As consumers’ use of digital platforms shows no sign of slowing, understanding the effects of online advertising across different devices is integral to any digital campaign. PwC predicts that over 50% of consumers will own an iPad, tablet or eReader by 2018. Unsurprisingly, the latest report from eMarketer forecasts a 38% rise in online advertising for 2016. As digital continues to dominate media planning, how are the latest technologies working to solve the issue of attribution and is the industry one step closer to achieving accurate cross-device measurement?

Technologies used to track attribution are changing

Attribution has been a coveted territory of all advertisers for years and not just those in retail. The problem facing the ad industry is the current technology to track the effectiveness of online ads across multiple devices is yet to be perfected. Instead, technology is limited to individual media owners and marketing companies.

It’s now widely acknowledged that simply using cookies is not enough to achieve accurate cross-device tracking. As a result we are seeing an influx of media owners and marketers rushing to offer a solution that enables conversions to be measured more accurately across different platforms.

A player that has shifted the game immensely is Facebook Atlas. At AdWeek, Atlas announced that Facebook IDs enable true multi device tracking and attribution. Atlas claims to now enable brands to track consumers as they move from mobile to web. All through the user’s Facebook profile. It also promises to deliver ‘people-based’ marketing. By using a Facebook ID, the platform can show it’s a real person viewing and acting upon an ad, unlike cookie-based tracking.

When used in conjunction with Facebook Audience Network, it’s also possible to link conversions with consumer demographics and behaviour to better tailor ads and arguably improve conversion rates. The benefits of Atlas though, only work if you want to stay in Facebook’s world.

Adapting our approach to tracking customers across devices

We are working closely with clients to prove the effect of online to offline, or ROPO: ‘research online, purchase offline’. Using accurate mobile geo location data, available through partners like Weve, allows us to test hypotheses in this area. Linking consumer movements to in-store, universal ID tracking cross-device and then stitching on more traditional methods of retail client data such as site sign-ins and store card data can answer the question.

An example of this in action is a campaign iProspect worked on with IKEA and Facebook. IKEA wanted to identify whether it could use Facebook to acquire new customers and connect with existing customers to increase sales. To find this out, iProspect worked with EE to analyse device activity for mobile numbers within a geo-fenced area around the Cardiff IKEA store. Facebook ads were then targeted at customers visiting the store to establish whether the ads have an effect.

By measuring in this way, we were able to demonstrate that Facebook ads drive a 31% uplift in store visits by 22-25 year olds and 11% increase in visits from 26-35 year olds. Such campaigns show that it is possible to prove the effect of online advertising, but only in certain scenarios where it’s possible to target ads without interference from pedestrians nearby passing by the store.

Where next for attribution?

The next steps in attribution will be for large players to follow Facebook’s lead in the attribution field and the engineers at Sizmek and Flashtalking to be working furiously to be able to connect those data walls that the large global players are re-erecting. It won’t be long before Google, eBay, Amazon and others follow suit, with the only surprise being that Google didn’t do this first, leveraging their large data stack and universal IDs.

Only once we are able to track, analyse and understand the data in a digital world can we start to fully understand and attribute the channel – then it’s just a case of linking this to offline retail purchasing. And while there have been great leaps towards solving the problem of cross-device measurement the race is still on to put the issues of attribution firmly in the past.