One of the most intriguing chats I had was one I instigated with the CEO of PlaceIQ, having read somewhere about the company.
Duncan McCall’s start-up, based in beautiful Boulder, Colorado, ingests mountains of data about places, things and people, and makes sense of it all for the purposes of targeted advertising. The company is heavily focused on mobile right now (following the money), but it struck me that what PlaceIQ does is really well-suited to Digital OOH.
McCall thinks so, too.
Mobile may be where the ad spend dollars and buzz are going right now, but McCall says that medium is somewhat limited in its ability to effectively target ads. You can push ads by handset types and apps and get a sense of user profiles. But the medium can’t really take advantage of all the hyper-local data that’s available to target ads based on location and context. Phones can’t set tracking cookies, for example.
Location and context, along with immediacy, are the biggest selling propositions for Digital OOH, and particularly vertical and aggregated networks that push the notion of hyper-local targeted advertising. The phrase has been used to death, but the medium really is about the right message, time and place.
With the right available data, and tools that make sense of it, the targeting can be far more sophisticated than the very basic level of audience profile, demographic and general category targeting being done on most Digital OOH buys these days. Better, deeper data opens up the possibility (or probability) that a lot of Digital OOH inventory will eventually be bought just like online through demand-side and real-time bidding/buying platforms. Based on very granular information, campaign buys will start to look and feel like something done by commodity market day traders. Already, companies like rVue are running off this general approach.
McCall says his take, developed from a lot of conversations, is that the Digital OOH sector is not yet at the scale, sophistication or acceptance level to use the data and tools his company markets. But he also thinks it will be.
“We’ve built this big system to essentially ingest, normalize, organize and make sense of all this location-based information that’s available,” he says, “and then we spit it out in this structured understanding of locations.”
The PlaceIQ platform takes piles and piles of public and proprietary information that, as isolated databases, have limited value. They are meaningful, however, when they get structure and available to query in endless ways. “We essentially transform time and location in context,” explains McCall.
The data getting steadily gobbled and chewed by the PlaceIQ platform includes the predictable – like government census and crime data – but also diverse information like mobile handset user patterns, the camera metadata on photo-sharing sites, event listings, retail sales reports, foot and vehicle traffic patterns, and social media check-ins.
The platform gets all the way down to analyzing what people are doing, where and when, and then segments that data into 100 metre by 100 metre “tiles” that define that area’s hyper-local profile data, hours by hour and day to day. For example, a downtown area may show a very strong affinity for business people during the day but completely transform its profile on evenings and weekends when the office lights go off.
So far, the company has profiled the 40 largest US markets down to the 100-meter tile level, covering about half the US population.
McCall says the licensing or fee model for Digital OOH networks is still being thought-through. But it’s more evident how the information would be applied. Tapping into the data would help both sellers and buyers, providing more context to the audiences and making recommendations of what time to buy, when and where. “We can recommend the types of ads to serve,” says McCall.
Right now, the Digital OOH medium is mostly about selling massive numbers of impressions. Network operators are probably not going to be all that excited about pushing a platform that more narrowly and efficiently targets buys and, therefore, reduces the overall value of the booked campaign.
McCall says he already saw that reticence from mobile ad network companies. That was an industry that had little interest in moving off mass impressions, for just that fear of eroding revenues, but has now been converted. “It’s pretty interesting how quickly that mindset changes.”
The debate is really just starting about the use of technology and data mining to make advertising far more addressable. There are arguments that better targeting increases campaign value, but there are also arguments that addressable is not even what some brands want, suggesting they just want reach.
But as a selling proposition the ability to mine data and really understand the audience and the context when planning ads – in some largely automated and dynamic way – seems pretty interesting.
Dave Haynes is the founder and editor of Sixteen:Nine, an online publication that has followed the digital signage industry for some 14 years. Dave does strategic advisory consulting work for many end-users and vendors, and also writes for many of them. He’s based near Halifax, Nova Scotia, on Canada’s east coast.