Skip to content
Go backGo back

Article

How the Financial Times brought data into the Newsroom

Data has been at the heart of digital transformation at the Financial Times.

Data has been at the heart of digital transformation at the Financial Times. As news consumption has shifted from print to websites and apps, it has become possible to build a deeper understanding of our readers. Where we once only had relatively crude measures of performance (e.g. overall circulation), hundreds of data points are generated from just a single reader’s visit to ft.com.

Understanding more about our readers has allowed the FT to pivot our business to digital subscriptions, and create the sustainable revenue that allows us to sustain our journalistic mission. While newsrooms are no stranger to technology driven change, new business models have required a reimagining of how editorial teams use data alongside journalistic instinct and editorial integrity. Our experience has helped us distil four key principles that can help bring data to the newsroom.

Principle 1: Know the data points that serve your overall strategy

One of the challenges of the Big Data era is cutting through the overwhelming volume of available information to discover what really matters. As a news organisation, one obvious place to start is in how our readers are using our website and app. Even with that focus in mind, most analytics tools can provide an overwhelming amount of information: page views, social shares, visit frequency, average time on page, bounce rate, and many more.

One of the most important steps at the FT was building alignment on which of these measures matter most. Ultimately, this was informed by our long term strategy – a North Star goal of reaching one million paying subscribers by 2020 (see ‘How the FT went from no digital presence to over a million paid digital subscribers’ on the FT Strategies blog). This North Star spoke to our belief that our future sustainability required a shift in digital revenue – from an ad-centric model, to a subscription-first one.

In ad-centric digital business models, volume metrics (e.g. page views) matter most – the more readers who hit a particular article the more ad revenue you can earn. In the world of digital subscription revenue, the economic logic shifts: while reach remains important, we become much more interested in a far smaller subset of readers who have the potential to generate far higher annual revenue per user (usually £100s per year, rather than a few pence per reader in an ad-driven model). Once you accept that not everyone will be a paid customer, your interest moves to increasing the proportion of readers who do convert, and in retaining them once they do subscribe.

Our own data analysis showed that the user behaviour with the strongest link with acquisition and retention was engagement. Specifically, we found that frequent and deep usage of the FT correlates strongly with a high likelihood to subscribe, and a low likelihood to unsubscribe. Industry-wide studies have confirmed this link at countless other publishers, whether you are a large publisher like the FT or a local news provider. For example, Medill Local News Initiative analysed 13 terabytes of data from the Chicago Tribune, San Francisco Chronicle and Indianapolis Star, and found a similarly strong correlation between regular habit and subscriber retention.

Graph
Graph
Graph
Graph
Graph

With the FT’s North Star goal focused on reader revenue, it followed that our editorial teams should be prioritising reader engagement over traffic volume. With this in mind, we were able to begin the work of making data actionable for the newsroom.

Principle 2: Treat editorial staff as your customers and build tools that meet their needs

Before we shifted our strategic focus to engagement, the newsroom used data only sporadically, with many editorial decisions being made on gut instinct. One early experiment with bringing more data into their work was a reporting dashboard for editorial called Bettsy – which included editorial metrics (e.g. comments, page views) alongside commercial metrics (e.g. registrations per article).

Data table
Data table
Data table
Data table
Data table

Unfortunately, Bettsy struggled to gain more than a small cult following. When we interviewed journalists about their experiences of Bettsy, we discovered some of the reasons it had not been more widely adopted:

  • The sheer volume of data points and use of colours were intimidating
  • Analyst speak was a turn off to a non-technical audience (e.g. “over-index”, “distribution”, etc)
  • Journalists had low tolerance of sorting through dashboards looking for their own insights

The overarching problem was that we were telling editorial what they needed from a Business Intelligence tool, rather than focusing on editorial needs.

Learning from these mistakes, we took a radically different approach. We decided that we needed to get closer to our journalists.. One practical step was to physically situate Analysts in the newsroom, working alongside journalists. Their sole focus and attention was on the newsroom, helping our journalists make sense of reader behaviour.

By spending more time understanding journalist’s needs we were able to replace Bettsy with Lantern. This new dashboard added jargon-free definitions to metrics, used clearer visualisations and provided benchmarks. These benchmarks allowed journalists and editors to quickly see whether content was under- or over-performing. Thanks to ongoing development and far closer communication with our editorial colleagues, Lantern has been widely adopted throughout the newsroom.

