Slow Churned: How To Target Retention Efforts So They Don’t Backfire

There’s a lot of talk in on the Internet about retaining subscribers, but very little of it has hard data behind it. Heres a look at several academic research papers on the topic of why retention efforts can actually increase churn, and how to tailor retention programs so that they work the way they are intended.

You can focus on your acquisition funnel and your lead generation and conversion rates all you want, but for a subscription business, you’ll be ignoring the biggest factor in total lifetime value of your customers unless you concentrate on retention. But don’t take my word for it; take the word of the academics who devote their lives to understanding and explaining these issues.

In 2004, Sunil Gupta found out that if you improve your acquisition cost by 1%, you boost firm value by 0.1%. If you boost your margin by 1%, you boost firm value by 1%. And if you boost your retention cost by 1%, you get a 5% increase in firm value. Increasing retention — aka, slowing churn — pays off in spades!

In an overview of the industry written by 12 different professors and released in April 2017, the importance of retention is a key point. Gupta, who teaches at Harvard, is one of the authors of that paper, titled “In Pursuit of Enhanced Customer Retention Management: Review, Key Issues, and Future Directions.” Don’t look for Web-friendly headlines in academic paper titles. Instead, look for insights like these:

Measuring retention is important for several reasons. First, predicting how long someone will remain a customer lies at the heart of any attempt to calculate lifetime value and customer equity. Second, retention drives firm profitability and value. Third, measuring retention rates over time can provide a key metric of the firm’s health.

And in fact, measuring retention has been a key academic focus for a decade or so. At last, researchers are now looking at what to do about it. That is, measuring retention is fine, but how do you improve it? And what if, when you try to improve retention, you actually make it worse? That’s what Eva Ascarza studies at Columbia.

WHEN GOOD MARKETING GOES BAD

Ascarza just finished a paper (dated July 2017) that will be published in the Journal of Marketing Research, called “Retention futility: Targeting high-risk customers might be ineffective.” She argues that companies who aim retention programs at the customers most likely to churn (that is, to unsubscribe, opt out, or quit subscribing) are making a mistake. She says, “contrary to conventional wisdom, customers sometimes react negatively to the firm’s intervention.” That is, when the company reaches out, the reaching out itself triggers the customer to quit.

Raghuram Iyengar, a professor at The Wharton School, said basically the same thing in 2015, in this video interview. He described his work with a cell phone company. The company figured out which customers were at risk of churn. Some of those high-risk customers were targeted with a campaign intended to boost retention. The rest were not. And the results were pretty scary, at least to a marketer:

For one group of customers who were randomly selected, they did a retention campaign. This was a pricing plan campaign. For another group of customers, they did nothing. The group of customers who were in the retention campaign actually left a lot more – in fact, staggeringly more. This was pretty bad for the company.

The customers who the company tried to keep were more likely to quit than the ones who were just left alone! The actual data (Table 3 at this link) showed that 10.0% of the test group quit after marketer outreach, but only 6.4% of the control group quit over the same time span. Iyengar says, “Sometimes, it’s best to let sleeping dogs lie.”

But why? Why do the best-laid schemes of mice and marketers gang aft agley? These researchers wanted to figure it out, and in follow-up studies, they figured out and described two kinds of customers who should not be contacted.

Uneducated customers who learn they have better options. Some subscribers quit after finding out from the retention campaign that there are better choices out there — maybe the marketing campaign inspires the customer to shop around, or provokes them to just quit. Ascarza puts it this way: “the intervention could make customers realize their (latent) need to churn.”

Why did highlighting past usage increase customer churn? … Contacting customers who are at risk of churning might stimulate a “latent need” to churn. … Proactively contacting customers can increase churn for some customers by drawing their attention to the amount that they are paying for their level of usage.

When customers find out how much they have been paying, it inspires them to shop for a different provider, or to stop using the service altogether.

Unhappy customers who are reminded that they want to quit. “Inertia” is the word academics use to describe the quality of being too lazy or busy or uncaring to quit. A marketing campaign targeting these at-risk clients may be the unintentional kick in the butt these subscribers needed to make the effort to cancel.

There is extensive evidence that inertia prevents customers from switching plans and providers. For example, customers in the treatment condition were encouraged to switch plans. Such a treatment would likely lower their inertia by, for example, making customers realize that it was easy to switch. We postulate that once customers realize that they can easily switch plans within the company, they might also explore competitors’ offerings, thus increasing their risk of churning.

These are the sleeping dogs who should be left alone, but who are motivated by the marketing outreach to quit. The campaign actually provides the kick in the pants that these customers need to unsubscribe.

PREVENTING BACKFIRE

So what can marketers do who have identified customers likely to churn — but who now know that reaching out to these customers will trigger more of them to unsubscribe than if you just left them alone.

There’s an old vaudeville gag — A patient says, “Doctor, it hurts when I do this!” The doctor replies, “Don’t do that!” The takeaway for marketers is similar: first, do no harm. If you suspect you will drive away more customers if you reach out, then don’t reach out!

But there are other options, according to the social scientists — testing! But not just A/B testing, which will not always do the job right. Here’s Raghuram Iyengar again:

it’s not enough to just do A-B testing. It’s important to analyze the data correctly. Let me give you a specific example. In our A-B test, one group was the people who received recommendations; the other group was people who received no recommendations. People who received recommendations and accepted them churned less, compared to the control group. Now, one might think the retention campaign worked. But what’s important to remember is that customers decided to accept the campaign. So, there is self-selection there.

So A/B testing can be effective if you are careful to account for self-selection effects.

But there is another experimental method that goes one better: field experimentation, the “gold standard,” says Iyengar. The purpose of this kind of testing is to identify the characteristics not just of customers likely to churn, but to identify the characteristics of customers likely to react positively to marketing efforts. That is, it is not enough to identify at-risk customers. It is far better to identify those for whom marketing works, and target only those.

It seems like common sense — aim your marketing at people who react well to marketing. As for the customers who react poorly, leave them alone. But if you can identify these people — Eva Ascarza says that they have “sensitivity to the intervention” — then you can increase overall retention and decrease churn by marketing to them.

We have demonstrated that customers with higher propensity to churn are not necessarily those who are more sensitive to the retention efforts. … We have demonstrated that, contrary to the conventional wisdom, proactively targeting high-risk customers might not be an effective strategy to reduce churn because by doing so, firms are wasting resources on customers who are not responsive (or even respond negatively) to the campaign. … Consequently, firms should first explore customer heterogeneity on the sensitivity to their retention efforts, and then target customers whose sensitivity is the highest, regardless of their intrinsic propensity to churn.

Using this methodology of identifying the customers likely to respond well to marketing, Ascarza conducted two field experiments and reduced churn in one trial by an additional 4.1 percentage points (compared to a control) and by 8.7 percentage points on a second trial. Using pilot programs, A/B testing, and randomized testing lets marketers get at this “heterogeneity” and then offer pinpoint targeted programs aimed correctly.

But be careful when analyzing results over time, according to this paper, for the danger of “survivor bias.” That’s the tendency to think that your retention rates are rising over time when what’s really happening is that the only customers you have left in your sample are long-time subscribers.

Caution needs to be taken when calculating aggregate retention rates because customer heterogeneity in individual retention rates can cause “survivor bias” [For example,] we have two types of customers, “good” and “bad”, with retention rates of 70% and 20%, respectively. We begin with 500 customers of each type. The 500 good customers experience a 70% retention rate hence they go from 500 to 0.70 × 500 = 350 in the next year, etc. The 500 bad customers, with their 20% retention rate, go from 500 to 0.20 × 500 = 100 customers, etc. This produces a “survivor bias” with the good customers dominating the aggregate retention rate over time as they are more likely to stay within the customer base.

Academic researchers are delving deeply into these and related issues. What incentives work best? What role does timing play? How should you manage multiple overlapping campaigns? What’s the best balance of acquisition vs retention? There’s every reason to think that keeping an eye on the Ivory Tower will continue to yield useful insights:

While we have learned a lot about customer retention, management attention and academic research efforts need to be broadened beyond identifying customers who are most likely to churn. More work can be done in this crucial area, but there are several other areas very fertile for future work. It is clear that the customer retention problem is not going away. Researchers need to guide managers on how to deal with this issue successfully. Managers need to think about retention more broadly.

Insider Take

It is not enough to identify customers who are likely to unsubscribe and then hit them all with a shotgun blast of marketing incentives. Doing that tends to backfire as you remind some customers that they are unhappy, and inspire others to shop around. Instead, test your marketing efforts on subsets of your user base to identify those subscribers who are likely to respond well to your retention campaign. Target those customers, and let the rest, the sleeping dogs, lie unpestered by marketing.