Perils and Promises of Proximity Marketing
Imagine you are strolling hand in hand in the park, having just enjoyed a wonderful evening with that special person in your life. On a nearby digital screen, this ad pops up:
Targeted real-world advertising is here.
The example above is really too simplistic — too easy for me to make. When we see these Minority Report futuristic nightmares, the fact that someone might learn your most embarrassing bedroom secrets, laying waste to romantic and familial relationships, is alarming, but far from the only threat. Private behavioral information can wreak havoc when used to target advertising in the real-world.
When real-world advertising arrives in full force, triggered by data that knows your browser history, and may even reveal your identity, the ramifications are profound. Imagine getting personalized offers for loans, on a gas station pump display that shares seemingly innocuous details about your spending habits, average daily balances and credit score range. Or consider offers for diet plans with your track record of battling the bulge; engagement ring branded campaigns that spoil the big reveal; ads for pill addiction based on your recent searches or even personalized LinkedIn recommended jobs popping up in response to your upcoming career change — in the presence of your coworkers.
Avoiding illegal and immoral activities can’t be our only strategy to deter us from uncomfortable or dangerous identity-specific proximity marketing situations. Keeping to a moral high-ground doesn’t mean you won’t be adversely affected.
Relationship troubles and embarrassment might be the least of our worries.
Creepy and Dangerous
Many of us already agree that cookies are creepy. The number of people who either actively block cookies or clear them on a regular basis, stays at the 50% mark. This online record of your past searches and browsing habits make relevant advertising possible, but they know too much — comprehensive, individualized data can be used to identify you personally.
The First Days of Proximity Advertising Are Upon Us
Proximity advertising, targeting out-of-home (OOH) ads, is already here. In fact, programmatic proximity advertising campaigns are already here too — and when programmatic takes a hold, you’ll see the larger networks really dig in. Programmatic campaigns bring scale, and they’re already dipping their toes in.
This type of advertising isn’t necessarily a bad thing, it can be very useful. Triggering relevant OOH ads is a win/win for advertisers, end users and publishers/developers. The user sees ads when they aren’t distracted, or that don’t clutter their screen, while giving developers and publishers a stronger monetization model (monetize without showing ads in the app).
There’s great potential here, but the same potential for good also has heavy implications for privacy. In the wrong hands, this type of advertising is creepy at best, dangerous at worst.
Analysis of an Aggregate is Adequate
Proximity advertising triggering OOH ads isn’t necessarily a bad thing, but personalization is — and it’s not even necessary. There is plenty of data that is useful for targeting, that isn’t an identity signal. Yes, it’s easier to take PII, like browser history, search history and cookies, perhaps combined with an advertiser’s first party data, to generate a highly targeted, personalized ad, but it’s overkill.
Advertising doesn’t require individual behavior history to be relevant — an aggregate of behavior is just as useful, and alleviates privacy concerns. If we can determine the demographic, we can trigger based on an aggregate of the behavior for that demo. For instance, if I am aware that the user is a 40–44 year old man, within a certain income category, living in NYC, with kids and if I can also associate some of his hobbies as that of an outdoorsman, I don’t need to know which sites he visited or who he banks with. I can activate advertising that is relevant, but not creepy, and I’ll get a better response. And this scenario is actually in the advertiser’s best interest, because an aggregate of that data, made up of numerous anonymous profiles, can help them better understand the bigger picture, how this group behaves, and how better to target the overall group, not just one person. This is proximity advertising that scales.
Who do you want to bring this to market?
The question is not whether or not this Minority Report future will arrive, it’s not even really a matter of when. The important question you should be asking is who do you want to bring this to market? Historically, companies like Google, Facebook and even Apple will buck public opinion because of our reliance on and proliferation of their products. If we don’t like the way they’re doing something, they just come up with a cute robot to sell it to us, or employ a tear jerking TV spot to distract us from the issue. When our back is turned, they go right back to doing whatever it was that they were doing before — problem solved. They make products that are so amazing we’re willing to swallow creepy practices to keep using them — but at what cost?
The ease with which these companies could create a world where the opening image of this article is commonplace boggles the mind. It’s coming because it’s easy. But what’s easy for them is not safe for us. There is a better way that can meet profitability goals while respecting the privacy of the users it’s built around and, ultimately, for. Viewing behavioral data points as pieces of an aggregate and “forgetting” what those behaviors were, gets both sides what they want.
To use the example above, being an Ashley Madison subscriber only indicates that there’s 99.9% likelihood that the user is male. Even better, let’s not even use web browsing as an indicator. We can understand what the most effective message to trigger would be based on the behavior of the aggregate. We can also algorithmically tie places we physically visit to categories of interest. While shopping for a car, location can already offer us advertising that moves the sale forward, and offers benefit to the customer, by revealing local sales and the best financing option. Physical location data, along with aggregate data, can signal a business traveler with an expense account, triggering offers that appeal to the demographic and the location.
The larger question isn’t whether or not behavioral targeting is a good idea, because not only is it good for users, it’s incredibly useful for advertisers. The focus has to be on making sure that when this arrives in full force, that the gatekeepers are benign from a privacy standpoint. The right types of companies to manage this are those who respect privacy, and have built their businesses with end user protections in mind. An aggregate of data is adequate, and an aggregate of aggregates adds even another degree of separation between the user and the behavioral data. The right way, is to, whenever possible, view the aggregated behavioral profiles as part of a larger set of profiles to determine the skew of a physically present audience. Again, forgetting the parts that made up the whole. Anonymizing any potential PII that is captured has to be the standard, and a desire to capture none at all is ideal. We shouldn’t have to give away our private information to use a phone, a browser or an app.
We’ve learned what happens when we’re irresponsible with advertising, we must make sure we get this right or one of the only ways to monetize this amazing channel, mobile, will be rejected and the implications of that are terrible. Developers deserve to be paid for their efforts, and the right advertising at the right time is a good thing for everyone involved. However, you don’t need to personalize this medium to monetize it — that would be the same mistake as trying to do desktop on mobile.
The big companies are not good stewards of our privacy, and it falls on the end user, and talented developers, to counter the ramifications of overreaching advertising practices. But managing this after it’s mainstream is going about it the wrong way. Let’s get in front of this before it’s fully upon us. This is going to happen, and it will be months, not years, before targeted behavioral advertising is mainstream — let’s just make sure that individualization is off the table.