The latest issue of Arkansas Business has an article on the cover below the fold entitled, "Quantifying Internet Traffic Remains Elusive". Since I make a living in the Internet Trafficking business, I am glad to see it get some front page news - thanks Nate. But I would like to elaborate on some of the points in the article.
The main point of the article is that it is difficult to obtain precise traffic estimates for sites that display advertising (see the title). Should it matter to marketers that these specific traffic estimates are difficult to pin down? In a word - no.
In traditional advertising, those selling the advertising could only point advertisers towards circulation numbers or television viewers and some demographic data to attract advertisers and set their rates. So naturally, traditional marketers still have this frame of mind when judging the Internet.
On the Internet, however, the marketers' focus needs to shift from circulation numbers to measurable ROI. There are great tools like Google Analytics and Omniture that allow Internet marketers to trace any digital campaign all the way back to a real ROI number. If I spend $10,000 per month advertising on SFGate.com and I can show an ROI of 40%, should I care if the site gets 2.4 million or 7.6 million monthly visitors?
The metrics that really matter are not ‘circulation', ‘viewers', ‘page views' or ‘time on site' but profit and loss.
Nitty Gritty Details
The article speaks to a specific niche of Internet Advertising - buying banner ads directly on another content-driven site. Think of the Metropolitan Bank banner ads that are displayed regularly across the top of ArkansasBusiness.com. ArkansasBusiness.com has made a good living out of selling space on their website in the same way they sell ads to statewide businesses in their print edition of Arkansas Business.
According to the Interactive Advertising Bureau, which is sited in the article, display advertising of this kind accounts for 22% of total Internet advertising revenue. The article doesn't speak of the advertising done on search engines like Google, Yahoo or MSN, which accounts for 40% total spending. For search engine marketing, getting the exact number of impressions (times someone was exposed to your ad), clicks-throughs (number of times someone clicked on the ad) and conversions (times someone took the action you wanted them to take for coming to your website) is straight forward and reliable.
Why are Internet traffic measurements elusive?
Measuring traffic, or visitors, on a website that you don't own is tricky. The main reason is because that is the proprietary information of the website's owners. As Nate mentions, there are services that can estimate a website's traffic for you. comScore, Nielsen/Net Ratings and the Audit Bureau of Circulation are sited directly in the article. In my opinion the best service for getting data on another website is Hitwise, which was not mentioned. Before advertisers start citing the wide variance in these traffic estimates, they need to understand how each of those services works so they can make intelligent analytical adjustments to the estimates. Maybe I will touch on the differences in the services another day.