With the economy shifting further and further towards e-commerce and digital services, it’s no surprise that 2021 saw a major increase in companies deploying subscription-based pricing. In particular, the subscription model has come to define the way consumers enjoy popular entertainment, from using services like Apple Music and Spotify for streaming music to an ever-expanding range of new video services like Disney Plus, Peacock and Netflix. 

According to an April 2021 report from Deloitte, the average consumer maintains nine entertainment subscriptions across video, music and gaming. With popular content fragmented across competing platforms, consumers must choose the services that meet their needs within a limited monthly budget. This market saturation is leading to frustration among consumers, with more than half of those surveyed by Deloitte reporting that they find it difficult to access content across so many services. 

As subscription businesses turn the corner into 2022, they’ll be forced to fight against a trend that is gaining momentum among consumers: subscription fatigue. And a trend that is gaining momentum among service providers: churn. Rather than pay nine or more subscription fees each month, many consumers are choosing to prioritize the services they use the most or those that offer the best value for money. This shift in consumer behavior is, in turn, giving rise to new technologies like artificial intelligence (AI) to meet the increasingly daunting challenges posed by customer flight and the increasingly critical need to focus on customer retention. 

The Churn Challenge  

As early as 2000, business experts found that customer retention would play a vital role in e-commerce success. According to a frequently cited piece from Bain & Co., a 5% increase in a company’s retention rate can drive profit increases ranging from 25 to 95%. On the other hand, the cost of bringing in a single new customer is significantly more expensive than holding onto an existing shopper.

More recently, the growing importance of customer data has provided more reason for companies to focus on retaining their best customers. With new insights into the behaviors and preferences of their shoppers, retailers and subscription service providers have made customer lifetime value (CLV) a guiding metric for their sales and marketing strategies. By maximizing the length of each customer relationship, the company enjoys the double benefit of increased revenue and decreased marketing expenses to acquire new customers.

So with retention — and AI — both becoming the name of the game, how can a subscription business optimize their operations in the coming year? Here are three key steps to applying AI to minimize churn:

1. Collecting and processing customer data: Every company will have a unique set of customer data based on the information they’ve collected throughout the sales process. Tackling churn with AI begins by cataloging and understanding each available data source, ensuring that the information is organized and accounted for. The data must then be preprocessed to clean and filter the information, separating out anomalies and arranging the data in a format conducive to analysis. For an AI solution to provide actionable insights, organizations must resist the temptation to cut corners during the tedious early stages of data collection: an algorithm is only as good as the data underpinning it.

2. Modeling and analyzing the processed data: Different data has different characteristics and requires a different approach for analysis. Whether a company is analyzing the total spend of a group of customers (numerical) or combining results based on shared characteristics (categorical), the data must be carefully inspected and modeled to produce useful insights. Of course, all personally identifiable information must be stripped from the data to achieve anonymity and maintain compliance with customer privacy regulations. Dividing the data into audiences and segments will allow the company to apply targeted marketing and sales strategies without compromising user privacy.

3. Interpreting the results and applying new strategies: To gain the insights needed to predict and reduce customer churn, the company’s data analysts must interpret the collected data. A strong customer churn model will be able to generalize the results of different data segments and apply those generalizations to potential future outcomes. Business leaders can ask questions of the model, for example, what will happen to customer churn if the subscription cost is reduced by a certain amount. The model will then determine the future results of that strategy, providing the end user with data-backed information to help drive future decision-making. As companies explore new products or business strategies, the AI-backed model makes it easier to predict the outcomes and move forward with confidence.

Entertainment and media companies shifted towards subscription models because of the predictability it brought to their operations — guaranteeing a set monthly fee from each subscriber. However, as customer behavior shifts as a result of subscription fatigue, the same companies are now dealing with new levels of uncertainty. With thousands of users in a subscription renewal cycle each month or year, companies must develop new strategies to ensure that their valuable subscribers don’t cut ties. And they can often benefit again from the enhanced predictability that can be surfaced from AI-generated insights. In 2022, we’ll see an increasing number of providers fighting over a finite number of consumers, and those companies with the strongest retention algorithms will be best positioned to survive. ■

The popular “subscription fatigue” narrative is that consumers have topped out on the number of over-the-top services they’re willing to pay for and are now in pruning mode.

But Parks Associates—which was one of the first research outfits to put the notion of subscription fatigue into the lexicon—now says that the number of OTT services in the average home is still expanding, and it’s traditional pay TV that’s getting the pruning.

According to some of the latest Parks research, the percentage of broadband homes subscribing to pay TV dropped from 87% in 2014 to 79% last year. But the percentage of households subscribing to at least one OTT service increased from 55% to 64% over that span. And the amount of homes taking two, three or four OTT services also increased significantly over that time period.

Consumers, Parks said, are “not only willing to pay for a subscription, they are willing to pay for multiple services, even premium-priced services if value is perceived.”

This is a change of course from Parks’ previous position.

In September, Parks & Associates released a study suggesting the subscription OTT market had become “saturated.” Consumers weren’t necessarily tossing away subscriptions to popular platforms like Netflix, Hulu and Amazon Prime Video, Parks found, but they weren’t adding new services to their monthly expenses, either.

"In talking about market saturation in 2018, we were saying that household penetration had leveled off (% of households taking any OTT video service)," said Parks senior analyst Brett Sapington, explaining the firm's position. "We still think that there is lots of room for OTT to grow in terms of number of subscriptions per household. Our figures for number of services taken continue to increase, and the number of households with three-plus services continues to jump.

"So, we are not big believers in subscription fatigue in OTT," Sapington added. "Rather, consumers evaluate each service on its own perceived value rather than comparatively, and current pricing allows consumers to pack on more. I think that the launch of the new Disney and WarnerMedia services will prove that out (or not). We are looking for the average number of services per household to jump by ~1 service the end of this year.