HuffPost CEO on leveraging big data
The Huffington Post’s global audience has grown from 45-90m users in the last two years, according to its CEO, Jimmy Maymann. Speaking at the Reuters Institute Big Data for Media conference in London on 8 May, Maymann said that the company’s target to launch in 15 markets before the end of 2015 was achievable, given that they have already launched in 11 and are in talks to produce Indian and Arabic versions soon.
Maymann described how HuffPost uses big data in its newsrooms to leverage content and produce better results. Insight into where users are entering the site is key, he said, and is tricky for a news site who has been used to driving traffic through its homepage. “The homepage has always been the front door to a brand, but that’s changing,” he said. “People arrive on our site in different ways now, so we need to do a better job of analysing user journeys, creating better paths and entry doors and exposing our audiences to the stories and content that is most relevant to them.”
In the newsroom, said Maymann, HuffPost uses editorial dashboards to “empower editors to make smart decisions,” to see which stories are trending in real-time and do A/B testing of headlines to see how it affects the success of stories. “Content is optimised by data,” he said. “One headline we changed saw 2.6m social entries in 24 hours. Using data, we can definitely make much better and informed decisions on when we need to change things.”
HuffPost has gone from producing 200 to 1,600 stories every day – a development that has been accelerated by social media, which has “changed the pace of the news cycle,” according to Maymann. “If you want to play the game, you have to produce more content,” he said.
But how do they make money? “We share our learnings with advertisers and suggest the kinds of conversations they should be having with their audiences,” said Maymann. “We have a team of staff creating content for advertisers which needs to be interesting, engaging and something that doesn’t turn our readers away, otherwise that would be a bad decision for us. By shifting and creating this new model, we’ve established a new revenue stream, which will this year contribute 25 per cent of our overall revenue. It’s very meaningful.”
The HuffPost’s current challenge is “passive personalisation”, which involves users being shown content and advertising most relevant to them. “We can identify user interest levels to tell us what is driving engagement on our site, which helps us to make better decisions on where to spend our time,” he said. “We need to connect the right content to the right people. The technology we use builds user profiles, so while we don’t know your name, we know your behaviour. This approach is creating better user experiences, dwell time and engagement on the site.”
Asked whether the data gleaned is used to make every decision for the site, Maymann said: “Sometimes we overrule the data. It’s not only about traffic if you’re a news organisation. We still need to cover topics that may not be shared on social. There is a distinction between these types of stories, but it’s important to still do A/B testing and for advertisers to stay relevant to the audience. If you lose relevance, you won’t be here in 5-10 years’ time.”