Is the metered paywall dead?
With the launch of Piano LT[x] — a new machine learning framework that predicts user behavior — the static paywall should get ready to take its last breath. LT[x] (or Likelihood To [act]) promises to change how companies interact with users, letting them create targeted revenue opportunities and build unique nurture paths based on a range of individual behaviors — not just number of pageviews.
First up: LTs, or likelihood to subscribe. The newly released propensity model gives companies the power to predict who’s most and least likely to convert. And that’s exactly the kind of knowledge that promises to put an end to the static paywall.
Introducing One-to-One Customization
By giving media companies the opportunity to target users based not just on a single metric — pageviews — but on dozens, LTs lets media companies look beyond a pageview meter or fixed set of premium articles or videos, opening up new opportunities to personalize the paywall and optimize the rules of engagement for each unique browser. That means media companies can present users most likely to subscribe with an offer to convert, while further nurturing those less likely — exposing them to paid advertising along the way before finding the right moment to move them further towards action.
And that’s just a beginning to the one-to-one customization it promises to offer. LTs also gives companies the power to target offers, promotions and guest passes more effectively and allows sites to nurture engagement and drive conversion at every stage along the way.
How Does it Work?
LT[x] was developed by the Piano data science team, who analyzed hundreds of subscription websites on the Piano platform and billions of monthly user interactions, testing different machine learning models. Through that testing, Piano developed a clear understanding of which behaviors contribute most to a user’s likelihood to subscribe, as well as the reverse — examining which behaviors negatively impact subscription. LTs scores browsers based on their individual likelihood to subscribe, and automatically adjusts over time as behavior and brand affinity change.
Factors like the time of day or week users come to your site, the device they use to visit, or how they arrive there in the first place, all impact the scoring model. In fact, a starting base of 76 unique metrics together form predictable patterns that demonstrate how users will eventually act. Each metric plays a different role depending on the site in question, and the LT[x] algorithm automatically fine-tunes its predictions — the machine learning adapts — based on each site, subscription offer and audience accordingly.
LTs is just the first of Piano’s LT[x] offerings. Our machine learning can be adapted to drive almost any user action: registering for an account, signing up for a newsletter, accessing an ebook, asking for a software demo (to name a few examples currently in the pipeline). The next major addition for media will be a predictor for likelihood to churn or retain, coming later this year.
Putting an end to the static paywall is just our first step — LT[x] promises to change the way the media industry interacts with its users at every touchpoint. And beyond that, it will also extend past the walls of media. By building advanced capabilities to optimize for any conversion action, LT[x] will provide a wide range of companies across industries the ability to truly turn data into action.
Want to find out more about LT[x]?