July 8, 2014

In-depth: Mark Beckwith, Head of Database Marketing at Financial Times

Mark Beckwith, Head of Database Marketing at Financial Times, will be among the keynote speakers at the Figaro Digital Marketing Conference in London on 17 July 2014, where he’ll be talking alongside Jade Tanner, Customer Success Director at Pure360. We caught up with Mark ahead of the event to hear about some of the issues he’ll be discussing.

‘Keeping Audiences Engaged in 2014’ is the title of your conference presentation. Can you give us a quick introduction?

One of the issues I’ll be discussing is how we leverage the information we get through predictive analytics. For us at the FT that means working with our data science toolkit to better understand our customers – where they are in the lifecycle, what’s relevant to them and the sorts of offers and messages we put in front of them.

We have our own in-house data science team with whom we work very closely. My own team sits between those data scientists and the marketing and customer services teams, who are at the coalface, if you like. My team’s job is to turn all that technical wizardry into something we can leverage value from. A lot of what the data science team does involves segmentation and data modelling. That really enriches our view of subscribers and lets us tell the marketing teams that a certain group of people are ready for a certain kind of communication or offer.

With so many sources of information about consumers available to an organisation as large and diverse as the FT, how do go about understanding that lifecycle journey in detail?

There are a few things we use which have generated significant success. The first is predictive propensity modelling. The FT’s subscription model means that anyone can register and gain limited, monthly access to content before they hit the paywall. Once users reach their monthly limit, they’re locked out until the next month begins.

We look to see who are the most engaged of those registered users. We look at the people in the buying window and the type of content they’ve consumed. Has something piqued their interest or made them get more involved with the FT recently? Is their activity typical of other people who then go on to subscribe? What’s their demographic profile? We put all that information into the mix and get a view of our most likely targets – the people we can put an offer in front of.

The other tool we use to understand our subscribers involves segmentation: we segment people according to how engaged they are, what they’re looking at and so on. They key thing here – and how we gain value from this data – is watching for changes in behaviour. For example, if someone’s been engaging on a regular basis and then suddenly that usage goes down. It’s incredibly important for us to spot that behaviour as soon as we can so we can proactively go out and engage the user with some form of relevant or related content to try to pull them back in.

We’re not just concerned with the size of our database, however. It’s the lifetime value of those consumers as well. This information helps us better target communications through the lifecycle. With renewal – or if a subscriber’s payment has failed – we can make an intervention, based on that subscriber’s value to us in the long-term.

So you have an agile model where you can respond to users rapidly, according to what they’re actually doing?

Yes. With our subscription model, even to access any of the free content, you have to log in. As a consequence we have access to every interaction on the site. That information is refreshed twice daily so we can see which articles someone’s reading, how they’re flowing through the site.

We integrate that information with our email interactions and we can also see from our email vendors who’s clicked on what. We’re in the process of integrating our customer service data into the data warehouse as well. This means we can see what we’ve pushed out to a customer and any incoming interactions; we can have a complete 360-degree view of what a user’s doing and what they’re feeding back to us. We’re also in the process of re-platforming our data warehouse to Amazon Web Services’ cloud infrastructure. That will give increased speed and decreased latency so we’ll be able to act on those triggers much more quickly than we can now.

If you had a couple of first steps for marketers who want to engage (and re-engage) users more efficiently, what would they be?

Be relevant and timely. That’s golden.

Do you have an optimal timeframe for engagement? Is there a point at which it becomes too late to send out a communication?

Well, to flip that question on its head slightly, let me give you an example of the benefit of getting to a user quickly.

Our FT subscriber ‘Welcome’ email used to go out on the first or second day after a sign-up, so we had some latency. We sent out a confirmation email straight after subscription, but you wouldn’t get the first content-rich, personalised sign-up email for a day or so.

Now we can get it out within 12 hours and open rates have increased by about 50 per cent. There’s been an uplift in clicks as well. A ‘Welcome’ communication is slightly different to something sent out after the event, so it’s difficult to say what’s too late – it depends on the product or the service you’re marketing.

For us in publishing, particularly in news, you have to be very close to the event because the news agenda changes so quickly. In retail, where the purchase consideration is longer, you don’t necessarily have to react to a click on a website within minutes. If you did that within a couple of days, you’d probably be okay.

Any other advice for more effective engagement?

Accuracy is vital. As a consumer, nothing is more frustrating than getting a communication about something you’re not eligible for or which you already have. If, as a marketer, you can’t tell who in your database is going to be eligible, don’t do it.

Consistency is also important. A lot of our subscriber offers don’t just go out via email. We’ll also present those same offers on the site. The message there needs to be absolutely consistent – the same offers, the same messaging, the same look and feel. That’s very important. I’d cite Virgin Atlantic as an example of a company that does this very well across email, on their website – everywhere. Everything they do feels very seamless, rather than being the work of different teams working on different channels.

Interview by Jon Fortgang