Kesewa Hennessy, the Financial Times鈥� digital editor for audience engagement, describes the newspaper鈥檚 past struggles and current successes with understanding its women readers, exploring the benefits of digital tools as well as their limitations.
The early, unofficial prototypes of the Financial Times鈥� women subscribers鈥� engagement projects were inauspicious. They started with women whispering in office kitchens about the urgent need for a more balanced output; with rudimentary spreadsheets comparing the number of women and men featured in bylines and skylines; and a on which three of the four writers were white men called David (the fourth was a white man called Michael).聽
The hope 鈥� back when this was still controversial (hence the kitchen-whispering) 鈥� was that data, however elementary, would call attention to the disparity in a way that simply saying that there was a clear imbalance hadn鈥檛. The hunch at the outset was that very 鈥榤ale鈥� coverage would be reflected in the make-up of the audience; and that a more balanced output would help broaden the subscriber base. Five years on, the FT has a company-wide goal of increasing women subscribers鈥� engagement championed by ; a growing cluster of well established newsroom projects and processes aimed at doing just that, which I lead; and a steady rise in engagement.聽
The first official women鈥檚 engagement experiment was launched in the newsroom in 2017: Project XX (named after the chromosome), where one story a day likely to be read by an above average percentage of our female subscribers was promoted high on the home page. Supportive senior editors, including Khalaf, then deputy editor, made it permanent after a trial showing clear engagement benefits. Our initiatives now range from dedicated promotion channels and targeted products that help ensure relevant content reaches the target audience, to editorial workflow systems and AI tools.
These initiatives, in all their disparate forms, broadly share the same simple goals. They're about increasing the engagement of our women subscribers, currently about 25% of the total 鈥� primarily by encouraging changes to both content and newsroom culture that will help the FT feel more relevant to a section of the audience who are already signed up but significantly less engaged than men. In focus groups conducted a few years ago when data first confirmed the disparity, a majority of women explained that they found the tone off-putting. If the FT were a person, they said, . So we set a goal of changing women subscribers鈥� perceptions 鈥� in the long run increasing their engagement and in turn building their loyalty. Internally, we want to learn more about what women subscribers consume and to encourage culture change to help us better meet their demands.
Collecting the data
The initiatives share a few common features. First, they鈥檙e all founded on models developed by our data science specialists that, since 2016, have informed us about what women subscribers read. The audience engagement team uses this information to develop and implement initiatives, tailored to FT newsroom culture, that help our journalists meet these readers鈥� demands.
Our understanding of women subscribers鈥� distinctive reading patterns is made up of several complementary metrics. More significant than high traffic (although, as with groundbreaking investigations into allegations of sexual misconduct in business like the , that is always welcome) is the percentage of women鈥檚 page views for an average FT story. This gives us a benchmark against which to measure the popularity of topics, sections and, importantly, stories for this smaller section of the audience.聽
Close monitoring of 鈥榦ver-indexing鈥� stories (in practice, that鈥檚 me staring at a dashboard of data every morning), more so than broader categories such as topics, has proved integral to how the projects work. It allows us to spot trends and share the information across the editorial and commercial parts of the business, informing departments from events to marketing. Crucially it also allows us to understand not just what has over-indexed but also to predict with confidence what will. It gives a fairly accurate feel for hard-to-quantify elements such as tone, too.聽
Within the FT, focusing on how individual stories fare means we can swiftly pick up on the distinctive ways women engage with new topics that dominate the headlines, such as coronavirus, the death of George Floyd and Black Lives Matter, and US president-elect Joe Biden and vice-president-elect Kamala Harris. This has practical value: for example, it helps us ensure headlines on these topics include keywords that reflect the aspects women are reading about.
Over the longer term, it helps us correct long standing misconceptions (鈥榳omen don鈥檛 read about 鈥渃ore 贵罢鈥� subjects鈥�; 鈥榤en don鈥檛 read about 鈥渟oft鈥� subjects鈥�). We can spot unexpected patterns (the obscure corners of that consistently over-index) and pinpoint newer ones (; the ). It highlights the demand for stories long dismissed as 鈥榥iche鈥� ( and diversity, managing , , the ) 鈥� and that the idea of 鈥榥iche鈥� is itself a myth (over-indexing stories are always read by more men than women, including those on the and, yes, the menopause). All told, this creates a more realistic picture of our audience as a whole.
Interpreting the patterns
Externally, understanding these patterns helps us to change audience perceptions by promoting more of our relevant content. Alongside Project XX, dedicated promotion channels include , a weekly round-up email of the FT鈥檚 biggest stories and best reads. It鈥檚 curated by women journalists featuring stories popular with women in a format that women told us they liked. Neither the home page slot nor the newsletter is overtly branded 鈥榳omen鈥檚 content鈥� 鈥� women told us in customer research they didn鈥檛 want that 鈥� but they both receive above-average engagement from women.聽
Understanding these patterns means that, as well as changing perceptions through promotion, we can produce more relevant content. It broadens the range of commissioning options for editors and writers, for example. Information on what engages women is also included in the specifications of new editorial products, such as , a platform showcasing the FT's coverage of climate change, business, markets and politics.
