TechBytes with Peter Lenz, Senior Geospatial Analyst, Dstillery

Peter Lenz Dstillery
Peter Lenz Dstillery

Peter Lenz
Senior Geospatial Analyst, Dstillery

Three things that rule real-time marketing today — location, location, and location. Geodata-based marketing tells you where to pitch your tent. Recently, Dstillery unveiled their Dscover Maps to serve the marketing and advertising industry better location data. We spoke to Dstillery’s Senior Geospatial Analyst, Peter Lenz, to understand how Geospatial Marketing Technology works and the impact it has on ROI.

Tell us about your role at Dstillery and the team and technology you handle.

Dstillery is an applied data science company that serves the marketing and advertising industry. We build high-quality audiences that are highly conformant using about 160 billion data points every day.

As the Senior Geospatial Analyst, my focus is on location data. Of those 160 billion data points, a significant amount contain location data. Sometimes it’s really seemingly accurate, sometimes it’s just a state or a zip code, or sometimes it just tells us “this is in the United States.” We take that information, clean it, and then we add it into our system. We then take this location information, turn it into something our machining learning algorithms actually understand and add it to our data so a visit to a physical Walmart is exactly the same as visiting Walmart.com.

How does Dscover Maps benefit marketing teams?

A lot of marketers still don’t know what they’re doing with geomarketing. There are some out there that are very sophisticated and understand what they’re doing. But there are certainly a lot out there that think adding location data in and of itself is magical, and that is plainly not true.

Dscover Maps makes it a lot easier for marketers to understand this geodata. Dscover Maps is a Software as a Service geospatial insights tool that provides a detailed view into market composition by geography or location.

If you’re a writer, you’re always told, “show, don’t tell.” Well, as a data scientist, Dscover Maps allows me to finally show, not tell someone why one location is a good or bad idea, and where they should be targeting their media efforts. It’s a map of The United States, with various different geographies built into it. It knows by default what kind of consumers live in zip codes and defined market areas, and can handle all sorts of custom segmentation for individual brands.

How does Dscover Maps directly impact advertising ROI?

It’s really impactful. A lot of times when a client comes to us with a geo request, it’s usually a desire to reach everybody that lives with a few miles of their store, or a competitor’s store. That’s a very naive way of thinking about how customers behave in the real world.

Here’s an example: Early on in the process of developing our geotechnology, a quick-service restaurant (QSR) client came to Dstillery with a problem. They had a location in downtown Baltimore that was not doing particularly well, and the client wanted to understand why.

QSRs are a great place to learn from geodata because people have to wait for their food, and any time anybody has to wait for something, out come the phones, and data flows into our systems. So we added this restaurant as a point of interest in the Dscover Maps database to look at all the devices that physically show up at this particular location.

The client strategy at the time was to send physical coupons to people within a few miles of this particular restaurant to drive them into the location. When we looked at the devices that were showing up, we discovered that a small number of them were actually from the area that was receiving coupons. Most of the people who were eating at this restaurant were well outside that range. They were actually suburbanites who were grabbing a bite to eat at lunch or after work.

We built a model that lets us take an informed guess about a rough area where a customer lives, based on device activity. The client then shifted the coupon strategy to target these specific locations that we had identified as likely hometowns of the people eating at their restaurant. As a result, they saw more business at the location, because they had a better understanding of what their customers were doing in the real world.

What is audience intelligence in the context of modern advertising?

Intelligence, in the frame of advertising, is this idea that we are getting better at understanding our clients and guessing less and less. Once upon a time, companies had absolutely no idea who their customers were. For example, consider the Edsel, which Ford developed in the late 1950s. At the time, Ford was one of the world’s largest corporations, one of the most intelligent corporations in the world, and the way they designed an automobile was that they guessed. They sat down a bunch of executives in a room and they guessed about who their customers were and what those customers wanted. As a result, the Edsel was one of the most epic flops in business history.

Nowadays, marketers don’t just guess at things. They can run surveys to understand what consumers say they want. Dstillery is an intelligence company, which is the next step beyond surveys. So, rather than going out and asking people thousands of questions, we build probabilistic models to predict what devices are interested in — what they’re thinking, what they’re doing, and what behaviors they have. That is a big shift. Every behavioral model in the system can be thought of as a question or survey. It’s like we’re asking everyone in The United States, everyday, “are you a candy lover?” “Are you a backyard chicken farmer?” “Are you a Ford driver?”

How do you organize your analytics stack for Audience Data, Customer Data and Intent Data? How are these streams different from each other?

Audience, customer and intent aren’t different from each other. They’re the same data. We look at the world as devices. We don’t look in terms of what a customer’s doing, we look at what anonymous devices are doing, and then we aggregate those devices. The only thing we get about these devices is their behaviors. Behaviors are intent and intent is behaviors. That’s how we look at the world.

Machine learning needs lots and lots and lots of data to be very good. To make these insights actionable, we have to make all these things look the same. So instead of trying to connect all three data streams together, like a traditional marketing company would, we have a holistic view of the world. All of that data is one type of data.

What are your predictions on Data-as-a-Service for 2018-2020?

Data-as-a-Service is the future. As I said before, audience intelligence means marketers are getting smarter and smarter. More companies are realizing that their current way of guessing is ineffective. A lot of it is based on gut or superstition, and marketers want more data-based solutions.

To make data-based decisions, you need data. We’re seeing more and more interest in putting our data in other people’s systems to help them make decisions. That is the future, and Dscover Maps is a step in that direction.

How should businesses unlock the value of their First-party and Third-party data? What makes Dstillery a go-to platform for such campaigns?

Dstillery started out in the advertising world by placing pixels on websites to build performance and high-quality audiences based off of that first-party data. Now we’ve built an analytics platform, in Dscover Maps, that can take all the audiences in our system, including those first-party audiences, and project them into the real world so that marketers can see where the people in their first-party or third-party audiences live in the real world.

The way to unlock this kind of data is by letting it play with other types of data. A lot of the data that’s out there is sitting in a system and not talking to anything else. It doesn’t have any contact with other kinds of data.

When marketers join their data with Dstillery’s data, it means we can take all of the context and other information we know about the world and connect it to the marketer’s data. Data by itself is lonely, it needs to have friends. We have the friends.

Thanks for chatting with us, Peter.

Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at news@martechseries.com

Previous ArticleNext Article

Leave a Reply

Your email address will not be published. Required fields are marked *