Geographic location has become an essential model for targeting end users including dynamic site content, search, social media, etc.
The existing geo-targeting practices for search and display media focus on targeting by country, state, region and city. This is all based on IP addresses many of which are dynamic or inaccurate.
You can however custom target by entering a street address and defining a radial boundary a mini Google Maps interface. Getting down to detail there is the option of using longitude/latitude coordinates which can be as accurate to within 100 meters.
So, how do you put in place an alternate solution bypassing the inherent limitations of IP targeting? In search, we often develop a second campaign, ideally with a larger targeting radius and implement geo-specific keywords. While the geo-targeted campaign may have the keyword “pizza” this may only hit a portion of the intended geo-targeted audience. To reach the remaining audience, we typically extend the geo-target radius within a completely new campaign with extended keywords such as “Fitzroy pizza”.
But let’s look at how this works when we use IP address location as the basis of location based targeting.
Google uses MaxMind’s database for mapping IP addresses to a geographical location. They claim it is 99% accurate. What is in the fine print, is that it is 99% accurate in determining the country. This figure will not improve, but drop dramatically with the rise in mobile devices, wireless connections.
Accuracy is also defined as being within 40km which means Melbourne and Frankston are local to each other. So searching for a local Pizza Hut while you are in Frankston might give you Universal Pizza in Lygon Street. Still, probably worth the drive as their pizzas are great. 😉
But from an end user point of view, is this considered local? Well no, not to me, I consider local to be within 5km. Like the ads for local businesses in your local newspaper.
The real happening place for location-based targeted marketing is clearly in mobile and mobile applications; in particular, social media applications that emphasize check-ins and connect you to places and people nearby. There is a world of difference, opportunity, and data between marketing location to where a computer resides and marketing location to the person who self identifies, checks-in, and announces their preferences and next stops.
Google has emphasized that location is an immediate and important relevance enhancer. However as the technology continues to evolve it remains both a limiting factor and a promise. HTML5 can allow permission-based, location-targeted ad delivery to consumers browsing on the mobile Web. Currently, it is supported in Chrome and Firefox, but not in Internet Explorer. The use of location-based data remains a point of debate, and just last week Apple posted a warning in its developer forums that if they use location-based data primarily for targeting ads the app will be rejected. Many apps already ask your permission to use your GPS location for app functionality, like the Google toolbar and Foursquare. You can even optionally geotag your tweets.
Likewise, the sharp increase in smartphone adoption all over the world raises both opportunities and challenges. Not only are more people using smartphones, but the devices now cover more of their needs and their day with a multitude of rich, engaging applications – many GPS driven.
Today the database at MaxMind states that it is 83% accurate for the USA and 62% for Australia. I believe this number is overstated and will only get worse which is why the data has not been updated since 2008.
All this makes geographical location services based on IP address unpractical and obsolete.