All these initiatives bring together expertise from across the business. Take November鈥檚 virtual Motherhood, Money and Making Career Moves talk. It鈥檚 our second event in partnership with , a British media platform featuring journalism for and by black women. The project involves the FT鈥檚 live events specialists and the personal finance and podcast sections, as well as the audience engagement team.聽
To encourage our fellow journalists to take part in these projects, we emphasise how they help increase their audiences 鈥� in particular, the fact that more representative output can make your work more relevant to a wider audience. FT analysis shows women are more likely to read stories featuring women.
So, for example, many FT teams have volunteered to join the BBC-led 50:50 project, tracking the balance of women and men writers, experts and other contributors in their output. The long-term aim is for each desk to reach a roughly equal monthly ratio. Asking busy journalists to spend extra time filling in spreadsheets for little direct benefit is not the easiest sell. Yet many in editorial and beyond have already signed up 鈥� very often because they believe in the principle but also because it can help them reach more readers.
It may also seem counterintuitive to ask colleagues to count imbalances by hand when, like a lot of newsroom processes 鈥� including several of our FT experiments 鈥� this could be automated. Automation has proved useful in providing a macro-level picture of imbalances in output but we鈥檝e also encountered limitations, including in-built bias. The 50:50 project has led a shift back to a more manual approach, closer to the kitchen-whisperers鈥� basic spreadsheet. This appears more effective as a way to raise journalists鈥� awareness of imbalances and encourage culture change. It鈥檚 easier to ignore a Slack channel quietly pumping out data than a spreadsheet you fill in by hand after poring over quotes, pictures and bylines in your own stories.聽
What next? As we learn more about our target audiences, focus on new ones, and this year in particular respond to big news events, the strategy continues to evolve. We are investigating the longer-term effects of the coronavirus pandemic on women鈥檚 reading habits. We鈥檝e focused specifically on American women during the election as part of a US growth strategy. And since the killing of George Floyd, plans to move beyond gender as we broaden our audience have gained momentum. I鈥檓 looking at what we can transfer from the women鈥檚 projects to help build a more ethnically diverse audience 鈥� the 50:50 Project, for example, and promoting through specific platforms, like Instragram 鈥� while acknowledging that it will inevitably require a significantly different approach.
And what about those who don鈥檛 wish to participate in such projects for whatever reason? Experience says time and energy is more efficiently spent working with those who do. Others will often join eventually 鈥� when critical mass is reached and/or when the benefits of participating clearly outweigh those of not doing so 鈥� and they鈥檙e always welcome. At the FT, as it turns out, there are enough people willing to take part to make it work.聽
Engaging with women readers begins with good data聽
-
Don鈥檛 ask permission, just start collecting data. Tracking by-lines and who is quoted in articles are good ways to get an initial indication of gender imbalances.聽
-
Build momentum for change: harness the demand for more representative output that is likely to arise naturally in younger, more female and perhaps eventually, more racially diverse newsrooms.聽
-
Find a few supportive senior editors willing to champion the cause.聽
-
To build women readers鈥� engagement, establish data models, set measurable company goals and lobby for dedicated newsroom resources, including for new roles.
-
Focus on demonstrating, through data and research, the quantifiable costs of not evolving and the business benefits of making your newsroom鈥檚 output more representative.
Every news organisation will approach this in the way that works best for them. The Financial Times women鈥檚 engagement projects are closely tailored to our specialist content and the culture of our newsroom. But, If someone starting from scratch asked, I鈥檇 offer a few simple suggestions: Unless you鈥檙e willing to wait decades, don鈥檛 ask permission. Instead, just start collecting data.
There鈥檚 no need for special skills or tech 鈥� tracking bylines and quotes in a single section using a basic spreadsheet is a good way to get an initial indication of any imbalances. The BBC鈥檚 50:50 project is a useful model. Harness the demand for more representative output that is likely to arise naturally in younger, more female and perhaps eventually even more racially diverse newsrooms (94% of UK journalists are reportedly white; 0.2% are black). It鈥檚 likely that colleagues will want to join you. Find a few supportive senior editors willing to champion the cause.
At the FT, this hastened the shift from kitchen-whispering stage to data models, measurable company goals and dedicated newsroom resources, including new roles. Focus on fairness, yes. That鈥檚 what motivates employees from already marginalised groups 鈥� and by definition often in relatively junior positions 鈥� to spend their private time and whatever scarce professional capital they may have pushing something that may not make them popular with their bosses.
But to get beyond heartwarming public statements about the importance of diversity, focus on demonstrating, through data and research, the quantifiable costs of not evolving 鈥� and, conversely, the business benefits of making your newsroom鈥檚 output more representative.
Kesewa Hennessy is the Financial Times鈥� digital editor for audience engagement.