ALERT: INMA, GNI to unveil CMS Vendor Selection Tool 2.0 at town hall Wednesday, register free here

Stampen Media experiments with article supply, paywall pages to increase conversions

By Håkan Hamrin

Stampen Media

Gothenburg, Sweden

Connect      

After the first quarter of 2024, we noticed the number of ad impressions were declining at Stampen Media. However, the editorial part of our company argued the traffic was at the same level as the year before.

And, although we had done a lot of activities to raise the number of ad impressions per pageview, we realised they were right!

Why was that the case?

A quick investigation on the site as a logged-out user gave us the answer, and, digging into the statistics, the numbers showed the same thing: We had a lot of traffic on the pages known as paywalls.

This is the page where a logged-out user only sees a little bit of the article but then has to log in to read the article. This page was very short, leading to very few ad impressions on that page. This was the reason we had the same page impressions but fewer ad impressions compared to last year. We had locked in a much higher share of the articles.

It is a delicate balance in figuring out which articles readers should be able to access freely while ensuring ad and subscription revenue remain top of mind.
It is a delicate balance in figuring out which articles readers should be able to access freely while ensuring ad and subscription revenue remain top of mind.

Of course, that was not the best scenario for ad revenue. However, the revenue from our consumer business was better. We had converted a lot of readers to paying customers, leading to an increased revenue from our users.

We didn’t want to lose that, but we did want to get back the ad impressions to meet the budgets for ad revenue.

There was a bit of discussion on how to handle this. One of the key factors in the option we chose was that we wanted to meet this increase in supply with an increase in demand.

So, what did we do?

We started decreasing the share of articles put behind the paywall during the weeks salaries were paid out. This was because those weeks always resulted in a higher grade of demand, with both higher fill rates and higher CPMs (cost per thousand impressions).

If we were going to increase the supply, of course it was better to do it when the demand was highest.

But, what about the consumer business in this solution? Wouldn’t we lower the conversion rate if we had a smaller share of articles behind the paywall? The fewer times the reader hits the paywall, the fewer opportunities to convert to a paying customer.

Yes, that is correct. However, we also see it’s quite rare that a customer converts during the first interaction with us. There won’t be enough articles behind the paywall. If they have read some of our articles before hitting the paywall, we actually see a much higher probability to convert.

More unlocked articles?

With that in mind, you could actually argue it would be better even for the consumer business to have more unlocked articles from time to time. There are more chances to interact with us. That should lead to more people reaching the critical number of interactions, thereby putting them in the group most likely to convert.

In other words, this would be a bigger pool for us to process.

With that analysed, we decided to give it a go. The effect was quite instant.

On the first day, we saw a little increase in revenue. The second day it went up a bit more. And on the third day, it was a significant difference in revenue — up by 20%-25%.

This happened in May, and we haven’t seen this kind of jump since. In August, we saw a 10%-15% increase, but the demand in August wasn’t like it was in June. It will be exciting to see the numbers when we approach the strong demand season in October and November.

The results?

So, we have the consumer business, though it’s a bit tricky to say what the effects are. It depends on what period we compare it with. It’s also quite difficult to take all the campaigns for driving conversion into consideration.

But, let’s give it a try: If we look back one year, there is no doubt we are at much higher levels — in both existing customers and conversions per week, or whatever period we chose. However, looking a year back isn’t correct, to be honest.

Looking a month back might be more accurate. Compared to May, we see no difference at all. Compared to the summer, we see April was better. But that’s not really a fair comparison either: The summer isn’t the best period for converting readers into paying customers.

If we try to break it down even more and check different weeks during the same month, we might see even more accurate numbers. But even then the numbers aren’t easy to analyse. Some months we have more conversions during the period with a lower share of articles behind the paywall. But, some months we saw a lower grade of conversion during that week.

What does that tell us? We have chosen to interpret this as it doesn’t have the highest effect.

If we have any campaigns during the period and if there are news-intensive periods, there is a greater effect on conversions. If we see little effect on the consumer business but quite a high effect on ad revenue, that gives us incentive to keep doing this.

Because we see conversion happens when a reader has interacted with us several times, we think this will be positive for the consumer business in the long run. But, analysing that is a matter for another blog post!

About Håkan Hamrin

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.
x

I ACCEPT