How TheStreet.com Used Data Modeling to Boost Renewal Rates and Price
In the relatively world of online publications featuring financial news and analysis, there certainly aren't many that can claim to have the sort of cache and name recognition of TheStreet.com. Today, TheStreet, Inc. boasts a rather formidable B2C unit featuring a suite of free newsletters, premium subscription products, an app, and more.
Launched in 1996, TheStreet was co-founded by the unconventionally gregarious money manager Jim Cramer, who is perhaps better known for hosting the long-running CNBC investment program "Mad Money." And yet as prominent and recognizable as TheStreet, Inc. and its co-founder may well be, it was, for a number of years, plagued by low renewal rates and poor customer retention statistics.
The Challenge: Boost Customer Retention, Renewal Rates, and Pricing, All at Once
About two years ago, as part of a management restructuring process at the company, marketing professional Jill Marchisotto was brought onboard as "the first person ever," as she said, "to really look at retention strategy for [the] company."
Her initial task as TheStreet's head of digital engagement and retention marketing, was simple enough to understand, if not necessarily to execute. It involved "modernizing everything about the company, and to prove results quickly," she said. "So, more revenue, more [subscriptions], better products.”
At the 2017 FUSE Media Summit, Marchisotto spoke during a publisher spotlight session to share the data modeling and price optimization strategy she and her team employed to improve the company's customer retention and renewal rates. Ultimately, Marchisotto's goal was not just to convince large numbers of readers to return to the proverbial fold, but to also increase the rates subscribers to the TheStreet paid where possible.
The Strategy: Bring in Outside Experts to Offer a New Approach to a Common Problem
TheStreet's old renewal pricing strategy was frustratingly basic, said Marchisotto. "It was just based on log-in frequency. So if you logged in a lot you were put into one group, and they got one very high price. If you logged in a little you were put in another group and got a low price. And that was it. There was also little historical price testing, so I didn't have much to go on."
It was fairly obvious to Marchisotto that one of her first orders of business should involve price testing with whatever data TheStreet already had. It had quite a bit: TheStreet had been fairly diligent in collecting customer data -- it just hadn't done much with the data it had. And indeed, Marchisotto and her team "did see some immediate positive results," after employing some standard A/B testing. "But we knew we could do better. We had all this data, and we knew we could leverage it in a better way to improve pricing and overall renewal rates."
Ultimately, though, the answer resided with the Minnesota-based Apex Decisions, a data science company that was brought in to create a unique one-to-one pricing model for TheStreet's customers. As Marchisotto explained, "We undertook a very in-depth proof of concept: We gave them half our file and we said, 'Take it. Look at all of these different attributes we have on it, and come up with what you think you can do for us.'"
The goal of the Apex Decisions project, Marchisotto said, involved the overall renewal and engagement strategy for all TheStreet's paid subscription products. TheStreet also wanted to significantly boost its renewal rates while also maximizing price. "And one or the other was not really an option that I was allowed to have on the table," Marchisotto said. "This had to be something that improved [retention] rates and maximized price in order to be successful."
The Results: A Win-Win Solution
Apex Decisions, as Marchisotto sees it, is a company that eats, sleeps, and dreams data and data modeling. And although she didn't share any of the specific details of what was clearly a very sophisticated data model, Apex was ultimately able to help TheStreet significantly improve both its retention rates and its pricing by offering personalized price points that made sense to TheStreet's customers.
The bottom line of the Apex approach, according to Marchisotto, involved the process of building multiple models that layered on top of each other to create one-to-one pricing. That's an interestingly novel approach, Marchisotto said, largely due to the fact that renewal pricing is generally based on groups of people -- not individuals -- and the testing itself is usually carried out in a fairly simplistic A/B testing manner. But with the Apex modeling approach, she says, "everyone here is getting a different price, basically. Once these models are laid on top of each other, Apex does their magic and spits out a score and is able to apply one-to-one pricing."
Ultimately, Apex used a total of 17 models, each layered on top of each other, to come up with the unique renewal pricing model it offered to TheStreet. And interestingly, because this sort of model building is an iterative process, it takes time to learn from itself, which means the pricing and renewal structures should become more and more accurate over time. As Marchisotto said, "This is something you have to be a little bit patient with."
Marchisotto also made the important point that the only reason Apex was able to apply its very sophisticated data model in first place was because TheStreet had that data aggregated. If the company hadn't been collecting the sort of detailed customer information it had been for so long, it likely wouldn't have been possible for Apex to apply its data models.
It should go without saying that Marchisotto's team is more than pleased with the data models that were eventually applied. "Over the last two months for all the products," said Marchisotto , "renewal rate is up two points versus control. And our average price is up more than 20% using the pricing model."
As for TheStreet's next steps, it has plans to increase the reach of the Apex data model throughout a number of its other product divisions. "As we move on," Marchisotto said, "the model complexity is increasing. So we need to get them more data."
Apex, for instance, isn't ingesting data on TheStreet's app usage or its email open rates to inform pricing. Nor do they know which customers are and are not attending TheStreet's live events and its webinars. "We're looking for ROI for [the Apex] model," Marchisotto said. "And we really hope to use it even more moving forward."
Dan Eldridge is a journalist and guidebook author based in Philadelphia's historic Old City district, where he and his partner own and operate Kaya Aerial Yoga, the city's only aerial yoga studio. A longtime cultural reporter, Eldridge also writes about small business and entrepreneurship, travel, and the publishing industry. Follow him on Twitter at @YoungPioneers.