Lantern dashboard
Lantern dashboard
Lantern dashboard
Lantern dashboard
Lantern dashboard
Principle 3: Do the hard work to find the right metrics for journalists

Newsrooms operate at a frantic pace. As our experiences with Bettsy show, journalists simply do not have the time to wade through hundreds of data points to find insights. So when it comes to metrics, less is more (and the more actionable the better).

In early versions of Lantern, journalists struggled to make sense of conflicting metrics and naturally gravitated to page views. With the FT’s focus on engagement, this is not the most important criteria for success. For a start, page views only measure intent to read – and cannot capture whether someone actually read an article (or got some value from it). Another consideration is that page views disadvantage niche content that might be supporting a smaller group of loyal readers.

Elsewhere in the business, many teams had adopted RFV as a useful engagement metric. RFV is a score generated for each of our readers. It is based on days since a reader’s most recent visit, the frequency of their visits in a given time period, and the volume of articles read in the same period. However, this metric was less helpful in the newsroom: RFV applies to an individual reader, not a piece of content.

To address these challenges, our analytics team developed the Quality Reads metric. Quality Reads is an estimate of whether someone read at least half of an article (based on word length of an article, time spent on the article, and the average reading speed of an FT reader). It is expressed as a percentage of all readers who clicked through to the story, who didn’t hit the paywall. It allows us to measure how engaging an individual article is.

We discovered that Quality Reads are linked to positive results when it comes to digital subscriptions. Non-subscribers that have visits including Quality Reads are more likely to subscribe, and subscribers who have more Quality Reads are less likely to churn. As such, Quality Reads allows the newsroom to assess the success of articles in a way that is meaningful in the context of the broader organisation strategy.

Principle 4: Use data to drive long term newsroom strategy

Making data actionable in the newsroom reaches its culmination when you link it to a long term goal or strategy. For example, editorial teams at the FT have a long term goal to focus on value over volume. Specifically the goal is to reduce the volume of content created by 15% each year in order to focus more time on quality journalism. To this end, the newsroom have found it helpful to compare Quality Reads with High Page views.

Infographic
Infographic
Infographic
Infographic
Infographic

By comparing along these two axis, we can divide content into:

  • High performing (High page views and high quality reads)
  • Niche but engaging (Low page views and high quality reads)
  • Need a closer look (high page views, low quality reads). In this case, the headline might have been misleading, the article might be slightly too long (perhaps we could have added prompts to help reader get through the piece?)
  • Candidates for cut (low page views, low quality reads). In this case, we should question whether we should have commissioned this content.

This is just one way that a newsroom can make engagement metrics actionable. In a 2019 INMA interview, Yasmin Namini spoke about her experiences at the New York Times. In order to drive frequency, she suggested that the newsroom could change the homepage curation, and make use of newsletters and alerts. To drive depth, the newsroom might plan follow-up articles, use “related stories” and other forms of cross linking.

Key to this success of such measures should not rest solely on the shoulders of journalists armed with dashboards. At the FT, these initiatives are managed by an embedded audience engagement team. Similar teams are present in most organisations focused on subscriptions.

Another example of how this might be structured comes from The Athletic, which hit its 1 million subscriber milestone last September (despite the challenges of a global sports shutdown). As INMA reported last year, a huge part of their success story has come from a data-driven and reader-focused workflow. This workflow sees several teams working closely with the newsroom: the Editorial team creates content distributed by the Engagement team, with well performing articles promoted with paid marketing by the Growth team – with the Data Analytics team providing feedback to each team on how content is performing.

Workflow model
Workflow model
Workflow model
Workflow model
Workflow model

How FT Strategies can help

Over the past year, FT Strategies has helped several publishers with the challenge of bringing data in the newsroom. We’ve seen that knowing where to start can be tricky, not least because this topic speaks to some of the biggest challenges of digital transformation.

Firstly, there are several common technological challenges: rationalising data infrastructure, selecting the right metrics and creating segments that enrich your data. Secondly, there are cultural obstacles: introducing editorial teams to a data-informed mindset, while helping analytics teams empathise with the specific needs of the newsroom. Finally, for this work to have an impact it needs to tie into a clear digital strategy, supported by organisational alignment in processes, tools and incentives.

These are exactly the types of questions that led us to create FT Strategies. So we’d be delighted to hear from you if you are about to embark on this journey. We have worked with all sizes and shapes of business in various formats – from 2 week design sprints through to longer-term engagements. And we’re always happy to start with a (virtual) coffee.

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.