One of the key things for media businesses to get right when it comes to subscriptions is pricing. A magazine leading the way in this regard is Time – the brand joining forces with Mather Economics, the global leader in subscription analytics and yield management, to enhance its pricing strategies and tactics through the use of data science and econometric modelling.
To discuss the hugely successful partnership, Laurie Truitt, Vice President, Global Consumer Growth at Time, and Luke Magerko, Head of Data Science for Mather Economics, sat down with James Hewes, President and CEO of FIPP, at a recent webinar. The online event – organised by FIPP and Mather Economics – explored in detail Time’s path to solving their subscriber revenue and pricing yield business questions.
“The data scientists are providing the right information to help a marketer make decisions,” said Truitt, talking about the things that make a partnership between a publisher and a data science business work. “But you can have too much information and then it becomes harder to test.
“So, it’s about trusting the person you work with on your data science team or a company like Mather and understanding the major points that are going to make a difference. You have to work out if the juice is worth the squeeze. Make sure the data you are pulling is valuable and not just noise on the side.”
Luke Magerko agreed: “The data science team will always look for the perfect answer. The most important part of the subject matter expert relationship is to put guard rails on the data scientists. We need to understand what we are trying to accomplish, and the fastest way to get there. Those rules come from the subject matter expert.
“When a model is produced sometimes it’s wrong, identifying things that are not the appropriate signal. The only way the data scientist will know that is through the subject matter experts saying, this doesn’t make any sense. So, it’s a relationship between the two groups that, when it marries, is quite powerful.
In these difficult economic times, media organisations are trying to assess whether raising prices is wise given the rise in inflation. Mather analysis has shown, though, that rather than being price sensitive, subscribers are volume sensitive, looking at the number of subscriptions they’ve got and just picking ones to get rid of.
“Consumers are actively reducing the amount of subscriptions, but what we are finding is that it’s not based on price,” said Magerko. “You can still have a very solid relationship with your customer while simultaneously increasing their lifetime value without having much concern about the effect it is going to have on the overall churn rate. It’s a little contradictory of what you might think but we are encouraged that we are seeing these results.”
The results indicate the competition publishers are facing has changed. “Your competition now really is the 17 other household subscriptions – the clothing subscriptions, the food subscriptions, the television and news subscriptions,” said Truitt. “(Your content) really needs to be a priority in people’s lives and how do you do that? You do that through engagement, value prop and really hoping that you hit that moment with them when they see you as a valuable asset to their monthly subscription.”
Determining your customer’s journey
The most important thing when it comes to subscriptions is figuring out the customer journey, according to Magerko. From a data science standpoint, the customer journey starts with the concept of renewal probability, which quantifies all available customer transactional data into one probability whether it’s High, At Risk, or Low.
Under Mather’s Data Science Model Flow, customers are put on a path and directed towards specific models with tailor-made pricing recommendations that work for them, whether it’s a Price Sensitivity model (for solid customers likely to renew), a Winback Pricing model (where some cancellation has taken place), a Churn model (for At Risk Renewers) and a Mail Effort Optimization model (for Low Probability Renewers).
At Time, Mather has three models in place: a print model designed to increase renewal pricing; a digital version of that model; and a Winback model designed to improve the renewal rate and the profitability of the people being brought back into the fold.
Some of the results of the testing have been remarkable. Time tasked Mather to look at one expired group and identify the people that are high-level, medium level and low-level churn possibilities. Once identified they were placed in different price points, with multiple prices going out in the expired group to try and find what Magerko calls “the floor and the ceiling”. Mather was able to ascertain that a $10 increase generated almost as much of a response as a $4 increase.
Mather Economics also estimated the price sensitivity of digital subscribers based on subscription characteristics and digital engagement. Prices were then assigned based on the estimated sensitivity, with low-sensitivity subscribers receiving higher increases.
Results indicated there is no statistical difference in response rates across subscribers receiving rate increases between 10% and 100%. That raises the prospect that there could even be a rate increase beyond 100% for certain users
“It was an extraordinary test because we thought we pushed it as far as we could, but in reality, we didn’t see the decline yet,” said Magerko.
With its Winback model, Mather estimated the price sensitivity on all recent inactive subscribers and provided price recommendations having divided people into large, medium and low increase groups. The prices were assigned on churn model results with lower churn probability correlating to higher recommended prices. The test showed that getting subscribers back at any cost is not the way to go.
“The concept of how you attract customers back to your brand has a formula to it that’s been around for a while,” said Magerko. “If someone falls off the active file and they are no longer a subscriber you treat them like you want to get them back at any cost. What you do is, allow them to go all the way back to the introductory offer and there is a hypothesis that exists that if you get to a certain price point they fall off in order to get that introductory offer.
“One of the things we are trying to help here is to say, is it conceivable that instead of dropping all the customers back to that introductory offer is there a way to identify them and, say, drop some to the introductory offer while certain people get a high price and certain people get a different promotion altogether?
“There are customers that don’t have to drop all the way back down to the lower price. By identifying the people that have fallen off the file you have the opportunity to get them back on the file and increase their lifetime value from the start, rather than sending them all the way to that lower price.”
According to Truitt, Time identifies four stages to pricing – the acquisition price, a price to stop people from cancelling, a winback price and an expiration price 90 days after the winback programme has ended.
“Then once they don’t meet the expire expectation in that fourth range of pricing, we let them go back and let them start the process over and maybe let them in a lot lower,” she said
“But there are those attempts that happen from the time they acquire to cancel what happens in between. What stage they are at, whether it’s winback or expire, and treating each one of those differently, definitely helps build up your revenue because instead of taking them from cancel all the way to the bottom, there are these middle stages where you can actually capture a large amount of people.”
A new FIPP report, Moving from Print to Digital: How to Manage the Economics of Media Change, has been published in partnership with Mather Economics, and can be downloaded here.