It’s time for that time-old question again: Are your batch and blast campaigns losing you revenue?

The short answer to this is yes. Sending one message to hundreds, thousands, millions of customers is just not going to be effective in this day and age. And spoiler alert—it hasn’t been working for a while.

But a lot of people rely on this campaign set up to get the word out. And sure, it’s the fastest, surest way to get your information on your customers’ radar right? Yes, true—but with only 10% of marketing messages getting any kind of engagement, you’re effectively teaching your customers to ignore 90% of your messaging.

And this is damaging to your brand’s reputation, of course, but will also do nothing to help your bottom line.  

Take this campaign as an example (and a lesson). 

One of our customers in the gaming space decided to test out Swrve’s propensity to churn machine learning model by designing an experiment to see which variation drove the most revenue and reduced the most churn.

The Campaign

Our customer sent one offer to all users with a holdout group set aside as a control to measure results.


The Results

The experiment showed a reduction in churn and increase in spend among high-churn score users receiving the offer—a pretty massive win. See the results in more detail:

  • 4% reduction in churn
  • 18% increase in spend

However, the experiment also showed that users with low churn propensity who received the offer spent less and churned more. In fact these users showed a: 

  • 14% increase to churn
  • 20% decrease in spend

This shows there was regret associated with messaging and rewarding these users. The obvious next step would be to restrict the campaign to those users with high propensity to churn. This might not just be users who haven’t been active recently—Swrve’s propensity scoring system will help marketers target the right users. 

Swrve’s Solution

Being able to predict churn is exceedingly valuable to brands. Knowing which users are at most risk of abandoning your app enables you to take action to mitigate it. A 1% reduction in churn typically equals 5% increase in customer lifetime value. Plus the fact that it costs 5x more to acquire a new customer than it does to retain an existing one means this is something that all brands should be taking seriously.

Swrve ingests and processes over 20 billion data points a day, giving marketers the ability to identify, predict and respond to customer needs in real time. The churn propensity score indicates how likely each user is to stop interacting with a particular app. Brands can then act on the score to create retention campaigns and deliver them to users based on their individual propensity scores, in addition to other key behaviors and signals.

To learn more about churn propensity and our other best-in-class AI and machine learning development, talk to one of our experts here