The folks at Facebook will be talking about what it calls ‘people-based advertising’ a lot as the social media giant goes about pocketing more digital advertising dollars.
- Take our Which-50 reader survey and go into the draw for a chance to win a pair of Sony H.ear Bluetooth headphones or 1 of 5 Google Daydream VR headsets
People-based advertising, tracking an anonymous individual versus a cookie, has a big impact on what attribution figures look like across different channels and devices, according to research conducted by Facebook and Datalicious.
Steve Lockwood, head of marketing science ANZ, Facebook outlined the measurement challenges for marketers during a presentation at the Ad:Tech conference in Sydney this week.
“We know there is a disconnected across different channels in a lot of the ways that we measure advertising. There’s no common denominator currently which makes a truly comparable metric of one channel against the other in generating a return,” Lockwood said.
And he is placing the blame on cookies.
“The vast majority of digital advertising decisions are still based around cookies which is obviously a problem. There’s going to be wastage and there’s going to be inefficiencies,” Lockwood said.
“When it came to advertising — whether it’s planning, execution or measurement — the whole internet infrastructure was built around cookies.”
Cookie measurement was a perfectly fine ID when people only had one device and didn’t know they could delete cookies, Lockwood said.
“Now people have multiple devices. Even if you’ve only got two, a laptop and a mobile phone, that’s different cookies. When you are using a cookie-based methodology it might look like you’ve reached two people once but you’re actually reaching one person twice.”
Lockwood said Facebook’s own research showed reach is overstated by 58 per cent on average.
If relying on cookies is bad then using last click attribution modelling isn’t much better.
Giving 100 per cent of the credit for a conversion to the final touch point leads to misattributing campaign revenue, Lockwood said.
“We’ve done our own internal analysis where we’ve seen, on average 22 per cent of incremental revenue is misattributed due to last click. Where mobile spend is high, more than 70 per cent of impressions in the campaign, as much as 54 per cent of the revenue is actually misattributed.”
To test what impact people-based measurement has on campaign results, Facebook allowed Datalicious to take a look at its data.
Datalicious used Facebook log in information to collect ad impressions of three ad campaigns conducted in Australia during October to December, 2016. The purpose of the study was to examine data from the campaigns at an individual level, rather than a cookie level, across multiple devices.
During the Ad:Tech session Christian Bartens, CEO & co-founder of Datalicious, presented a handful of key findings from their research with Facebook.
“What we found was every person, on average, has three different cookies associated with them,” Bartens said. “Two out of those three cookies will get no credit for a conversion. That’s a big, big problem because that tends to disadvantage certain channels more than others.”
Mobile tends to perform badly in a last click, cookie scenario because it is rarely the device consumers use to make a purchase. According to Bartens mobile plays a more influential role earlier in the consumer journey moving them from initial disinterest to awareness, and from awareness to intention.
The problem is only going to get worse as mobile penetration rises. More smartphones, mean more cookies.
“This is an issue that is becoming worse and we need to address it sooner rather than later because the proliferation of mobile devices is not going away, it’s getting worse,” Bartens said.
Taken as an average from the Datalicious study, using a people-based attribution method means 1.8 times more conversions should be attributed to mobile campaigns than they currently are.
Fusing desktop and mobile together under one identity also makes the customer journey 38 per cent longer. Under this scenario, Facebook and display can claim more credit for contribution to the final conversion. Facebook’s gains come at the expense of organic search and direct.
“Most marketers in the room are probably undervaluing how much Facebook and Instagram advertising actually is contributing to their sales numbers,” Bartens said.