Understanding Users’ Paths to Subscription: Content Attribution Scoring for Editorial Decision-Making

Deep.BI’s Deep Content Attribution Score (DCAS) is designed to create impactful insights for editorial and data team decision making. We are excited to present the newest expansion of DCAS, which has been designed to help editorial decipher content value by taking a deeper look at article performance in driving user engagement and conversion.

Following in true reader-first fashion, we know that by using DCAS to visualize a user’s progressive path to subscription, we are able to generate much more valuable insights about the value of content than those provided by outdated metrics (such as last-click attribution or simple pageviews). This yields a much better user experience and the delivery of content that matters to the reader.

Knowing the realities of an article’s effect on users can strongly impact editorial strategy. Clearly, content is no longer simply news. It is now a significant driver for engagement, as well as the key for converting users. With the insight that DCAS provides, editorial teams can now focus on getting the most from their content.

A Deeper Look: How did they get there?

The Attribution Score is comprised of two scores; the first is the Acquisition Score, and the second, the Conversion Score.

The Content Acquisition Score 

This score is heavily focused on article order - the number and timing of article reads, and to a lesser degree on time to acquisition. Looking at the graph, it is clear that the score is decaying exponentially from the first encounter until the final moment. 

Despite the apparent gap in time between initial interest and further engagement, the score remains consistent as the user is re-acquired and becomes an engaged reader.

Deep.BI Content Acquisition Score

The Content Conversion Score

This score grows exponentially from the very first engagement and is based heavily on time to subscription with some consideration for article order. Looking at the graph, it is clear that as the subscription moment approaches, the content is scored similarly; depending on the user behavior. For example, if 4 articles were read during one session, their scores are almost identical, while those that have a time gap between them receive a slightly higher score since they contributed more to habit formation.

The last piece of the puzzle is the content consumed in the last 24 hours - this content receives a larger bonus. The key here is the following: user conversion into a subscriber depends heavily on their habit of using the site/app/digital asset that the publisher offers. This habit eventually culminates in an engagement level at which the user decides to subscribe - those last articles are given a much higher score due to their push for the user to see value in the subscription.

To explain further, as each article is read it will receive both a Conversion and an Acquisition Score, however as articles read early in the subscription funnel do not contribute much to a user’s Conversion Score , these will receive a conversion score of almost, but not quite, zero.

Deep.BI Content Conversion Score

When the two scores above are combined, they yield the Deep Content Attribution Score (DCAS), a measurement of the content that takes into account both its acquisition and conversion aspects. This score is the weighted sum of the two scores where the Conversion Score is given approximately twice the weight of the Acquisition Score.

Deep Content Attribution Score

Because there is only one subscription at the end of this journey (combining the DCAS of all articles read by a user until subscription must amount to 1), all of these scores must consider the total number of articles read by the subscriber up until the moment of subscription. This is why the scores/article of a user who has read 10 are much higher than a user who read 100 before subscribing.


Finding deeper value in the details

In realistic terms, DCAS will enable your organization to separate content into 4 categories: “Conversion”; “Acquisition”; “Habit-maintaining”; and “Unsuccessful”. They can be explained as follows:

  • Conversion Articles: articles that have a Conversion Score within the top 25% of the most read articles by future subscribers.

  • Attribution Articles: articles that score within the top 25% but with an average RFV lower than 45 (signifying that they are being read by future subscribers but not to the same degree as above).

  • Habit-maintaining: these scores fall within the bottom 75% of the average Attribution Score, but the RFV score (+35) indicates that these articles are being read by users who are already engaged. Specifically they are being read by users who have begun to establish a constant habit over time.

  • Ineffective: Articles that were not on the path to subscription of any users within a time period of a month from subscription. This is not to say these articles are not necessary for other types of readers or not important in driving ad-sales, for example ,in hybrid revenue models. They simply did not appear on the subscription path of users within the past month and must be strategically evaluated by the editorial team.

It is worth remembering that articles can change categories over time as their attraction to users changes as well. For example, an article might initially be classified as unsuccessful, but may eventually prove to be a highly specific acquisition article for a user type that requires a lot of cultivation before conversion. This is why DCAS is viewed over a period of time for the most accurate results and is fit best for strategic decision-making in content creation and distribution.