McDonald's sets sights on DMP as it plots two-year roadmap to get its data working harder

The move was set in motion with a restructure of how it prioritises media spend in May this year to focus on five key markets outside of the US, namely the UK, France, Germany, Australia and Canada – which combined make up 40 per cent of its global media plan.

McDonald’s European media director, John Thekanady, who is leading the charge out of Europe, told The Drum that it has already made some strides in its approach to automated media trading. Today, up to 50 per cent of its display ads now bought programmatically, but moving forward it’s looking to get more targeted.

To do this it's gathering more third party data in a bid to get a clearer offline view of its customers so that it can better map that to online behaviours. This is allowing it to go beyond targeting people based on demographic data such as age and gender, by further segmenting these audiences based on what other stores customers shop at, how much TV they watch, plus the genres of this programming.

“It’s general but more in depth than we’ve ever had,” said Thekandy, describing this as the “intermediary stage” in its two-year plan.

The advanced stage is a DMP but before it can enjoy the benefits of having more ownership of audience insights, Thekandy said that the brand needs to further sharpen its first party data, describing the third party tools being used as making it “very disparate and siloed.”

One of the biggest barriers is that its web platforms differ from region to region – for example McDonald's in the US is completely separate to McDonald's in the UK, and both generate two entirely different data sets that it is struggling to combine.

According to Thekandy, rather than start from scratch, McDonald's is in the midst of employing a single platform provider that different markets can use as a global website. From there, it can plug in an analytics package with the aim of producing consistent first party data. It will look to further expedite this with its own CRM programme in the UK.

“Once we’ve got that first party data we will be able to combine that with third party data we’re already collecting to create a holistic view of the customer, wherever they are on the internet, that’s the goal,” said Thekandy. “But we’re not going to get there until 2017 in most places.”

McDonald's will soon find itself facing the same challenge as other marketers – measurement. The ultimate measure of success is sales in a restaurant – in other words, offline. It has been testing various click-and-collect order services for the last 18 months, which it has said are helping it build the consumer profiles that match the depth of those held by Amazon. But as it stands, France is the only country currently able to partially link online and offline activity through the app, and so for the most part soft metrics such as a click, view or some other form of interaction are the key KPIs.

Thekandy conceded that this means McDonald's is still grappling with which online marketing channel is most valuable and currently relies on econometric modelling to understand what part of its media spend is generating the greatest return.

However, it is taking this a step further in Denmark where it will soon inject econometrics data into the programmatic buy. This should see it, for example, be able to injest the sales data for a particular product at any given time to decide what product or creative execution should be prioritised in an online ads. “It won’t be real-time but as close as we can be,” he explained.

Thakandy is also working on a project to help it better define this customer end point. It will use a panel of people to link those that bought in a restaurant to their online behaviours and how they engage with McDonald’s offline touchpoints. From there, it will try to pick out key trends and assign ads to the most relevant screen.