Is it the “new normal?” Last week was another week of great performance of online subscription sales worldwide, reported Piano, a publishing business platform, based on the statistics of 295 paywalled news sites. 

• In the week of April 19, sales in Europe were up 97% in comparison to the weekly averages in January and February. They have been stable around the 100%-increase level for the third consecutive week.
• Sales in the United States were up 78% vs. pre-Covid averages. They have stabilised, too — for the second week, holding around the 80%-increase. 

A month ago, we dived into the data of a mid-sized European news site, supplied by Deep.BI, an analytics company, and discovered the bump in online traffic and subscriptions were separate phenomena. It was the existing readers who were more likely to subscribe than the new ones attracted by the COVID-19 coverage.

Today, we revisit the data set and check what happened to the readers of the site between March and April. 

On the surface, all is fine — the number of online users doubled in March and then by April it decreased just marginally. Drama, though, is hidden under the surface.

The publisher enjoyed an unprecedented surge in online traffic. The number of users almost doubled from 2.71 million in February to 5.09 million in March! In April, the traffic still is way above the pre-COVID levels — between April 13 and 23, it had 4.81 million users.

Together with Pawel Blazejewski, product manager at Deep.BI, we dive deeply into the data to analyse the segments of users defined by their engagement. Similarly to the industry leaders such as The Financial Times, this mid-sized publisher uses RFV scores to measure engagement of each and every user. RFV is a composite metric built on scoring Recency and Frequency of visits and Volume of pages viewed.

• Firstly, we see the publisher has successfully engaged the heavy users, or those classified as “engaged” and “addicted.” These are the segments of users most likely to subscribe. Most of the current subscribers have been members of these two segments.
• Then, secondly, we see a danger for the future acquisitions. As the new readers flooded the site, only a few engaged enough to become “light users.” In fact, the share of “light users” shrunk rapidly from 31% of all users in January to 19% in April. 

The data on the flows of users between segments over a month uncovers a struggle in engaging the new users attracted to the site during the COVID-19 pandemic.

• We see 3.06 million potential customers visiting the site in March just once. They are so called “flybys,” of whom most are referred by search engines and social networks. Then we find that only 420,000 stuck to the next month.
• What about 1.34 million “light users” from March? Almost half of them were gone a month later and did not return even once!
• Had we visualised the segments as an engagement funnel, we would see a shape of an umbrella. Broad at the very top and then a narrow in the middle. 

Recommendation for the publisher? Focus on increasing the visit frequency and depth to deal with the bottleneck in the engagement funnel. Consider the following initiatives and tactics:

• Audit the user experience on the site as the problem might be the way it works. For example, check and work the page load time.
• Audit the performance and experiment with different calls to sign-up for push notifications or newsletters, which will let you to remind users about new content.
• Audit the performance and experiment with content recommendations on article pages to improve the time spent and the number of pages viewed, in particular test different selections, formats, and positions.
• Track and retarget “light users” with your best-read content and the most popular newsletters on social media and ad networks. 

Interested in the executive briefing on the bump and the best practices? Check the programme of the INMA Virtual World Congress, starting the next Tuesday on May 5. We talk reader revenue on Thursday, May 7. 

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .