Written by:
Alexandra Borchardt
Data-driven decision-making remains a challenge for the industry. Here are some practical tips from Anja Noster on how to succeed with data-driven work – from selecting KPIs along the funnel to continuous monitoring with the right tools.
“You need metrics from which actions can be directly derived”
You develop dashboards for and with public-interest media organizations. How do founders develop a data strategy that aligns with their strategic goals?
At the beginning, it’s not about the dashboard, but about the company’s strategy. Most of the time, the problem isn’t a lack of data, but a lack of prioritization.
When I work with a media organization, we start with a workshop involving all relevant team members. These are usually the founders, often supported by someone from community management or marketing. In the first step, we go through the following together: What is the company’s goal? What do they want to achieve, and what data do they need for what? For example, is the company primarily focused on reach, revenue, or user engagement?
Once the goal is defined, in the second step we look together at the user journey along the funnel. We map out how users become aware of the offering, what happens on the website or landing page once they’ve discovered the project, and how the relationship is maintained after an initial conversion. If the goal is newsletter sign-ups, for example, we analyze the entire process: Where does the traffic to the landing page come from – social media or offline advertising? What convinces users to sign up? And what happens after sign-up to turn subscribers into loyal readers?
In the third step, we take stock of the existing data sources. The content management system, the newsletter tool, the CRM, Meta Business Manager, the website, the payment provider – all of these generate data. We examine exactly what that data is. Building on this, we define: What should the key metrics be, how frequently do they need to be tracked, and who is responsible for this.
It’s crucial not to be blinded by metrics that suggest some kind of growth but say nothing about the strategy’s success – so-called vanity metrics. You need metrics that directly inform your actions – in other words, actionable metrics. For example, if you need reach, page views are naturally important. If you want to build a community, you’re better off looking at how much feedback you get on the newsletter or reactions to a post. In any case, it’s important to think about how you want to measure success right from the start.
Many employees are uncomfortable with data, especially journalists. What is the best way to communicate the data strategy to the organization, and how do you follow through on it?
It’s important to clearly communicate how the data ties into the company’s strategy. Numbers shouldn’t stand alone. Employees need to understand what they’re supposed to do with the data and what conclusions they can draw from it. That’s where things often get stuck. Especially in newsrooms, there are often skeptics who say, “Our data alone doesn’t say anything about the quality of our journalism.” Of course, there’s some truth to that: just because you have metrics doesn’t necessarily mean you have good journalism. But you can optimize a lot with the help of metrics.
That’s why visualization plays such a big role. It’s the only way to identify trends over time. This is where dashboards come into play. It’s good to have a company-wide data tool that everyone can access. This way, everyone works with the same data, and no parallel truths emerge in individual departments. It’s effective to have a dashboard within the data tool where all employees can immediately see whether the company is heading in the right direction – for example, based on key reader and revenue figures.
At the same time, role-specific dashboards are needed; the marketing team focuses more on conversion rates and wants to optimize advertising spend, while the editorial team looks more at reading time and interactions. Not every role needs every metric. It is therefore beneficial if every role in the company can use the data tool independently, but one person bears overall responsibility – for example, a business analyst or a growth manager – who ensures the data is presented clearly across the various dashboards and serves as the point of contact for the others. It usually makes sense for this person to be one level below the founders and to understand the language and goals of the editorial team, the tech team, and the revenue team.
Established routines are also important. Ideally, you should hold a meeting with all relevant employees, during which key metrics are discussed and courses of action are defined.
What tech skills and tools are absolutely necessary to implement a data strategy, and which ones are “nice to have”? Please explain this for beginners.
Solid Excel or Google Sheets skills are a must. Of course, you shouldn’t be afraid of numbers in the first place. Basic statistics help, but you can also read up on them or use AI for assistance. Technical skills like SQL, Python, and experience with APIs and data warehouses are also helpful but aren’t absolutely necessary at the start.
I always work in two phases. In the first phase, I test which data infrastructure makes sense. I do this entirely manually. I export CSV files weekly from the most important systems – CRM, newsletter tool, payment provider, web analytics, Meta Business Manager, and so on – and upload them to Google Sheets or Excel. One tab per data source, neatly structured, with one dimension per column such as region, date, open rate, revenue, and so on. And please, no color coding! I usually add three months of historical data as a starting point. In principle, you can already perform analyses in Google Sheets or Excel itself. However, based on this, I visualize the most important metrics using a data tool, such as Looker Studio, if the team is already working with Google Suite. This is cost-effective and quick to implement. My goal is to quickly understand which metrics are truly relevant and used regularly.
Only once you know what you need can you move on to phase two: automation and building interfaces. In the beginning, for example, you can build a lot using Zapier or Make. But proper data warehouses are a whole different story. There, you can tailor the technical setup to your specific needs and, for instance, rely on providers other than Google or work with open-source solutions. If you don’t have the necessary expertise yourself, it’s best to collaborate with developers who have experience with APIs, webhooks, and data warehouses.
In summary, my most important message for beginners is this: You don’t need a complex data warehouse or a large tech team at the start. You need clarity on which questions you want to answer. The technology follows the strategy, not the other way around.
Dr. Anja Noster is a freelance coach who advises media companies on developing and implementing data strategies. In parallel, she conducts research on journalistic innovations at the Hamburg Media School. Previously, she was responsible for strategic growth and publisher relations at the media startups ada and Opinary for several years. She earned her Ph.D. in journalism funding from the Bauhaus University Weimar.
Last updated: March 27, 2026
