This is a follow-up to an article, “Why do Subscribers Subscribe?”, posted on Web Analytics World in March. The key insight from that article was that engagement with content and the community around that content was crucial for transitioning an anonymous user to a paying subscriber.
Here, we discuss the factors that keep individuals active paying subscribers once they have made that transition.
Subscriber Acquisition Isn’t Easy
Subscriber acquisition is not an easy or inexpensive process. Working with our publishing clients, we have observed many digital product launches and the subsequent subscriber acquisition and retention processes. Following the launch, there are a number of customers who will become paid digital subscribers without much effort on the part of the publisher.
Unfortunately, the number of these customers is relatively small and never sufficient for supporting the business model for the product. Once these early adopters are acquired, the challenge of marketing subscriptions to the rest of the audience truly begins. Tactics for marketing subscriptions are the usual channels, and the cost of the marginal subscriber acquisition increases. Keeping subscribers once they are acquired is critical to sustaining operating margins and growing the business.
The good news is that the data available for modeling customer behavior and engagement with digital products far exceeds what is available for traditional print platforms. Capturing data on individual customers, both anonymous non-subscribers and paying subscribers, enables a publisher to identify predictive metrics that are important for customer acquisition and retention.
Which Metrics are Important?
We have found that a number of predictive metrics are consistently important in our models of retention of digital subscribers in our work with publishing clients. The degree of importance for each metric differs by publication, but the core set of metrics remains remarkably similar. We can classify these metrics into consumption, interaction, attitudinal, time and socio-economic categories, although there are other metrics that do not fall into these groups.
- Consumption metrics describe the quantity, frequency and time spent with the content during a particular period of time. Number of visits, articles read, videos watched, time per visit and time between visits are examples of consumption metrics. For general interest sites, the breadth of content consumed across sections of the site can be an important predictive metric for retention.
- Interaction metrics describe actions taken by the customer while on the site or while they are engaged with the content. Comments, social media shares, contest registration and survey completions are examples of interaction metrics. There are several tools that can analyze the content of comments to quantify them so that they can be included in models of customer behavior.
- Attitudinal metrics are those that measure the level of enthusiasm or loyalty an individual has to a topic or community. Customers that are very interested in cooking will reveal this interest through subscriptions to multiple titles, gifting subscriptions to family and friends, membership in clubs or political parties and purchases of related equipment or clothing. These data are often hard to obtain, but they can be powerful predictors of behavior.
- Time metrics reflect when events occurred during a subscriber’s lifecycle and the overall time of activity on the account, often called the account tenure. Seasonal variations are often important in retention activity, particularly in certain geographical areas or when a product or service has particular relevance or importance to a customer at a particular time, such as a wedding, birth or retirement.
- Socioeconomic metrics include factors that characterize an individual’s demand for the product, such as disposable income, price sensitivity, age, gender, macroeconomic indicators, education and household type. For digital products, we have found these indicators to have some relevance, but they are often secondary to the consumption, interactive and attitudinal indicators. For non-digital subscription products, these socioeconomic and demographic metrics are often the best that are available due to limited data availability.
To use these metrics in churn models, data for each metric by customer needs to be captured and collected in a common data set. Analysis of this data using statistical regressions or other approaches can identify the effect of individual metrics on churn probability, and an individual’s relative level of churn risk can be estimated. Those customers with the highest levels of churn risk can be targeted for proactive retention efforts, such as emails, text messages or promotion offers.
The use of A/B testing, in which a statistically valid sample of customers is used as a test group for evaluating the success of a particular retention tactic, is a best practice and one that enables a learning agenda within an organization. These tests also validate the predictive power of the metrics identified in the models.
We find that organizations often under-invest in retention efforts relative to acquisition, and the reason may be the greater difficulty in measuring success. The points discussed here should help organizations understand the ROI on retaining existing subscribers compared to getting new ones.