The power of location audiences: A buyer’s guide & scorecard
There has been a lot of discussion about proximity marketing and location data. If you haven’t been following the emergence of location intelligence platforms, they are data platforms that ingest location signals collected from GPS, WiFi and sometimes beacons, and combine those signals with other data sources to do a variety of things, including:
- Create location-based audience segments/targets. Signals collected can be used to create location-based audience segments such as moviegoers, retail shoppers and others. These audiences are created when the signal, such as lat/long, is collected and combined with other sources, such as a point of interest (POI) database. Additional attributes and insights may be added to the data, such as frequency and demographics, to create highly targeted audiences based on the individual’s offline behaviors — e.g., frequent (three or more times per month) department store shoppers. These segments can typically be found on various marketplaces, including DMPs and DSPs (data management platforms and demand side platforms).
- Optimize in-flight campaigns. The ability to track and measure the impact of an ad impression served to a given location-based audience segment/target and its ability to drive the walk-to rate or visit. These real-time campaign optimization tools are typically used by agencies and allow the users to drive incremental results by monitoring the impact of multiple creatives and/or geographical regions. A/B testing and the corresponding “lift” reporting allows campaigns to be adjusted in-flight to further drive marketing ROI.
- Inform future audience and campaign efforts. Post-visit reports, typically managed by the agency and location intelligence platform, allow brands to not only track lift associated around footfall traffic, but also to better understand audiences and visitors to specific locations, including their activities, journey paths and brand affinities.
But not all location data is equal. So whether you’re an agency or a brand interested in utilizing location data, it is important that you do your homework and ask the right questions before investing in location data, buying audiences or looking at offline attribution. Here are some key questions to ask:
How was the location data collected?
Yes, methodology matters! Location data and audiences can be collected from various companies and in a variety of ways — for example, through a network, which is typically a method closely associated with telco operators. Data can also be collected via an ad request — otherwise referred to as bid stream or on an ad open — via a panel or through an SDK (software development kit) integrated into an app.
Understanding the data collection methodology can yield valuable insights on the quality and scale of the data.
What data is collected, and what is its density?
Just as the methodology differs, so does the quality — including the data density collected. Density can vary dramatically. As an example, SDK-based solutions typically run off apps where there is persistent background location running, such as a weather app. Because these SDKs tie into the device’s GPS and WiFi capabilities, they typically can collect a lot of data points based on the device’s location and the WiFi signals detected without the app being opened.
On the other end of the spectrum, bid stream data collection typically has less than 10 percent of the data density because it relies on the user to open up an ad in order to track location, providing less data on the user’s location and journey. While data density matters, it must also be combined with a strong POI database in order to provide the proper context to the location.
Is the data accurate?
The demand for location data is increasing dramatically, and the lack of standards around how such data is collected and used has resulted in a large discrepancy in data accuracy. In fact, according to the “MMA Demystifying Location Data Accuracy” white paper:
Up to 60% of ad requests contain some form of location data. Of these requests, less than 1/3 are accurate within 50–100 meters of the stated location.
Network-based solutions like telco operators rely on triangulation across cell towers and can typically identify a user’s location within a ZIP code or neighborhood. While bid stream data may also collect lat/long, its accuracy is often questioned, requiring many providers to further scrub the data.
SDK-based solutions that leverage GPS and have some WiFi verification, including signal strength, results in greater accuracy and precision (10 to 100 meters). When combined with WiFi and beacon signals, these solutions can, on average, report accuracy within 10 meters 85 percent of the time.
But accuracy and precision may not be enough. Critical context, such as dwell time, should be considered. For example, a dad picking up his daughter at the movie theater clicking on an ad served while waiting in the parking lot is significantly different from a dad at a movie theater with a dwell time of two hours. One is a moviegoer, the other is just her ride. Buying the right segment requires this level of diligence.
What sources were used?
In mobile, “the more the better” is usually the rule of thumb, as scale matters. Is your location’s source limited to an individual app, based on a panel, an individual network or ad calls? Or does it represent a diverse and large universe reflective of the US smartphone population itself?
What is the approach to privacy?
What is the privacy framework? Is personally identifiable information (PII) collected? Do publishers have control? And does the provider give the user the ability to opt out at the publisher level?
Once you’ve selected a location intelligence source for audiences, campaign management and transparent footfall attributions may be a consideration. Look for location intelligence platforms that can provide real-time test and control optimization tools for campaigns and provide footfall lift analysis at the state, DMA (designated market area) and store level with full transparency. Attribution reporting should include visualization of the data, including impression and geographic POI placements, in-depth impression analysis including day parting and hours, visit analysis and uplift reporting, audience analysis and brand affinities, time analysis and more.
Still need help? Check out the attached (Excel) score card, allowing you to consider, weigh and score critical features and function in your location intelligence platform partner.