The Direct Marketing Association will introduce a primer of sorts at Advertising Week that's designed to help marketers and tech companies speak the same language on cross-device measurement.
Marketers are increasingly trying to understand whether and when they've reached the same consumer on multiple devices, hoping to calibrate the frequency and content of their ad messages appropriately. But retaining vendors to help with that is complicated by the differing and evolving jargon in the field.
The DMA's document will define terms and suggest questions that marketers might include in requests for information from vendors.
Providing more information on cross-device marketing and developing a set of standards was the top request from DMA members, according to the association.
"Nobody is totally clear on what is really capable of being done today," said Morgan Digital Ventures CEO David Kohl, who spearheaded the initiative for the DMA. "Things like how well does this work, where are the gaps? There's also a lot of confusion in the language where expectations are missed because buyers asked for one thing and sellers give them something completely different."
The sprawl in terminology and confusion seems to reflect the fact that people are using more connected devices. Adobe says the average consumer currently uses six devices each day. These can be work PCs, personal laptops or tablets, smartphones, TVs, wearables, connected cars -- the list can go on and on.
So it's no surprise that a slew of companies have propped up offering a solution for marketers to tackle this problem. Thing is, the technology is still young and the language has become so discombobulated that what one company tells a buyer can mean a completely different thing at another company offering the same solution.
The DMA saw this gap and says it wants to close it. "If someone can ask questions that are clear and specific and don't leave room for interpretation, but can only provide a singular answer using language that everyone understands, then the remainder of the gap should be dramatically reduced," Mr. Kohl said.
Although many companies provide cross device solutions, many also call their technology something completely different. Tapad, for example, trademarked the term "Device Graph" and as such, calls it that. Other companies, like Oracle Data Cloud, use "identity graphs." There's also an acronym: XDID.
Jargon aside, there are two important terms that remain constant in the cross device space: deterministic and probabilistic.
Deterministic processes work out who a person is (anonymously) using things like social login IDs, credit card transactions, reward cards and utility billing. Probabilistic techniques aim to achieve similar results by looking at things like location and browsing history. Some companies, such as Google or Facebook, offer completely deterministic data. Others, like Tapad, focus on probabilistic. Then there are those such as Oracle Data Cloud, which argue that a blend of the two provides the most accurate picture.
The DMA's RFI guide will offer marketers a more comprehensive understanding of the two approaches, among other things.
"We are hoping sellers and buyers get more aligned on how to use cross device, the right questions to ask when buyers are looking for a provider and to eliminate some of the friction between buyers and sellers," said Kurt Hawks, senior VP for cross device and video at Conversant.
Another problem facing the cross-device field today is accuracy.
There is no authoritative third-party source to verify the accuracy of information that data providers come up with. "Who do you trust hasn't been established in the same way that it has in TV and radio," said Philipp Tsipman, VP of product at MediaMath. "You have a whole set of companies using probabilistic algorithms and the best you can do with that is 60% or 40% or 20% accuracy. Some companies are going out and saying they are 90% accurate."
Companies in the business are mixed on whether an authoritative source is needed to verify cross-device solutions, questioning among other things who should be trusted with the data necessary to do the work.
Then there are those who say the field is far too young to even be consider bringing on a third party to verify information.
"Everyone is looking at each other, saying, 'You have a data set, I have a data set, are you going to trust me to identify it?'" Mr. Tsipman said. "It is going to be a while, but it is so needed."
In the meantime, the DMA is releasing its RFI template to first get everyone on the same page.