Piecing together the geotargeting puzzle

compass geo locationThe geotargeting of internet content has long been an established method of connecting with consumers in real-time. The benefits of doing so are readily apparent – consumers are 30-300% more likely to click a geotargeted ad versus untargeted; localised ecommerce offerings can increase conversion rates by as much as seven-fold; and geolocation can help reduce online fraud rates dramatically. As such, geotargeting technologies are deployed in a wide range of online applications, including advertising, ecommerce, publishing, social media and banking to present a safer, more relevant user experience.

While the benefits of geotargeting are apparent, marrying location to a relevant communication is far from simple, and the advent of new mobile and social technologies has increased the complexity of geotargeting choices available to marketers. A comprehensive strategy is required to capitalise on the opportunities geotargeting technologies present, and a good starting point is to understand each method of geotargeting and its relative strengths and weaknesses. Outlined below is a hierarchical view of the seven main geolocation methodologies that comprise the geotargeting ecosystem:

IP targeting enables the identification of a user’s location based purely on an IP address, identifying where the user accesses the publically routable internet down to the Internet Service Providers’ end-point equipment (essentially the “node up the road”). There are no privacy or legal issues with IP geolocation as the individual is not targeted, just the ISP infrastructure (not individual computers or mobile devices). IP targeting is recognised as the most versatile targeting method as it is anonymous and does not require opt-in, and it is universally available and easy to deploy. As such, IP targeting should generally serve as the ubiquitous first line of geotargeting for websites, apps and online advertising – types of content that can be targeted to a general local area (postcode level), possibly also enticing users to opt in to supply more information. IP targeting can also complement mobile device targeting technologies, providing accurate postcode location of web connections via WiFi devices and filling in the gaps when mobile users choose not to opt in to location-based services on their devices. However, as ISP end-points tend to serve a 3-5 mile radius of end users, IP targeting is limited to postcode-level geolocation and does not penetrate beneath the final 1,000 feet.

Location-as-a-service solutions are cloud-based offerings of derived mobile user locations from mobile network operators using mobile phone numbers and accessed mobile phone tower locations. Each LaaS solution provides developers access to location data from a micro-ecosystem of strategic carrier alliances, limiting location data to the network of mobile carriers included within the LaaS. This methodology can be quite granular in its ability to identify location down to a mobile phone tower triangulated radius of thousand(s) of feet or less; however, speed of service can be an issue, with delays of up to several seconds to resolve location. Additionally, users must opt in to be targetable by this service, limiting the addressable audience considerably. 

WiFi triangulation technology is used to geolocate mobile devices using the MAC address and location information of nearby wireless hotspots, the locations of which are compiled via “wardriving” (the act of searching for wireless networks by a person in a moving vehicle). To maintain an accurate database, WiFi triangulation service providers are required to constantly re-drive areas to update their database of WiFi hotspots, many of which can come and go frequently and affect the technology’s accuracy. The technology can be useful in dense urban areas where GPS signals do not penetrate and precise accuracy is of the utmost importance. Accuracy typically matches or surpasses mobile phone tower triangulation, identifying location within hundreds of feet; however, results may not be instantaneous, and again users are required to opt in to the targeting.

User supplied location information gathered through registrations or drop-down selections, for example, is only helpful to the extent that users agree to provide it. Even when they do, it is not always accurate – users often provide inaccurate information about their location. Sometimes this can be for privacy reasons; sometimes a web visitor is interested in content relating to another geography; or sometimes a user may supply home location information but actually be surfing the web at work or while traveling, etc. And by definition, the coverage of user-supplied data is far from comprehensive.

Cookies can be used to store a user’s location derived by a website using any of the other geolocation methods identified here, and then used to geotarget the user upon subsequent visits. However, cookies can serve the wrong information at the wrong time. For example, if a user is cookied in London and then travels to Madrid, targeted information on local florists could be off by 1,000 miles. Further, cookies are commonly deleted by users and anti-virus programmes, reducing their effectiveness for geolocation, and cookie-based technologies are coming increasingly under fire for being invasive.

Location-based proximity networks provide extremely accurate data as they generally operate within stores and are able to locate opted-in users within 200-900 feet of the point of purchase. Mostly used by big retail brands, shopping centres and other large venues, this method involves a controlled broadcast of relevant content sent through proximity access points strategically placed throughout high-traffic areas. Opted-in users receive information on promotions and offers, and interact with retailers through their mobile devices. However, retailers can only take advantage of this method with users who opt in, and the uptake is generally very low.

Global Positioning System (GPS) is the most accurate targeting technology for mobile device users (to within a few feet of the user), provided that the device is within unobstructed line of sight to four or more GPS satellites. Location services on most mobile devices do not actually use GPS most of the time, however. Actual GPS signals are usually the third fallback and are often not actually available even if people refer to the location services as “GPS”. Most mobile devices rely on triangulation of mobile phone towers or WiFi hotspots to determine location. Regardless of which actual technology is used for location services, accuracy on mobile devices can vary significantly, and utilising mobile device “GPS” information requires user permission to retrieve and deploy on smart GPS-enabled devices. GPS is also not applicable to desktop users. Finally, there are many legal and privacy issues, latency challenges and performance issues in dense urban areas that limit the scope of GPS-derived geolocation efforts.

Piecing It All Together

With all of these options, what dictates when one method should be used over another? Much depends on the objective, type of message, target audience and the required level of action and engagement. A layered approach is often the most effective, starting with an instant ‘’first cut’’ of geotargeted content using IP targeting and then encouraging users to opt in to a service in exchange for more relevant content, at which point even more precise geotargeting technologies may be deployed. Information is currency, and by immediately providing end users with content they value without opt-in barriers or invasive techniques, users are much more likely to engage and provide greater information in return. 

Frank Bobo is Vice President Operations for Digital Element and leads Digital Element’s international direct sales, partner channel management and strategic alliances. He brings more than 20 years’ experience in business development and operational leadership to the role.