‘We Need To Talk About Geo-Data Quality’
For PlaceIQ’s Drew Breunig, the burning issue affecting the geo-marketing space is not “direct” vs. “indirect” location data. The first step: figuring out what data’s good and what’s bad.
The topics surrounding geo-marketing appear fairly simple: the ability to target or direct advertising messages to consumers based on the signals emitted by their digital devices — and, ultimately, understanding the patterns created by that activity. But once you get into the details, things tend to get murky and frustrating, or so says Drew Breunig, PlaceIQ’s VP of Strategy. We asked him to sort out the various problems that presently plague location-based advertising.
At a conference a few months ago, you talked about geo-marketing being touted as a great solution for marketers. However, you also mentioned that marketers still don’t see it as an honest, good solution because data quality is so unreliable. What’s your sense of the issues and discussion around location-based advertising?
Candidly, that’s a bunch of hand-wringing. It’s like coming out and saying, “Hey, 70 percent of sushi tastes awful.” We know you don’t go to sushi places that have dirt on the windows. You don’t go to places that look bad. You only go to the ones that have a good reputation. For [tech companies in the location space] to try to raise this specter and say, “Hey, everybody freak out because there’s a bunch of stuff you can’t control,” just creates a problem that doesn’t exist. And then to say, “Here we are, we’re the saviors, we’ve got the sole answer,” well, it’s nonsense.
What exactly is the hand-wringing over?
I love when the truth is simple. There are two types of location data. There’s directly-acquired location data, which is taken just from the operating system API; then there’s indirect, which is what happens when you don’t ask your consumer for the privilege to use their location, and you have to derive it using IP address and other things.
If you get direct location, it’s all the same. The direct location that ThinkNear sees, that NinthDecimal sees, that Verve sees, that every company in this space sees is all the same. In other words, if you ask the smartphone user if you can use their location, there’s nothing special there. You don’t pick if it’s GPS or WiFi or some other connection. The only reason they’re talking about it is because they have nothing else to add to the conversation.
So the debate about the value of location data is largely a false one. Is that correct?
I think that conversation’s been over and done for many years. It effectively ended when Apple reversed its stance on the ways you can use location data. In the past, the only way you could use location data legitimately within the API was if you were offering a map or some other location service. Now, you can use it even if you’re an advertiser. That was five years ago they made that switch. After that happened, there was no reason to worry about this stuff. It’s rather boring and it’s kind of frustrating that we have to keep talking about it, because it’s solved. It’s been solved for years.
So what should the industry be talking about? What’s the burning issue that does need to be discussed and addressed by location technology companies?
Being able to understand why someone is in a place is our focus. That’s something that we should be talking about. The cool thing about location is that if you can understand what it means when someone is in a place, you can understand who they are and what they do, what their behaviors are, and what the demographics are. Location is a thread that connects different bits of data, and the only way you can do that is if you build a platform that can do those comparisons at petabyte scale.
The topic of beacons and other methods of in-store targeting, such as magnetic positioning has been getting a great deal more attention from the mainstream advertising world lately. How have you and PlaceIQ been thinking about the challenges when it comes to indoor?
It’s something we keep our finger on, but there’s no scale yet when it comes to indoor marketing. The hardware’s not installed in the stores themselves. Because our resolution of the world is in tiles that make up a hundred meters [of where a connected device is], you can divide the entire US into a billion tiles. But if you’re only targeting a handful of tiles, you’re not going to have scale. That’s why you have to have those data points to tie it together.
Targeting one McDonald’s with one message isn’t worthwhile. But targeting every McDonald’s in the region is a great tactic. When you get down to the aisle level, we just aren’t there yet. Even if a retailer had that hardware installed in all their stores, would consumers actually use the apps? Would they even install them on their phones? That’s a thing most people forget about beacons; it’s just a piece of hardware. You need to have the apps deployed as well. If you don’t have the apps deployed, it doesn’t matter. If you install a beacon in the forest and no one has the apps… well, that’s exactly what it is.
Still, it’s fascinating technology but quite frankly, it gets a lot of flak. You can look at every article about beacons, or at least most of the big articles about beacons, and you can replace the word beacons with [electronic inventory tracking systems such as] “NFC” [Near Field Communication], “RFID” [Radio Frequency Identification], and it’s all the same story. We’ve been trying to come at this 20 different ways. But a beacon is a technology, not a platform with practical implementation yet.
We were also wondering about how the use of geo-fencing is evolving as a location-based advertising tool. Do you think there’s a good understanding of the right use cases for this kind of ability?
First off, let’s be clear about our terms. Geo-fencing is another word that is basically a technical term that gets confused for the practical usage it gets. A geo-fence could be considered one of our one-hundred-meter-by-one-hundred-meter tiles. For example, we know an overlap with a Wal-Mart, because we’ve drawn the boundary with the tiles that correspond to it. Or, it could be a five-mile geo-fence that people want to target because you’re in the range of something. I think it’s important to make that clarification.
Clarity is one issue with this technology. Is there anything particularly good or bad with the way geo-fences are deployed on behalf of retailers?
The problem with geo-fencing is that it limits your choices when things go wrong. Say you don’t have our platform. You don’t understand what’s in every tile across the US. The only things you understand is where your store locations are and where that consumer’s phone is.
When you start under-delivering to your one-kilometer geo-fence, what options do you have? You don’t know where anything else is. You only know where your stores are, and you’re already targeting them, so what can you do? You can make the geo-fence bigger. That’s a silly approach, because all you’re doing is guessing at that point.
There’s another common problem with the use of geo-fencing. One of the most popular tactics people use us for is the ability to serve a different message to people that shop at their store, versus people who shop at their competitor’s store. The problem with that is, most people and most retailers’ competitors tend to be located in roughly the same place, such as a same strip mall.
My favorite one is one of the reasons geo-fencing is so popular with auto clients is that auto dealerships are all on the same block. If you have a geo-fence larger than a couple hundred meters, you’re buying the entire block. If you say, “Oh, I can do competitive,” your geo-fences are all stacked on top of each other, and you have no idea why someone is actually there.
The more important reason to use geo-fences is as an organizational tool. You can’t be sure why someone’s there. If you put a mile radius around a Starbucks, are you going to assume that everyone in that mile radius is in the Starbucks? That’s being foolish. As an industry, we need to be clearer in defining what our tools are and how clients can get the most of out of them. It’ll still take some time to get everything sorted out until we’re at least talking about the same thing when we bring up geo-fences or location data.