With all the recent writing about local search I was prompted to reflect on what are the “real” local numbers? What do you mean “real”?
I’ve consistently argued that the Internet – from a commercial standpoint – is ultimately more about local than almost anything else. For example, the US Commerce Dept. data reflect that ecommerce is about 3% of total US retail (about $3.6 trillion), but almost $400 billion in offline spending is Internet influenced.
The recent comScore data (July 06) asserted that 13% of overall web search is local, or approximately 849 million monthly searches vs. 6.5 billion in general web search. When I was at The Kelsey Group we developed an empirically based estimate that roughly 20% of web search manifested a local intent.
comScore also says that 63% of the US online population (roughly 109 million people) conducted a local search in July. Those are clearly huge numbers. But as I’ve said, the definition of “local search” (IYP + search) that comScore uses doesn’t tell the whole local Internet story very well. So I want to lay out some of the numbers to offer a better sense of how many users are looking for local information (broadly defined) online.
- Total US Internet audience: 172 million (comScore, May 2006)
- “Local Search” (search + IYP sites): 109 million uniques (comScore, July 2006)
- Jobs: 49.8 million uniques (comScore, February 2006)
- Cars: 42 million uniques (excludes OEM sites; comscore, April 2006)
- Real Estate: 42 million monthly uniques (comScore, April 2006)
- “Classifieds” sites: 37.4 million uniques (e.g., comScore, July 2006)
- Newspaper sites: 55 million uniques (Nielsen//NetRatings, February 2006)
- Mapping sites: 65.1 million uniques (comScore, March 2006)
- Travel: 41 million uniques (top 10 sites only; Nielsen//NetRatings, August 2006)
- Personals/Dating: 31.6 million uniques (comScore, January 2006)
- Shopping (shop online, buy offline): 80.8 million uniques (Dieringer Research Group, 2005)
The first thing you might feel the impulse to do is dispute some of these traffic numbers as “local.” For example, Travel is a heavy ecommerce segment and thus is potentially controversial as a “local” category. But fundamentally travel is about place and drives related local spending. (comScore has pointed out related site visits before and after visits to travel sites that are inherently local.) Dating is also often not local, but typically is. Visits to car sites may or may not be local (but transactions overwhelmingly are).
In addition, there’s clearly overlapping usage in these categories so the numbers are not additive. I don’t have any way to provide unduplicated uniques. But all of these categories and some not mentioned (e.g., social networks/social directories with a local element or emphasis) have “local intent” traffic not captured by the narrow comScore local search definition above.
Some people might also be inclined to argue that the traffic data above underestimate actual usage in certain categories and some might argue the opposite. Putting aside the precise accuracy of the traffic data in any given category, these are extremely large numbers of users. And in the overwhelming majority of cases the transactions that result from the user behavior are local.
Some of the behavior seen on these sites is not going to be like that of search, where a user is actively seeking information that will ultimately result in a transaction. For example, people might be on newspaper sites reading news stories that are not inherently local and won’t drive any consumer purchase behavior.
But lots of behavior that is “directional” is happening on these non-search sites and thus should be considered in thinking about the “local search” marketplace. In one way of thinking about it, “local search” is ultimately about actively seeking information online and buying something or engaging a service-provider offline.
And that’s a much bigger deal than the “niche” or “vertical” that most people assume local to be.