The definition of a Dynamic Paywall assumes that the paywall shows an offer when certain conditions are met, regardless of whether these conditions are based on content category, device type or location. The most important factor paywall actions depend on is the dynamic of a particular user’s engagement during their interaction with a publisher’s site.
The first part of rule-based segmentation is generally easy and offered by almost all paywall providers. The second element is extremely difficult as it requires Data Architecture to provide the capacity for real-time user engagement scoring.
Many solutions are able to run such scoring only by weekly or daily batch processing which seriously impacts their effectiveness because user-performed actions, such as purchasing a subscription, are very much in-the-moment decisions.
“Emotional (or impulsive) purchase” is an old trading concept used by many vendors. It is driven by the fact that purchase is carried out when the need to possess intersects with growing emotions/impulses, creating a specific moment when the probability of purchase reaches its highest peak.
After such a moment ,the probability to purchase decreases significantly - and publishing is no different in this regard than any other business.
Consider this scenario: a user comes to the site with increasing frequency and consumes more and more content with each session. One day, the user becomes really interested in a story or series of stories and wants to read them immediately. This user’s emotions are heightened at that moment. If we are not able to detect this short-term flare-up, we will lose the opportunity to sell them a subscription at a time when they are most likely to buy.
Simple rule-based metric modeling (e.g. IF a user is coming from Facebook THEN show paywall) or batch processing will not work in this type of case.
Let’s consider a further example: a publisher chooses a metered model with a limit of 5 articles per week. Then two new users come along: User A, who came one day and used up their limit of all 5 articles in one session, and User B, who comes every day, 5 days a week and reads one article each day.
User A did not have a chance to build either habit or loyalty to the publication before they were met with a paywall. User A will not wait another week just to come back to a refreshed paywall, thus User A never comes back.
Now User B comes every week but 5 days a week. Their engagement pattern is very clear and is growing with every visit during the month (due to a growing frequency of visits and a smaller time between each of these visits - otherwise known as recency). Then, one day User B finds a set of articles which holds their interest during the visit, causing them to read 3 or 4 articles during that session. Because User B has developed a high level of engagement, they are ready to be offered a dynamic paywall and are quite likely to subscribe.
If you are not scoring your users dynamically in real-time but only in batch processing, you are not showing them a paywall at the specific moment when their emotions are at their peak. You instead show the paywall the next day (when your batched user scoring process shows you the results). Unfortunately, the emotions/impulses are gone and probability to purchase will be much lower than the day before. Thus the reader will ignore the paywall and return to their routine of consuming one article daily, 5 days a week. In the case of batch processing of data, we’ve actually created a negative habit for that user rather than pushing them to subscribe.
If the paywall offer is displayed “IN THE MOMENT”, it will meet the criteria of already existing engagement (the need) and the emotional/impulsive need to possess this locked content, thus capturing the user when their probability to purchase is at its highest.
After the purchase, if subscription on-boarding is well executed, user engagement should grow very quickly. Most importantly, those users that purchase a subscription with a stronger awareness, which has been reinforced with engagement, stay longer then subscribed users who were randomly caught by paywall. We have observed this reality in many cases of individual user behavior.
It can also be seen with user engagement levels on user journeys required to execute certain actions.
When constructed properly, Dynamic Paywalls deliver consistently better results over any other methods and do not create negative habits that result in users dismissing communications from publishers as “just another pop-up or paywall”. Instead they are seen as valuable invitations to invest in what interests them at the exact moment when they’re most likely to invest in the subscription.
Start using real-time user scoring to power your paywall and increase subscriptions!
