You will be hard-pressed to find a successful digital publication that is not data-led. But measuring everything might lead to having too much data, bringing more confusion than clarity. Or worse than that, relying on incorrect metrics means we end up looking for answers to the right questions in the wrong place.
To start or improve your data-informed editorial strategy, you first need to get your data in a good shape and then make choices about the next action.
"Hard news is not enough to attract audience regularly or with sufficient engagement," says Dmitry Shishkin, digital transformation and innovation specialist well-known for his 'user needs' work at BBC World Service, speaking at a webinar organised by Smartocto, an editorial analytics provider, last week.
"We increasingly see a misalignment between what audiences want in digital and what they currently get from the media outlets."
One of the reasons why audiences are less engaged, according to Shishkin, is because newsrooms lack focus on user needs.
Six principles of actionability
Editors are overwhelmed with data but they do not have the clarity, time and skills to interpret it and deliver change. Although bigger organisations usually have audience engagement teams, smaller newsrooms cannot afford a dedicated data specialist.
To help them out, Smartocto teamed up with Shishkin to start a Triple N project, which stands for news, needs and notifications.
Growth comes when different user needs are addressed consistently, creatively and strategically, says Shishkin, and where content and product together represent the whole offer.
Sort out your 'data spaghetti'
Most editors have mountains of data. This leads them to be overwhelmed at best and draw wrong conclusions at worst. The first step is to define what metrics support your editorial and revenue model.
This can be page views, returning visitors, conversions to paid subscribers, and so on. There is no wrong or right answer here, each organisation has a different model and the metrics you need will depend on what you are looking to achieve.
The next step is to check whether everyone in the newsroom uses the same definition of success and understands how the metrics contribute to that. Getting everyone on the same page is essential to create a meaningful change.
Clean up and organise
Once you determine what metrics you need for success and sort out the data mountain, you need to start measuring what matters.
For example, if your objective is audience growth, you must start measuring new users. For subscription growth, consider measuring conversions from anonymous to registered or paying reader.
"Making decisions based on wrong or incomplete data is even worse than making them based on no data at all," says Shishkin.
Get the right people involved
Once you set up the analytics you need, make sure the data goes to the right people. This includes understanding everyone's preferences and needs, such as receiving a report in an email once in a while or having access to a real-time dashboard.
Do not assume you know the answer and ask your colleagues directly, then make consistent decisions.
Right place, right time
When and where does your audience actively look at your content? To find out, you may need to experiment with different publishing times on all platforms, including your websites, any broadcast channels, newsletters, and social media.
Match your content strategy to these moments and places of engagement, rather than broadly increasing your output. Overwhelming your audience with notifications or newsletters, for instance, will only lead them to unsubscribe, creating a vicious circle.
Test your decisions
To understand how your audience behaves over the course of the day, look out for what sort of stories or topics they consume, at what time and on what channel. Then you can start testing decisions based on this data. For example, if an article performed well in the morning, it should be included in the afternoon newsletter.
Another well-known technique is A/B testing, which allows you to measure engagement based on headlines or where an article sits on your webpage. Ideally, you automate as many of these tasks as you can.
Finally, assess whether the individual metrics led to the desired result you set out to achieve. This is not about numbers, but about the validation of your data and actions so you can keep on improving them.
Join our Digital journalism bootcamp and learn how to create outstanding digital content, master web analytics and hone your multimedia skills.
Free daily newsletter
- 'Anger at "fake news" can stop journalists from admitting - and correcting - their errors'
- Decisions, decisions: weighing up long-term revenue models in digital publishing
- Licence fee, accountability and overseas opportunities central to BBC's future
- 'Actionable user needs' make for more efficient newsrooms
- Direct publishing: what you need to know about this new technology