capabilities
A SaaS company’s salespeople assure customers prior to onboarding that minimal integrations will be needed with company databases. But in reality the integrations are often fairly complex for their customers’ small IT departments. The company’s product team is surprised to hear customers say that the integration is more “time consuming” than they thought. Company analysts using speech recognition software recognize the pattern amongst customers that don’t renew, and discover the expectation-setting problem from the salespeople. From there, the sales discussions are adjusted to better prepare clients for the integration required.
Product - There is a seemingly minor malfunction or bug that turns customers off to the product
A mobile app from a major media company combines different sources of news media. The app has an undetected bug that causes the app to not refresh with breaking news quickly enough - certain sources take up to 45 minutes to be loaded into the feed. Customers that call to complain or cancel say frequently that they use “other sources” for breaking news because of the “update speed.” Focusing in on those keywords, company analysts can identify the problem, get it fixed, and track keyword frequency to ensure that the problem was resolved and that the fix reduces customer defection.
Service - the company couldn’t quickly & easily solve customers’ problems with their product or service
An online clothing store notices that its customers defect at an abnormally high rate when they return products. The company uses speech analytics to identify that customers say they “can’t find” their order numbers, which are required for returning products, and it often takes representatives several minutes to find the number while the customer is on hold. The organization shifts the way it sends and stores order numbers for quicker & easier resolution with customers returning products.
In each of these examples, if the solution proposed doesn’t reduce the frequency of the negative keywords, then the organization can continually test new solutions with a small subset of customers until it finds one that works; a method known as A/B testing.
Using A/B testing, companies can test solutions in two different environments: one that has an experimental proposed solution, and another that uses the current approach. A small percentage of customers are randomly selected to be offered the experimental solution, while the rest receive the current offering. Then, the performance of the new solution is compared against the current solution by comparing metrics on keyword frequency and customer turnover. This approach gives companies an unbiased, scientific way to evaluate how new solutions affect customer retention.
Using this kind of testing, businesses can not only identify if their solutions reduce the amount of negative keywords, but also pinpoint which negative keywords correlate to a reduction in customer churn. This iterative testing method could make all the difference when trying to reduce attrition by just a few percentage points and fuel business growth. That’s why competitive companies need a speech analytics platform like LucidVueCX that can not only accurately capture data from speech, but can also combine that with all the other sources of customer feedback and make it ready for analytics and decision making. Businesses need this source of valuable customer intelligence to stay competitive, grow, and keep clients coming back.