Editor’s note: This is one of 17 case studies featured in INMA’s strategic report “Making Big Data Smarter For Media Companies,” released in December.

The metered model of Financial Times’ Web site has allowed it to build a sizeable database of users and their behaviour. Its direct relationship with its customers gives it a unique view of its audience across various channels and devices, says Tom Betts, who oversees data analytics at the Financial Times and is vice president of customer analytics and research for its parent company, Pearson Professional. 

Because its customers log in wherever they read Financial Times content, its team is able to triangulate usage data to give them complete visibility of where, when, what, and for how long readers engage with its content. 

The company uses customer profile information — including topics of interest, time of access, and type of device — to develop a unique “customer DNA” for every user. It describes the user’s behaviour and demographic data, as well as many other derived attributes, such as customer segmentations, Betts says.

Financial Times uses insights from the customer DNA to guide actions, such as using a “trigger” to initiate an e-mail or telephone call to a prospective customer to tell them about a new product or service of interest.

Financial Times uses this data to understand how customers engage with its products and where they do or don’t meet user needs. But for years it has also used the data to drive marketing efforts, building new audiences and engaging existing audiences.

Its data analytics allow its team to identify and predict, with high precision, the tipping point where people might begin to pay for its product. It uses this intelligence to power its marketing activity.

Betts ascribes more than 15% of its digital subscription growth to marketing activity directly linked to consumer behaviour. 

In recent years, Financial Times has had considerable success in improving subscription revenue through a variety of data-led initiatives, Betts says, including the optimisation of its customer acquisition efforts and the engagement of its customers to improve existing relationships and thus its retention rates.

The media company also uses data commercially to offer smarter product development and marketing, including managing renewals and personalised products.

Similarly, its business-to-business sales team uses data to understand and manage corporate relationships, giving the team insight into where demand for the publisher’s content is greatest. 

“Rather than competing for volume, publishers like the Financial Times should position themselves as niche players that offer unique value and deep engagement with their audiences,” Betts says.

“Our direct relationships mean that we have proprietary first-party data that our advertisers can trust and use in reaching the audiences that matter to them. This information allows us to offer improved measurement, segmentation, and targeting, helping to grow our digital advertising revenues.” 

Beginning with a pilot programme in 2011, FT.com has used a content-matching service called SmartMatch, which is provided by the content marketing firm Smartology.

SmartMatch goes beyond the industry standard of keyword recognition, using instead an emerging technology called semantic profiling that analyses text through weighting concepts and categories to match relevant display ads with news articles in real time.

SmartMatch has helped increase campaign performance by as much as 20 times, Betts says. 

Financial Times uses Deep View 2.0 to offer enhanced reporting on campaigns for advertisers on its Web application, Betts says. This enables it to report back to an advertiser not only on the number of impressions and the click-through rate, but also the types of users exposed to the campaign and which actually clicked on their ads.

Mobile advertising is showing 9% growth year on year. 

In October 2014, the publisher began experimenting with using attention-based metrics to increase advertising revenue. It has been working closely with Chartbeat to measure not just whether an ad is seen, but for how long, Betts says.

The amount of time the target audience is exposed to the advertiser’s message directly affects its impact, as was seen in a 2014 study that showed readers who saw an ad for at least five seconds experienced, on average, 79% greater brand recall. 

The company can tell its advertisers how long each impression has been viewed and the total duration of exposure across the campaign. Says Betts: “This is a metric that we feel is closer to the actual outcome an advertiser is seeking. Impressions themselves are just a convenient mechanism to trade.” 

Financial Times developed its core analytics platform and data warehouse internally. An internal engineering team manages them, but it has recently migrated this infrastructure to Amazon’s cloud-hosted database, Redshift.

The company was an early adopter of this technology, and migrating its data warehouse to Redshift allows its team to work faster and scale its capacity in minutes to crunch large volumes of data.

While core functions are managed internally, the company works with technology partners for certain analytical solutions, using an advertising data management platform from Krux and Web analytics tools and instrumentation from iJento. 

Not every data project has been a success. Betts believes it is impossible to innovate without sometimes encountering failure and that “accepting this is an important part of building an innovative and data-driven culture.”

“Some things just don’t do what you expect. We have been using data extensively for a number of years and have built a significant capability around it. We do a lot of multi-variate testing, much of which is ‘unsuccessful’ — insofar as we do not manage to improve upon our baseline results.”

Betts believes that data analytics leading to action almost always work better when data specialists and subject matter experts collaborate.

“Data analysis in isolation, either internally — or worse as an outsourced activity — has a high rate of failure,” Betts says. The Financial Times has found that democratising data and putting it into the hands of those who will use it increases opportunities to scale the benefits of that data.

The publisher is alert to privacy concerns and has a transparent policy that gives users clear information about the cookies it uses, empowering them to make informed decisions about allowing the use of cookies, Betts says.

It doesn’t analyse and use data on an individual level — only by general demographic cohorts — and it doesn’t grant its journalists access to its subscriber databases.

Looking toward future uses of customer data, Betts believes there are applications for analytics that can succeed across news media companies. Every company can benefit from having a greater appreciation of the customer.

He believes an analytically mature company is one where most data analysis happens outside an analytics team: “Non-technicians or data specialists should be able to use data to help bring them closer to the customer and their needs, attitudes, and behavior.

“If we are to be nimble and able to adapt quickly to changing needs of our readers and changing environmental conditions, then we need to equip people with the necessary tools and skills to use data successfully,” he added. “We have made great progress in democratising our data, but I would like to equip everyone across the organisation with a view of our customers that is tailored to their specific needs.”

This doesn’t mean everyone should follow what the data tells them. The goal is not to be “data-led” but “data-informed.” Says Betts: “Understanding the difference there is crucial, because we’re not in volume game, but the quality game. And that is exactly the kind of existential problem that I think we can use data to help us crack.” 

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .