Trust pyramid by Axel Springer
Use this model if…
- …you have made trust the goal and nobody can yet say how to measure it.
- …visits and page views still steer the newsroom, and you want a metric that sits next to them instead of fighting them.
- …you use AI to publish more aggregated content and want to keep track of how much is still your own reporting and original thinking.
It's not the ultimate framework, but as soon as you try to frame it somehow, I think that helps already.
— Janis Kitzhofer, Axel Springer
Reach is only context. Trust comes from quality and continuous user value.
A simple measurement model for journalistic quality, user value, habit and brand trust.
A pyramid of five layers, listed here from the top. The measurement runs from the base up, from originality and journalistic quality to brand trust, and the trust at the top is earned over time.
Brand trust
Do users credit us with credibility, reliability, timeliness and relevance?
Benchmarks, brand tracking, panels
Relationship and commitment
Do users choose us again and again?
Registered users, subscribers, active 1 day or 3+ days a week, retention
Engaged visits
Do visits show real interaction, and how deep does it go?
Article clicks, media starts, page views per visit, time spent per visit
Content value and satisfaction
Does the content create engagement, meet expectations and create value?
Quality reads, recirculation and further interactions, time spent, qualitative feedback, immediate cancellations
Originality and journalistic quality
Do we produce accurate, original content that users can’t get anywhere else?
Exclusive, own reporting, transparent, nuanced, error reports, ASK feedback, letters to the editor
Basic principle
Trust is granted by the audience, through repeated evidence of reliability, relevance, transparency and direct user value.Product and experience trust
Trust is also shaped by the experience of using the product: ease of use, transparency, quality of service, a fair paywall and little friction.Types of evidence
- Perception: studies and panels
- Behaviour: habit and commitment
- Experience: satisfaction and UX
- Quality: scorecards and transparency
The trust and value pyramid by Janis Kitzhofer, Axel Springer. Translated from the German original.
The pyramid started with one sentence in a board meeting where it was agreed that reach was no longer what mattered most. Trust was. Then every department came back with the same question.
Our CEO said at some point that reach is not most important for us anymore, but trust is. And then all the different departments said, okay, so trust is now most important, but what do we do? How do we measure it? What is the action here?
Janis Kitzhofer, Axel Springer
Janis Kitzhofer, who runs editorial insights at Axel Springer, had his own reason to take the question on. He had doubted volume metrics since his first day in the job. Page views and visits show how big a publisher is and what happened yesterday, and a headline or a picture can move them. They cannot show whether the reader found anything worth coming back for.
His answer was to make trust something the newsroom could count. A gut feeling can't be shared or argued with. A number can.
If you can make something measurable, that's always one of the most important steps for something to be democratized, for people to actually make it tangible and make it have an effect. Otherwise, if it's just a gut feeling, the hard foundation is missing.
Janis Kitzhofer, Axel Springer
The result is the pyramid above. Each of its five layers asks one question and names the signals that answer it, with original journalism at the base and brand trust at the top. The arrow beside it says the top is earned over time.
How the pyramid measures trust
The pyramid does not reduce trust to one number. The slide lists four kinds of evidence alongside it.
- Perception is what readers say in studies and panels.
- Behaviour is habit and commitment.
- Experience is satisfaction with the content and the product.
- Quality is scorecards and transparency about the journalism.
Only perception asks about trust directly, and it sits at the top. Axel Springer follows BILD's trust scores in the Reuters Institute's Digital News Report from year to year. Its own user survey, once a big annual exercise, now runs every quarter, with questions on trust that avoid asking readers outright.
What I did, for example, is get these numbers from all the Reuters news reports and see how that developed. We used to do a big annual user survey. We do that more often now, quarterly, and we added more specific questions towards trust without saying, do you trust our brand more?
Janis Kitzhofer, Axel Springer
The rest is read from the layers below, from the share of original reporting up to how many subscribers come back three days a week. None of these is trust in itself. Over time they show whether the newsroom is earning it.
If we start with original content, showing how much of our content is original, seeing that users are highly engaged with our content, they're engaged in every visit, they come back more often, then this all counts towards brand trust.
Janis Kitzhofer, Axel Springer
The limit is plain at BILD. Behaviour can look like trust without being trust.
All this together is only a proxy to get closer to trust, but you don't really know. You could say, if someone returns every day, there must be a lot of trust, right? But not necessarily. If you look at the numbers from surveys, they say BILD is the least trustworthy brand, yet it is the most popular and widely used in Germany by far.
Janis Kitzhofer, Axel Springer