Loading ...« | Home | There’s Something About Craig »
Does Close Enough Really Count?
By Cameron Ferroni | November 19, 2007
IP targeting of local sites has always been a hotly debated topic, and some of the recent posts around the Web about the SEO challenges around it have just added fuel to the fire. As such, lately I’ve been thinking about what the best approach truly is, and when is close enough, good enough, if ever.
To frame - IP Targeting (at least in my parlance) is when a site determines the location of a user based on their IP address, and provides them with content specialized to their location, rather than a generic site. We’ve done it both ways here at Marchex, with varying results. The old arguments were pretty simple - you can’t get 100% accuracy of determining someone’s location from an IP - with AOL customers for example everyone can show up as being from the same IP, independent of where they are located. Or you may have issues with a hotel in one city using an ISP in another city, causing people to get results for Florida when they are sitting in their hotel room in New York. Add to this the fact that there are many different companies providing these targetting services, and, well, you just can’t get 100% accuracy. Estimates vary widely, but in general you could expect about 80% accuracy within a 25 mile radius. Certainly better than baseball players, but not exactly A+ performance.
That said, given the choice between getting it right for 80% of your users, and having 20% having to pick/change their location manually on first visit, versus having 100% of the users having to pick/change on first visit, the 80% was a reasonable win. A recent post at http://www.localseoguide.com/geotargeting-location-by-ip-address-seo-death/ (written by one of our faithful readers, which I’m sure will spark more commentary!) has made this a tougher argument however. You should check out the full post there, but the gist of it is that IP targeting also has a pretty negative impact on SEO. The reason is that when the search engines come in and get geo-targeted, they get results for that region, versus all results. Now there are lots of ways to get around this - whether by detecting the engine coming in, or having the local specific stuff in an i-frame, or using ASP or PERL to sub the dynamic content in, but again - most of those are sub optimal, and/or a lot of work.
Ignoring my biases about having to hack around with user experience just to make search engines happy, it still leaves us in an interesting predicament, which is the question I have - how do people feel about this? Is the user experience worth it even in the 80% case? Do users really care about having to pick their city first? Is the SEO kicker as bad as everyone says?
Topics: Local SEO, Local Search, Content |
November 19th, 2007 at 6:58 pm
Nice linkbait!
In my experience the SEO kicker is pretty bad v. the benefit. I guess if you could show a large improvement in conversions (i.e. more page views, ad clicks, etc.) from displaying geotargeted content it might be worth it, but I have yet to see a geotargeted app that was so unbelievably compelling that it was worth the SEO headaches.
November 23rd, 2007 at 6:11 am
As an investor, I couldn’t help but notice you all own Beijing.com. I would hope to see that asset used for the ‘08 Olympics, and I think a good opportunity to get the Marchex name out there and maybe drive some traffic through the balance of your local sites.
January 8th, 2008 at 12:54 am
At first read I would have thought that I would let the IP Targeting remain with the PPC folks and their creation of unique landing pages that are location specific.
80% is a good number, but what worries me is the 20%. The one thing you can’t tell is that how many out of each group (80 and 20) are going to buy something on impulse when they arrive at the website.
But then the odds are so random that it’s a tough nut to crack.
Personally I find it annoying and wonder about the technical capability of the company I am visiting when I have to enter the correct information so they can present me with information for my locality. Oh yes, I do hold them accountable for their actions.
Tough call though and with all the possible random IP’s and anonymous surfing or IP blocking that can be done, where do you have to draw the line. When do you cross the line of infringing upon someone’s assumed privacy before they give you permission? Do we even have to worry about those kinds of things?
“Inquiring Minds Want to Know”