Understanding the opportunity of the data revolution

Rob Bradley

No more so than at a company like CNN. We’ve built up a huge digital audience at scale over the years, but the value of that audience is diminished if we can’t understand it, respond to our users’ behaviour and harness the insight for our advertisers. Sounds a mighty task, right? Well, it would be downright impossible if we didn’t have data insight.

Take, for instance, how data works in understanding and targeting an audience. Is it better to start with a large audience data-set then narrow in and optimise targeting or better to start with a tight data-set and broaden if delivery is hard or it’s underperforming?

Balance is vital

Having worked with publisher data using several data management platforms for four years now, I’ve come to understand that this balance is vital to the success of a data-targeted campaign. Utilising the platform to create both broad and tight data-sets to optimise dynamically against is key. Specific segments should be fluid because one set audience won’t always behave the same for every campaign. By using the real-time data insight and reporting back to the client, the marketing message can be tailored for best success during the campaign or the next.

The bad news for smaller publishers is that only large scale allows you to start big and tighten in segments to meet performance goals. Without the scale in the first place, where do you go to optimise? Some Western European and US media owners can utilise third party data, but other regions will struggle to find meaningful volume of accurate data, or indeed, any data at all.

Scale can be bought in other ways, of course - huge audiences are available to any advertiser happy to be cast adrift in an ocean of inventory. Even I can put money into Facebook to boost a post on a hobby page if I wanted to target a group of users by demographic or interest anywhere in the world. However, what about context, environment and a site’s natural audience harnessed over years of providing quality content? This means something before data even enters the room.

Programmatic trading struggled for credibility in its formative years, but as premium publishers entered the arena programmatic became a means to reach upscale audiences. With premium publishers also comes reassurance around ad fraud - we like humans seeing our ads not bots - and marketers could take heart that even without the data to back up the results, they broadly knew what type of upmarket users their message would be reaching.

Positives outweigh negatives

We need to be wary of an over-reliance on data because a great creative solution in a relevant environment can succeed on its own, and we don’t want to risk creeping out a user by over-targeting them; but, overall, the positives far outweigh the negatives as long as we’re sensible. In fact, data-informed decisions about capping frequency of ads and re-targeting can enhance the user experience and make our audience less likely to want to install that pesky ad-blocker.

However, for me, the biggest opportunity for using data to its full potential is in the area of reporting. Not so long ago a tedious chore delegated to the most junior member of staff, data-rich reporting is enabling publishers to deliver true audience insight to clients. At a time when there’s so much competition for ad dollars, the onus is on the publisher to go beyond the click through rate.

There’s a whole range of complex metrics - viewability, engagement, dwell-time and audience behaviour before, during and after they visit our site - that we can use to tell an important story: what type of audience saw your ad, how did they react to it, what was their next step, and how we can better serve them in the future. Only then can we as an industry truly say that we’re not just relying on scale, and that we really understand the opportunity of the data revolution.