Measuring the success of digital media advertising in a fractured media landscape

According to a study by Aviva, in 2020 the average UK household had 10.3 internet-connected devices. This amounted to a 26% increase on a similar study in 2017, which revealed UK homes had an average of 8.2 connected devices. Based on an average UK household of 2.4 people, this works out at around 4.3 connected devices per person, many of which will be devices capable of distributing media such as smartphones, laptops, desktops, connected TVs, connected radios, tablets and e-readers.

What challenges does multi-device use create for marketers?

This presents some challenges for digital marketers trying to understand the impact their display, audio and video ad campaigns are having on website traffic, let alone low-funnel conversions like sales or leads. After all, despite what you might have heard, most people do not spend all their time on their phone, and many people’s internet browsing habits have now evolved to the point where they think nothing of switching between devices for different tasks.

It’s easy to imagine a scenario where someone sees an ad on their connected TV, researches the advertised brand on their phone and clicks a search ad, then navigates directly to the site on their laptop to complete a purchase. This kind of behaviour is common and may even be something you yourself have done. But from an attributional perspective it is highly complicated: here is a customer journey with three touchpoints spread across three different devices, with no direct way of connecting the ad clicks and impressions to each other, or the final purchase.

In website analytics platforms like GA4, this kind of cross-device behavior can manifest as a general underestimation of the impact of prospecting activity on conversions, and an overestimation of the impact of lower-funnel activity such as search ads. In the example above, for instance, the purchase would probably never have happened if the person had not been served an ad on their connected TV to begin with. But GA4 would have no idea that CTV impression had even happened. At best, GA4 would give all the credit for the conversion to the paid search ad the user clicked on their phone. At worst, GA4 would assume the conversion on the user’s laptop was from a direct website visit, and not attribute the conversion to any advertising activity at all.

These kinds of reporting biases aren’t just minor irritants for pedantic ad ops managers like me. Marketers of all kinds have grown used to using data to support tactical decisions, and for many advertisers the reports generated by analytics platforms inform where budget is allocated. This means that any measurement bias which is not accounted for could lead to a cycle of worse-than-expected marketing performance.

What might marketers be missing if they disregard these tracking difficulties?

Despite the fact it is often undervalued by website analytics, when it’s done right cross-channel prospecting and brand awareness advertising can have a big impact on conversions. In 2017, 31% of sales in the US involved touchpoints across multiple channels, and that figure is likely to have increased since. For brick-and-mortar retailers cross-channel advertising is even more important, with omni-channel shoppers averaging a 30% greater lifetime value than single-channel.

Market-making media advertising may not drive conversions with the same immediacy of paid search ads, but over the longer term it assists in brand recall and customer retention and leads to higher volumes of website traffic and in-store footfall which, with an optimised sales journey, can lead to more conversions. Put simply, there is a reason Coca Cola pays for TV ads in premium slots over the Christmas period, and it’s not because everyone who sees the ad is immediately rushing out into the wintery night to grab an ice-cold cola for their stocking. As a company, there is no way for Coca Cola to deterministically link impressions of these TV ads to specific sales, but the overall benefits to their brand and bottom line are enough to justify the expense.

How can marketers overcome the cross-device tracking difficulties?

Google provides a couple of solutions for marketers looking to more accurately attribute cross-device conversion journeys. GA4 has a data-driven attribution model and provides a ‘conversion paths’ report based on the gclid and utm parameter data of website visitors. For advertisers who are ad serving, Campaign Manager 360 offers cross-device conversion reports. But both GA4 and CM360 are reliant on Google’s statistical attribution modelling, which itself is dependent on logged-in Google account data. When it comes to something like connected TV or radio, or even mobile display advertising, marketers should be asking themselves how many people are likely to be logged-in to a Google account on those kinds of devices, and how this might be biasing Google’s modelling in favour of search ads?

As things stand, the best way of accurately attributing cross-device conversional data is almost certainly to develop a bespoke model that takes data points from multiple sources and combines them to produce a more nuanced picture of the influence of different channels on conversions over a long period. If this sounds like a daunting task, that’s probably because it is one! Luckily equimedia have a wealth of experience and expertise in this area, so if you would like to discover more about cross-channel media advertising and attribution, reach out to us now to find out how we can help.