The shift toward programmatic buying has mixed implications for local. On the one hand, the ability to buy inventory algorithmically provides marketers with unprecedented reach to local audiences at meaningful scale. On the other hand, local marketers who may only want to reach a limited audience do not benefit from the low marginal cost of automated buying.
Street Fight recently caught up with Frost Prioleau, chief executive officer at Simplifi, a Dallas-based programmatic advertising platform that specializes in localized campaigns, to talk about what programmatic advertising means for local marketers. He will speak on a panel at Street Fight’s Local Data Summit in Denver next week.
What impact has programmatic advertising had on the local media industry, and then how will that change in the next 12 months?
To us, the big thing happening is this ability to customize audiences to the local needs of local providers — which is a huge benefit for local advertisers because they don’t have to stick with the audience segments defined by national advertisers anymore.
What programmatic enables is the ability to target very precisely audiences that might be of interest to local advertisers. It’s consolidated a large amount of inventory and enabled local advertisers to then be very selective about where and when and to whom to show their ads. It really brings the ability of some very tight targeting to some smaller advertisers which in the past has been more easily available to larger national advertisers. Being able to go into a programmatic platform and target that same audience against all those medias is a big advantage.
Often, people frame hyperlocal as referring to the size of a geography, when it’s meant more to describe an “ intelligent” understanding of location. Can you talk a little bit about the shift from pre-packed to custom audiences and what that means for local marketing?
One of the challenges of local marketing is that many advertising platforms try to apply a nationwide audience segment to a very local market. We know local markets are very different across many industries. For example, in the auto industry the Toyota dealer in Palo Alto, Calif., has a lot full of Priuses and hybrid cars on its lot, and is after a very different audience than the Toyota dealer in Fort Worth, Texas, where he has a lot full of Tundras, pickup trucks and SUVs. If you were to apply a national audience segment to local Toyota dealers you’d be targeting an audience that wasn’t ideal for either of them.
With respect to geographic targeting and the traditional DMAs that are used, many are too broad for local advertisers. For example, in various verticals, grocery stores like to target within a seven-mile radius, auto dealers like to target within a 20-mile radius and other stores like to target much more closely than that.
Historically geo-targeting has been based on the location of the IP address of the user. We are now able to target on GPS data which is really precise, as well as data from beacons, which is in short supply, but coming, and that enables entirely new types of audiences to be targeted based on both the current location of users and the recent location of users.
In a programmatic environment, where marketers target data — not content — local media no longer has a monopoly on reaching local users. What advantages, if any, do local media firms have over their more general competitors?
Local media has a lot of advantages. The question is how much they put those to work. They have longstanding relationships with local customers, and based on those relationships they understand the needs of local customers. They have the ability to bundle advertising on their own properties with advertising across the web — whether that’s display, mobile or video advertising. Most importantly, what many don’t use is the ability they have to extend their existing audience that comes to their websites across the web.
Context does matter still is still a very important signal to the quality of the context of where the advertising is shown. It’s a very important signal to how effective that ad’s going to be.
Most people focus on a wholesale shift in media dollars from desktop to mobile. But as cross-channel targeting becomes more efficient, and data can flow from a mobile device to a desktop, how do you see the datasets used to identify intent on mobile, namely location, fitting with the cookie-based data typically used online?
What’s interesting about cross-device is the ability to flow both ways. Initially we are seeing a lot of search data from desktop devices being put to use for targeting users while they are on their mobile device.
But on the other hand we’re also starting to see the ability to take audiences generated by mobile location data and target those users when they are on a desktop or laptop, maybe a device where they tend to do more of their purchasing. We’re definitely starting to see that data flow both ways.
Let’s talk about incentives. Facebook and others have made a push to ditch post-click attribution models recently. What advancements in data are allowing companies to solve attribution in ways that it couldn’t five years ago?
The big move in attribution is from last-click attribution toward full-funnel attribution models. The definition of full-funnel attribution models is expanding. Not only are companies able to track attribution from other digital advertising or conversion to online, but we’re also starting to see attribution from purchases offline through CRM data linking and also through physical visits to a store by tracking beacons or GPS location.
The big move is the definition of what goes into full-funnel. It used to be full-funnel was just online activity now we’re seeing full-funnel include offline activity as well.
Retailers and others are collecting huge amounts of information about their own customers. How can marketers use what they know about their existing customers to better inform how they message new ones?
Look-alike modeling has been done in the online world for quite some time. Separately, there’s been look alike modeling done in the offline world for quite sometime. The developments over the last few years have been able to bridge the two types of look alike models so retailers can take offline data from their CRM systems and point of sales systems, profile their ideal customers/users and then target those profiles online with display, mobile and video ads.
Liz Taurasi is a contributor to Street Fight. This interview has been edited for length and clarity