Publishers looking to collect first-party data should skip registration walls and offer low-cost subscription trials instead, according to Norwegian media company Schibsted which has shared its tips for other publishers.
Gustaf Eriksson, executive vice president of consumer business at Schibsted Media, told Press Gazette why free registration walls could cause friction between publishers and potential subscribers when sharing insights on how publishers can best approach first-party data collection.
Schibsted, which owns Norway’s two largest newspapers and the Nordics’ biggest tabloid, has around 1.5 million subscribers across its news brands, including 1.25 million digital-only subscribers.
How has Schibsted built a large online subscriber base?
“Scandinavians are pretty far ahead when it comes to how online they are,” said Eriksson.
Schibsted is heavy on its newspaper production but also owns two large TV stations and a podcast platform. Its business spans text, video and audio.
“So broadband usage was very high from the very beginning in both Sweden and Norway,” he said, adding this meant high consumption of online news since the late 1990s.
Schibsted was then very early to translate these “digital habits and preferences” to an online subscription service in around 2014.
“I think that’s when the first-party data also set off, because when you have online subscriptions, you’re required to have your own account, and that’s the basis of your first-party data,” Eriksson said.
He added this guided the publisher in understanding what readers wanted and didn’t want, which was discovered through A/B testing.
Why move from open web to first-party data collection?
Eriksson said a publisher “cannot succeed in any online business” without collecting data on its audience.
“It translates to not only how we work commercially, but it’s also how we work in the newsrooms.”
He added that analysing open web data signals leads to counting page views encouraging clickbait and sensational headlines.
“If you try to optimise your business to have loyal happy readers and customers, your top list [of stories] is fundamentally different, right?
“And if you look at not the total number of clicks from social media, but you look at what your paying subscribers spend time on reading, it’s a totally different world.”
What registration incentives have proved the most effective?
Eriksson said Schibsted has achieved high levels of logged-in traffic almost entirely through offering subscriptions rather than registration walls, and having “the right paying model for subscriptions” drives the most first-party data.
“The paywall is almost always your best option if you have a premium product…
“If you create a lot of friction to logging in, you may as well have at least a trial period subscription, and you achieve the same thing.
“But the behaviour from a paying customer, at least the trial-paying customer, is a stronger relationship than just a login user.”
He added you can “achieve quite a lot” using a straightforward paywall instead of a metered paywall but you have to begin with a “low-threshold offer”, such as £1 for a month then £9.99 per month thereafter.
Skipping a registration wall and going straight to the paid paywall achieves “two things at once”, Eriksson added, receiving payment and data from users.
Why is first-party data so valued?
“I think first-party data is definitely more sought after than any other type of third-party data in the ad business, right?” Eriksson said, adding it has higher commercial value, and it is stronger when attached to a logged-in user and has high frequency – a user logging in multiple times.
“It’s very robust. You can see how many people [you reached], not how many cookies have I shown [someone] to.”
First-party data is also beneficial for targeting specific audiences: “If it’s a geographical area, or if it’s an age group, or if it’s based on the gender… it’s very tightly linked to that type of user groups’ behaviour, and what they want, and what you can offer”.
What are the common pitfalls in first-party data collection?
Eriksson said creating “friction” at a paywall checkout by complicating payment options is a common mistake.
“If you’re starting out, I think it would be viable to have a look at what type of account solutions you can tap into,” he said. “I do not know what’s available in the UK, but in Sweden and Norway, there are some that are linked to the banking system where it’s basically mobile pay has one solution, and there are others [where]…you can tap into something where they already have an account, [so] you can then you skip one step.”
How should different sized publishers approach first-party data collection?
The approach should be the same, said Eriksson.
“I think you should have a culture of experimentation,” he said, describing this as being “agnostic” and not yet knowing what’s best for your newsbrand.
“You should have the instinct to test before making long-term plans, that should be true across publications, but the results of what you find out might be different…
“Once you reach three successful visits a week, and you see that churn on that segment is very low, what about these unhappy users?” Eriksson said.
“Who are they? What do they want? Can we ask them? Can we look at their general site [data], what they read, and provide more of that? It leads to the right questions internally of what you do to create as many happy customers as you can.
“You’re user-centric, almost like a religion… it drives behaviours that lead to better products, and I think it leads to a stronger foundation for us as publishers, as well.”
Should publishers collect the data themselves, or go to an outside company for data collection?
Eriksson said smaller publishers should learn from competitors and use consultants where needed, while larger organisations could benefit from building an internal, dedicated team.
“I also think you need to be, you need to learn to fail fast, right?” he said. “I think just keeping an open mind and have that as part of your planning is the first piece of advice.”
Tiered subscriptions versus one-size-fits-all
Eriksson said publishers should match products to different levels of willingness to pay, “and the question then becomes, ‘how can we design tiers that are both valuable and user-centric?’”
He said the execution of this is the “tricky” part, but an example is offering different products like an e-paper for older readers.
“Another feature could be like, if you have a family, you’re probably more established, so family sharing is another product feature that fits well with a higher willingness to pay, but you need to figure out what works for you and what fits your context.”
How should publishers approach going from open site to paywall?
Eriksson urged publishers moving from free to paid models to collect first-party data to prioritise audience growth over making money fast.
“Try and build volume, it’s more important than building quick revenue… Reach your audience first and then try to work with price over time rather than doing it the other way around.
“If you do it because you’re in financial peril and you need the money fast, I think you won’t reach your full potential. Try to find yourself in a situation where you can afford to look at this long term.”
[Read more: Schibsted signs deal with OpenAI]
Email pged@pressgazette.co.uk to point out mistakes, provide story tips or send in a letter for publication on our "Letters Page" blog
