How New Location Data Tools Are Making Attribution a Reality
The global advertising industry is a $500B-per-year market spanning mobile, OOH, television and countless other mediums and verticals. And while the underlying technology differs across these mediums, the goals of advertisers do not. The core objective of advertising is to influence a consumer to purchase a product. Whether it’s a bar of soap, a car, or a hamburger is irrelevant — in the end it is always about a purchase.
What is lost in the frenzy of cyber Monday statistics and “free shipping” is the fact that in North America 90% of products are still purchased in a physical location. The problem up until now, regardless of advertising tactic or technology, is that it has not been possible to determine definitively if any ad message resulted in driving a consumer to an actual store. Not one type of advertising can deliver this proof. Sure, many CMOs and Brand Managers naturally gravitate to the use of coupons and loyalty programs as a means to measure proof of ROI, but the results of these tactics represent but a fraction of total purchases — and even if accepted as “proof” — the path of influence is still a black box.
But the world of marketing technology evolves quickly, and in a single year the relationship between retail, the consumer, and the ad industry has already changed.
First key factor: In the U.S. today over 75% of consumers own a smartphone, up from 35% in 2011. Among 18-to-29-year-olds the penetration is 92% (more baffling to me are the 8% who don’t have a phone).
Second key factor: We are undergoing a wholesale acceleration by the advertising agencies and brands away away from insertion order-based advertising to programmatic buying. As an example of this acceleration, according to eMarketer, continued investment in programmatic advertising technology and transactions will see programmatic’s share of total US digital video ad spending rise to 82.0% by 2018.
Conclusion: The coupling of the adoption of mobile handsets with this shift to automated buying has resulted in the creation of a scalable in-store attribution platform the likes of which has never been available to marketers before.
How so? First, today’s handsets are not the voice driven flip phones we used 10 years ago — they are now geo-aware supercomputers that people keep within arms length no matter where they are — and are increasingly connected to more and more objects and things.
Second, the buying of advertising is not advertising at all — and hasn’t been for quite some time. In today’s media world we buy “audience,” which is really another way of saying we are buying data. The advertisement or banner ad is just an action layer that rests on top of a data decision. The combination of this geo-aware computer in your pocket and this data-driven advertising model is what now enables in store attribution to be truly measured.
Being able to pull location data out of a mobile applications bid-request has provided a slew of location-focused tech firms with the ability to infer in-store attribution. The reality is that it’s really just a guess (no matter what they tell you). Fraud is rampant in the real-time advertising space and naturally as a consequence affects the accuracy of the location data that is being presented — up to 70% of the time. If we want to be romantic and assume that this location information is in fact legitimate it still does not solve the challenge of determining if Consumer A actually went to a specific McDonald’s and not the cafe next door to it. The reality is this methodology can not prove this action. Period. Sorry, no “tile” “blueprint” or polygon does this. Without the installation of some actual hardware infrastructure inside a store it is impossible to deterministically state that Consumer A has been seen in a McDonald’s or any other store.
While you digest this there are a few important aspects of location marketing that brands need to be thinking about in 2017 and beyond:
1. 1st party data, 2nd party data or 3rd?
Any platform that relies on anything less than 1st party data cannot promise true in-store attribution and can never do real-time data. It’s that simple. Any platform that relies on third party data (the most common form being that which comes from the bid-stream) as their methodology should be categorized as a probabilistic attribution solution — and evaluated and priced as such.
2. How does the data get tied to a specific store?
Similar to the source of the data, how a company ties the advertisement to an actual store is the next important question. Beacons, sensors, Wi-Fi, etc. — these will all allow for the closing the loop. If the platform/vendor does not have these themselves, there is no validation and their results are estimates.
3.What is the privacy of the data?
If the data is not opted in, you are in a grey area. Nobody really wants to talk about this, as it is the foundation of almost every ad company in existence and it is easier to pretend this isn’t an issue. Privacy is the third rail of advertising and those who sell profiles and solutions without the consent of the user are exposing themselves to enormous risk (see Vizio) are asking for trouble in the months to come. Companies who have thoughtfully worked out a clear privacy policy are likely more compliant across the board. These companies should be treated as premium.
The future of retail and attribution is evolving quickly and allowing brands for the first time to have a better understanding of how effective their advertising is. While the search for in-store attribution is at the top of the every marketer’s wish list it’s important that all know the strengths and weaknesses of each methodology. Now that we have made great strides with viewability and fraud detection – in-store attribution will quickly emerge as the next third party metric for validation. Why? Because if you can prove the ad is real (fraud detection) and seen (viewable) the next step is to prove it is effective (in-store attribution). Maybe then we can all agree to move on from “click-through-rate” entirely.
Neil Sweeney is the founder and CEO of Freckle IoT, a first party data company that solves in store attribution for brands and agencies.