When most people think of location they think of an exact point that is both precise and accurate. In reality, location has levels of precision and not all location data is created equal.
Consider a content publisher like CNN. With nearly 15 million pages of content there are hundreds of millions of opportunities for ad placements. The idea is for them to get the most value for those placements and the way to do that is to make sure that they convert for their advertisers. In other words, the value of the ad depends on getting the right message to the right person at the right place and time. Relevance to the user increases the chance they will convert. The more precise your location data, the better chance you have of creating a relevant experience. Let’s step back and take a look at location precision and accuracy and then talk about what we can do.
There are two ways that a publisher can obtain location: Using an IP address or via a device that has location services turned on (on-device location). While both of these methods result in a latitude and longitude, only one is the accurate and precise location of a device.
Network Topology or Top-Down Approach
In order to understand exactly why an IP location is less precise, you have to understand the technical details to the top-down approach, also known as network topology, for how location is obtained including:
The finer points of IP addresses
How IP addresses are assigned by a carrier
The methods carriers use to estimate location: trace routing or pinging the network to time the signal from origin to the estimated router at the given location.
The amount of error in the result leads us to colloquially call this fuzzy location.
Nearly eighty percent of location is obtained by publishers using an IP address. This method is neither precise nor accurate. Most of the time, the latitude and longitude returned using IP location is the center of a civic area (let’s call this a centroid) with a large margin of error. The average margin of error range is a square mile, but we have seen scenarios where centroids are placed in the middle of the country and in the wrong state.
It’s like this:
Irma Paul wants to invite her friend over for dinner. “Hey Addie, would you and your boyfriend like to come to my apartment for dinner tonight?”
“That sounds like fun, IP!” says Addie. “I’ll bring a salad and a bottle of wine. Where is your place?”
“It’s somewhere in New York City.” says Irma.
“Can you be more precise? That’s going to be hard to find.” says Addie snarkily.
“I think it’s in New York City.” says Irma.
Another way to obtain location is on-device location. First, the location is accurate and precise. This is because it uses a combination of signals from Wi-Fi, GPS, cell towers, the IP address and even the device sensors to determine the location down to 10 meters. You may also hear the term ground truth location, meaning that the location is obtained from the closest possible sources on the ground.
Additional battery and bandwidth efficiencies are gained if a publisher is running Skyhook’s SDK as it smoothly and efficiently loads information (in tiles) onto the phone to allow us to determine location without making requests. This also means that once location is obtained, we can capture additional location points as the device moves, depicting continuous motion through space. This is what we consider good location.
If you are a publisher, which would you rather have? The geographic region a user is in or would you like to build personalized experiences based on precise location behavior and the venues a user visits. Good location allows you to do more with loyalty programs like figuring out if someone is spending a lot of time at Starbucks and rewarding them for it. Or delivering relevant content to the right person at the right place and time.