Understanding the customer journey is the grail for many publishers—for Hart Energy, it’s now a reality. At the upcoming Business Information & Media Summit, Hart’s chief digital officer Mark Chiles will show how Hart applied data integration and predictive analytics to generate a clear picture of customer priorities—and revenue opportunities for Hart, including the potential for a dynamic paywall, reduction of subscriber churn, more effective advertising targeting and better understanding of event attendees needs, including how to upsell those attendees to other paid products across the Hart portfolio.
At a recent Connectiv Digital Media/Audience Marketing Committee meeting, Chiles offered a preview of his presentation on how Hart was able to turn customer data into actionable information. Data is pulled from a “frankenstack” of platforms that includes customer data platform Treasure Data, Marketo, Cvent, Omeda and others. “We can cross-query the different data sets and understand first party data versus third party data and how we reach out to that person—is that via an ad network on social media, do we want to serve them targeted content to keep them engaged, do we want to get a marketing message out to a certain persona type to potentially get them to subscriber or buy?" said Chiles. "We’re starting to wrap our heads around the data and right now we’re just scratching the surface.”
“We want what everyone wants—to get them in the door, to nurture them, to keep them engaged,” added Chiles. “We don’t want to lose them from a purely transactional standpoint.”
KPIs of Engagement and Reducing Subscriber Churn
Hart creates a series of questions that varies by the customer they’re trying to understand. In comparing users (registered but unpaid readers) and subscribers, Hart asks questions such as,
- How many subscribers log in to one site?
- How many subscribes log in to multiple sites?
- How many registered users log in?
- How many active users look at 5+ articles within the last 30 days?
- What is our subscriber churn propensity?
- What was the persona type of the users that view said article?
“We can obviously look at what they’ve purchased but that doesn’t necessarily mean that’s what they use,” said Chiles. “We want to get to a KPI understanding of engagement metrics versus just anecdotal. For something like how many articles did a user view in the last 30 days, we’re looking at a dynamic paywall for 2019 and we needed to know what that represented how many articles are viewed before they subscribe.”
Chiles said Hart is now at point where it can do predictive modeling. Hart compared two subscriber data segments—active and expired—to identify characteristics that indicated a propensity to churn.
Hart was able to create four segments—Possible, Likely, Marginally and Unlikely. “That helped us start to understand how many people could we lose month to month but what we really wanted was to start to get an understanding of the behavior behind it and the criteria that correlates to that,” said Chiles.
The company created a graph of users with demographic detail such as job titles and companies but more importantly, behavioral features such as expire data that offers a look at the elements and attributes of a subscriber who is in danger of dropping. “We now can understand what that looks like by using machine intelligence and algorithms from Treasure Data and see with a great deal of accuracy what a user is going to potentially do,” said Chiles.
Identify Event Audience (and Upsell Them)
Earlier this year, Hart held its first DUG Permian Basic Tech event and was able to gather detailed insights on attendees and exhibitors, including different types of registration (full conference or exhibit-hall only) and what they paid, which helped Hart determine the average rate (or “blended rate” as they refer to it).
“It gives us a sense of what people are looking for,” said Chiles. “When we look engaged personas across all our events, we’re able to look at what people needed and were willing to do to get the information they needed—including buying a subscription or a conference pass.”
Personas and Targeted Messaging
Hart is starting to use its customer data to create personas around certain titles and demographics. “It’s easy to say ‘pull a list of CEOs,’ but that doesn’t mean you’re getting all the top execs at a company,” said Chiles.
Using its customer data, Hart can identify includes the title of CEO but also executives whose responsibilities overlap with CEO (such as president, chairman, etc. “From an account-based marketing and even personal marketing basis, we can create a very specific message to that persona,” said Chiles.