Understanding IP Targeting for PPC Campaigns
IP Targeting is one of the technologies that Google, MSN, and Yahoo employ to target searchers. IP Targeting is often misunderstood. It has both advantages and disadvantages; however, before those can be discussed, we must first look at how the technology functions.
Assigning IP Addresses
Whenever you logon to the web, your computer is assigned an IP address from your host (your internet access provider). This is essentially your computer’s address on the web. However, as it’s assigned from your host, your address is tied to your host’s as much as it’s assigned to your computer.
The major engines (and many others) map these IP addresses to physical locations. Since there is an actual location associated with your computer, ads can then be served based upon your actual location.
The advantage of this technology is regional targeting. The job of an advertiser is to connect with searchers. Connecting on a local level is one method advertisers can employ so that searchers relate to the advertisement.
However, the disadvantage is that advertisers may be connecting with the incorrect geography.
IP Targeting Map
In this example, the grey area represents a DMA (Chicago in this example). There are three searchers, and two host providers.
Searcher 1 is located inside Chicago, and the host provider is located inside Chicago; therefore, the ads served to searcher 1 are geographically appropriate.
Searcher 2 is located outside Chicago, however, the host provider is inside Chicago; so this searcher will see ads targeted to the major city in the area, however, they aren’t as specific as possible. This most often happens in the suburbs of large cities, or when there are few host providers in an area (usually rural).
Search 3 is located inside Chicago, however, the host provider is outside the city. Therefore, searcher 3 will not see geographic ads for Chicago, however, they might see them for the location where host 2 is located.
How often are irrelevant geographic ads served?
I don’t believe there are exact statistics of this measure, as when it occurs, the technology thinks its being mapped appropriately.
Based upon what I’ve heard, what I’ve seen, etc – I don’t think it happens that often. However, it happens often enough that the limits (and advantages of connecting geographically with a searcher) should be known.
However, for those users it does occur for, it can be quite annoying. I’m a case in point here. I live in northern Chicago and my host provider is in Evanston (the city just to the north of me), so when I’m at home, I see Evanston ads, not Chicago ones. The advantage is that I often travel to Evanston for shopping as its closer than downtown Chicago. The disadvantage is that I’m downtown almost everyday, so if I’m searching for downtown Chicago locations, I see inappropriate ads. At my house, I’m searcher 3.
The good point is that when I’m at the office (located in the loop), I see Chicago ads; so I’m searcher 1 while in the office.
In addition, there are other geographic technologies at play. This is only one of the mappings that the major engines do to serve geographic ads – there are others which can help reduce how often IPs are incorrectly mapped.
Geo Qualified Keywords
Using geographic targeting in the engines does not mean that only those in your targeted region will see your ads. The engines use geo-keyword mappings to also show your ads. I’ll make this the subject of another article – however, it is important to note that if you only serve ads to Chicago, and someone in New York City uses the geo-qualifier Chicago (with your normal keywords) – your ads could be shown.
Drilling down on geographic keywords is an article that goes into depth about the geo-keyword possibilities.
The AOL Conundrum
For several years, all AOL users appeared to come from Virginia. This was because AOL routed all IP addresses through their Virginia servers. However, AOL is no longer a major issue for IP targeting.
There are many more AOL broadband users than dial-up users these days. Broadband users are assigned IP addresses differently than dial-up, so they can be served regional ads.
Those users still using dial-up (this stat came from someone at Google) is less than 0.5% of web traffic. It should also be noted, that the engines keep a list of providers and IP addresses that are black listed from being served IP targeted ads because of known limitations with those hosts.
There are other items that are also commonly used to serve regional ads such as user registration data (Microsoft Passport accounts, Yahoo accounts, etc), geographic keywords, and others.
IP targeting should be noted as a technology used in PPC regional targeting, not the entire process. It’s a common mistake to refer to regional targeting (or geographic targeting) as IP targeting. IP targeting is just one technology among many that are used to serve regional based ads.
However, it is very important as advertisers to understand both the benefits and limitations of the technologies being employed. Ultimately, you should control your ad serving, and being informed is the best way to take full control of your PPC campaigns.
I have the exact targeting issue Brad has, except much more extreme. I am in Dallas and get Geo-targeted to St. Louis.
Unlike, you I rarely travel 2.5 states to purchase anything. This makes any geo-targeted results suspect at best.
I can do a search for Dallas Car Dealer and right in the middle of the page I will get a geo-targeted ad from a St. Louis car dealer, and often a second ad saying Missouri as well.
Odds are that you’re logging in through a corporate VPN or a host provider that’s located in another state.
This is when IP targeting falls down and does not give users the proper experience.
Although, IMO, an explicit local query for ‘Dallas car dealer’ should bring up Dallas ads and not Missouri ads (should, doesn’t always do so).
This is where I’d like to see the SEs put in some time to refine their local targeting algos.