When measuring website traffic or planning a new website there are two common questions:
- How many people visit my competitor’s websites?
- How many people can I expect to visit my new website?
While each question is simple – accurate answers are critical to making good business decisions. The common tools that people use to compare metrics are Comscore, Alexa, Compete.com, Nielsen NetRating, Hitwise and others. Each has some advantages and disadvantages. To understand how the numbers are calculated we should understand how the ratings are acquired:
These measurement approaches are similar to the Nielsen TV rating we have all heard about. A group of people/families volunteer to have their web surfing behavior monitored and compiled.
- Advantages – panel member demographics are well understood, traffic is accurately measured.
- Disadvantages – traffic patterns are extrapolated from a small group resulting in less overall accuracy.
Alexa, ComScore and Nielsen use this approach. The Alexa and Comscore numbers are actually calculated based on click measured directly in a users browser sessions.
These measurements are compiled by merging the traffic behavior of a group of ISP measurements of web browsing behavior.
- Advantages – large numbers of web surfers can be measured so total traffic estimates are more accurate.
- Disadvantages – demographic behavior is less well known so demographics of visitors are not as well understood.
Hitwise and Compete compile ISP traffic data as the basis of their metrics.
In our testing we have seen a large underestimation of overall web traffic in the tools available. Compete.com which does provide a ‘people’ estimate showed visitors of 50K per month on average for a client of ours who more typically sees visitors in the 200K range. Does this mean that their competitors are also seeing 4 times as many visitors as indicated by Compete.com?
The same issue comes up when planning a new business. While investigating the traffic for potential competitors – the online measures were all on the low side. One site with a know revenue for $7M / year was shown with about 5K visitors a month. This is equivalent to a conversion rate of 233%; obviously some of the metrics are out of sync. If you are starting a new business which statistics do you believe?
So online metrics are not accurate – so what do you do?
Should competitive metrics still be used? And if so how can the metrics be used when the accuracy is still in question?
The answer lies in the fact that we have measures where before there were no measures at all. And like all decisions when it comes down to it you have to weigh the accuracy of the information available and make the decision.
This does not mean settling for questionable metrics, obtaining several different sources of information is critical to gauging the accuracy. In addition comparing competitive metrics against known baseline data will help to further evaluate and compensate the available information.
As always analysis is required before you make decisions based on any web metrics.
For further reading:
“Web Numbers: What’s Real?”, Business Week, http://www.businessweek.com/magazine/content/06_43/b4006095.htm
“How ComScore can track your mouse clicks”, The Register, http://www.theregister.co.uk/2008/05/12/inside_comscore/
“Competitive Intelligence Analysis: Why, What & How to Choose”, Occam’s Razor Blog by Avinash Kaushik, http://www.kaushik.net/avinash/2006/08/competitive-intelligence-analysis-why-what-how-to-choose.html
“Website Traffic: Why Can’t We All Agree?”, Nat Ives, AdvertisingAge April 14, 2008, http://adage.com/adnetworkexchangeguide/article?article_id=126134