comScore’s ‘IYP’ Rankings: Whazzup?

Here’s the comScore release. I’ve made a long post at SEL about the data. I’ll reproduce the charts here (click to enlarge):

Search volume:

Market share:

Engagement (page views):

Compare the July, 2006 data:

As I said at SEL, clearly we’re not getting an “apples to apples” comparison between July, 2006 and the data released today. I surmize that comScore has altered its methodology and/or definitions.

The data released today put local search at 11% of overall search query volume, or 808 million monthly queries vs. 7.3 billion in general search. A big chunk of general search volume has a local intent and is not being counted as such. Recall the Webvisible-Nielsen study that showed 51% of respondents didn’t use a geographic modifier when searching for a local service business:

Now back to WebVisible’s research, which appears to support the idea that a large percentage of searches with a local intent don’t appear as such because they lack geographic modifiers. Here’s what the research determined about respondents’ local search query formulation:

• 51% used a general service term to search (“dentist”)
• 49% used a general service term and regional term (“dentist in Cleveland”)
• 23% used a specific business name (“Dr. Bob’s Dental”)
• 19% used a specific service term to search (“root canal”)

(Respondents had the option of answering more than one)

Interestingly, younger respondents (18-24) were more likely to use a geographic modifier than older users in the sample. But overall 51% of the actual, local search behavior didn’t carry a geo-modifier – that’s striking.

It strongly argues that search engines should be serving locally targeted ads against commercial queries in almost all service categories where there’s no local modifier because the probability is extremely high that the user is looking for a local business. In addition, geotargeted ads tend to perform better for the engine and the advertiser.

There are few meaningful distinctions between IYP sites and “local search” (except history and ownership). The consumer intent and use cases are the same. The critical issue is having some visibility on user intent. Here’s a less empirical but equally powerful case-in-point using Compete, Inc. data on the top retailer search terms:

As I said in my original post discussing this:

These are direct navigation queries for “online retailers.” Or are they? While all these companies sell online, eBay and Amazon are the only two true “online retailers.” All these other names are well-known retail brands where the overwhelming majority of purchases happen in local stores.

And what’s the most popular area of most brand retailer websites? You guessed it: the store locator.

These types of searches are not visible as local searches to comScore or Nielsen or Hitwise or Compete unless you “close the loop” and track what happens from the search engine to the ultimate conversion (offline). So what we are operating with is a conservative/tentative definition (and corresponding methodology) that undercounts actual local search volume in the market because it’s often invisible and hard to track.

If we want to get real about it here’s the definition we should use:

Local search is a process whereby users seek information online with the ultimate intention of conducting an offline transaction (service & retail).

That’s what’s really going on out there and accepting that as the definition of “local search” is radical because it starts to invert the relationship between local as a category and everything else online.

 

Advertisement