Comments came into my personal site last week suggesting that AutoTrader CEO Chip Perry, in his keynote at Kelsey’s ILM conference in Philadelphia, may have hyped up his company’s considerable lead over Cars.com and other auto portals.
“Based on my own personal brand awareness, I wouldn’t have guessed that AutoTrader was that dominate over some of the other competitors. Just for fun, I did a search for Chevy Suburbans within 25 miles of Philly (19019 zip actually). I only came up with 162, which is much different than Mr. Perry’s claim of 425,” said “JackOfAllTrades.”
“CarGuy” also disputed Perry’s assertion of dominance. “You are right, JackOfAllTrades. The numbers don’t match at all I checked Compete.com for some traffic estimates. They show AutoTrader at 5,564,966 UV/month and Cars.com at 3,603,126 which is hardly a 4:1 ration, if I am doing my math right … it is possible that Mr. Perry might be referring to Page Views or Searches or some other metric, which are extremely susceptible to differences in site design and web tracking methodology.”
But Perry, in a detailed response to us, notes that research is always complicated, and it is all readily explainable. “I can assure you that every statement I made at the conference was entirely factual,” he said. Here’s the text of his response:
Regarding the question about the number of listings I cited, our search results pages display on the first page the number of ‘matches’ or listings that fit the consumer’s search criteria. The number I cited at the conference was the number of actual vehicles contained in the search results, which is greater than the number of ‘matches’ that we show at the top of the page. The difference comes from the number of actual NEW VEHICLES that are included in the search results along with the used vehicles that match the consumer’s criteria. For example, when you do a search for ‘all’ (new and used) Audi A8s within 10 miles of zip 30309 (where I live in Atlanta) you will see we display 11 ‘matches.’ You will also see that the first ‘match’ actually contains four individual new A8s from Jim Ellis Audi. We are the only site that presents new vehicle inventory this way. We call this kind of listing a ‘New Car Aggregate Listing’ because we aggregate all the new vehicles from a dealer’s inventory that match the consumer’s search criteria into one listing. We do this to present a more easily scannable search result list than if we listed all the new cars individually. This type of search results presentation has been very well received by our users in both surveys and usability tests. So if I had used this example in Philadelphia I would have said we have 14 Audi A8s even though the top of the page says 11 ‘matches.’
I probably should have made this clear during my presentation, but I hope this clears this up and I hope you agree that the comparison I made was indeed fair. By the way, we have a change in the works that will display the total number of vehicles, not matches, at the top of our search results pages. And you may also be interested in knowing that this unique way of displaying used and new inventory (we are the only major automotive site to do it this way) is very well received by our audience which is made up of 33 percent new vehicle intenders and 67 percent used vehicle intenders. Many consumers bounce back and forth between considering a new or used vehicle during their shopping process, and many consumers like to see a new car dealer’s actual new vehicle inventory before visiting the store or generating a lead.
Regarding the question about our traffic levels compared to our competitors, the data I cited came from a publicly available October 2006 Comscore Media Metrix report, which does indeed show that our users’ average monthly time on site is about 2 and 1/2 times higher than cars.com and 4 times higher than most of our other competitors including KBB and Edmunds for example. During my presentation I cited comparative statistics only about time on site, not unique visitors or any other audience metric.
I don’t dispute the Alexa or Compete numbers but in the audience measurement game there are many different sources and you know as well as I do that they all have their various and sundry strengths and weaknesses. We are very confident that among the automotive shopping sites we have by far the largest audience of in-market car shoppers and we are, in fact, facilitating by far the largest number of actual vehicle transactions. We have a large amount of anecdotal feedback from our dealers that supports this, and to prove the point more broadly we’ve just completed a major study on this which we will be releasing at the NADA conference in Feb 2007.