Skip to content
Go backGo back


Data in the newsroom: in conversation with McKinley Hyden

The FT’s Head of Insights, McKinley Hyden, talks through how data has been incorporated into the newsroom, its wider impact on culture, as well as key commercial metrics.

How has the Insights team evolved during your time at the FT?

Insights wasn’t actually a stand-alone function when I first joined the FT. I initially joined the Data Intelligence team, which now has been split out into different functions: Insights being one, Business Intelligence the other.

When I joined, there were three people: myself (looking after B2B), my manager and another analyst looking after B2C and “everything else” (as he often put it).

To sum up my role now, it’s all about being truly immersed in the business; really understanding the context, and being a translator between analytics and data. We’ve since grown the team to thirteen people.

FT Strategies’ clients often ask when data was first brought into the newsroom. Were you involved from the beginning?

When I started, the FT was already reaping the rewards of making key decisions using data. For example, when optimising subscription conversion.

But this transformation didn’t extend to the newsroom.

Once people finished patting themselves on the back for the great wins from utilising data to transform the FT’s model, there reached a point where there was a concerted attempt to have editorial fall in line with the data-centric model that the rest of the business followed.

How did things move from an idea to action?

Once the decision was made, we brought former Editor in Chief, Lionel Barber, together with (now) Chief Data Officer, Tom Betts, and a few other senior members of the commercial team.

The conversation went something like this:

Non-Editorial People: “Here’s our important metrics, here’s how you use them, now crack on!”

Editorial People: “@%#&?!”

It’s not really so much a conversation as a mandate and there wasn’t an attempt to understand and account for Editorial’s perspective. As a result, everyone agreed to disagree and data stayed out of the newsroom.

Where did things go from there?

Slowly efforts were made to regain the trust that had been lost. It really started to change when we introduced a new analyst role - which I initially took up - designed to help bridge the gap between data and Editorial. I came in to live & breathe all things editorial, to get a much better sense of their culture and to try and shape the data and analytics strategy within that team.

The first year of that project was really spent getting to grips with the way journalists operate within the FT, right down to the way they talk. One of the dirty secrets of analytics is, no matter how good your data is, or how smart your analysts are, if you can’t communicate those insights in an effective way that allows somebody to really understand what you’re saying, and now how to use it, there’s just no use.

The FT’s North Star metric, RFV (see post link on RFV) is used across the business to guide decisions. I’m curious how this was interpreted in the newsroom?

It was initially difficult to really appreciate why editorial couldn’t get more out of the North Star metric. After all, it tells us how much our customers are engaging in the content they produce. But does it? A 90 day metric covering all of a user’s behaviour is actually not very helpful when trying to assess the engagement with a particular article. Editorial had a point so actually the question morphed from “how do we get Editorial to use RFV” to “what is the most useful measurement of engagement for Editorial?”

Is this what lead you to uncover the new ‘Quality Reads’ metric that was rolled out?

Exactly, RFV is a customer-centric metric. This is usually a good thing, but for years, the typical metric used to measure the success of an article was page views - a product-centric metric. The problem with this measure is that page views don't give a real indication of the consumption of an article. It really should just be called “headline click” - that’s all it is.

It also undervalues articles that aren’t going to bring in millions of readers, but are nonetheless providing quality content to customers.

The number one question that journalists wanted answering, which page views couldn’t answer, was: ‘what do subscribers actually think, and are they really reading?’

Over the course of a few years, the insights team created a ‘Quality Reads’ metric, designed to work in tandem with page views. ‘Quality Reads’ was a consumption-based metric: how much of an article has someone actually read? This metric was designed alongside a number of different teams: internal products, multiple data teams and technology teams.

Does this correlate with what RFV tells us?

Yes actually, it does!


So this was a metric that editorial could really get behind?

This is something journalists intrinsically care about. It’s sort of intuitive, really. They want somebody to read most of the article you’ve spent lots of time working on. Previously: a page view, was a page view, but it only gave a surface level insight into the work.

If I was a journalist in the FT newsroom now, what data would I look at on a day to day basis, and how would I use it?

The data available within our analytics tool, Lantern, which is presented twice a day in a meeting with all of the editors.

Lantern data dashboard
Lantern data dashboard
Lantern data dashboard
Lantern data dashboard
Lantern data dashboard

We also now have a newsletter which goes around the entire business which talks about editorial successes across the week, which is great to align the rest of the business around the work that’s going on.

Some data-savvy journalists are now actually coming to Analytics before starting pieces to try and inform the scope of their work.

Do you think data is now informing editorial decisions?

I think it’s a combination, which I’m definitely OK with. There are definitely some limitations in using data. People who have years of experience in the newsroom, can feel things in their gut. That’s not to say data should be ignored. There are probably things when people should use data to inform things, but have got to go with their gut, particularly as they’ve got to work very quickly and it may take Analytics a number of days to pull out the correct insights.

Which is why it’s really important to apply an idea of data-informed journalism versus data-driven journalism.

Finally: are there any key pieces of advice you’d give to media companies looking to start bringing more data into the production of their journalism and content?

I think the big thing is really listening first. Transformation is a two-way process. It’s like any relationship: if you move in with a partner, you don’t walk into a new house you’re sharing with a list of demands. It doesn’t work in your personal life, so it shouldn’t happen in business, even if it often does.

If you want to undergo real cultural change, it’s vital to spend time finding common ground, get a real understanding of how a department works, and the language they speak: that then enables you to create more relevant, empathetic insights, and then start to make compromises along the way.

About the author

McKinley Hyden is now Head of Insights at the Financial Times, having joined in 2013 after graduating from the London School of Economics with a Masters in Management. Prior to that, McKinley lead niche supply chain research at FC Business Intelligence.

You might also like

FT Strategies brand symbol
FT Strategies brand symbol
FT Strategies brand symbol
FT Strategies brand symbol
FT Strategies brand symbol

Get started

We've helped companies around the
world future-proof

their businesses
- and we can do the same for you.