Can we please stop trying to measure the web like we measure magazines?
Some of you longtime readers know that I’m not a tremendous fan of media plans that are put together solely on the basis of what the various syndicated research companies say. I prefer to look for contextual fits for a client’s product, rather than depend on a syndicated study’s assessment of a site’s demographic composition. I’ve been seeing more and more planners use these studies to justify a buy, and it really bugs me.
Here are my problems with using syndicated audience research on web sites.
- Web users make different uses of the sites they visit. What does the average Yahoo user look like? How about the average Women.com user? Or the average LookSmart user? My answer is: Who cares? Almost all demographics are represented on the web. And what we’ve come to refer to as a “portal” can contain a plethora of different demographic and psychographic groups, all using the portal in a different way. Would you say that the people looking at stock quotes on Yahoo Finance are interested in the same things as people who are surfing through the content at Yahooligans!? Not likely.
- We consume information differently via the web than we do via magazines or television shows. When I pick up my monthly copy of Guitar World magazine, I’m interested enough in the content to read every single page of the magazine (ads included). Do you read every single page of every web site you visit? I didn’t think so. For this reason, measures of “reach” attributed to web sites for given demographic audiences should be taken with a king-size grain of salt.
- If the average buyer of pet food in the real world is typically female and between the ages of 35 and 49, should those who sell pet food target W35-49 online? Heck no. I’m a guy, between the ages of 25 and 34, and I faithfully plunk down $10 or so a month to feed my pet fish. If you’ve got pet supplies to sell, you’d be better off trying to reach me by advertising on a pet-related site, rather than going after an age/sex demographic. Moreover, advertising with a pet site will give you a better chance of reaching me while I’m in the proper mindset – while I’m either shopping for pet supplies or looking for pet-related information – and will likely give you better results.
The web is the ultimate narrowcasting medium. It is a giant collection of niche interests. Some of those interests are extremely obscure. Others are more mainstream. To conduct a successful advertising campaign, the best place to start is by capitalizing on synergies between your product or service and those niche interests. If you’re selling stereos online, the best place to start advertising is with sites that cater to audio enthusiasts, and not with the sites that happen to have the highest coverage of your traditional media buying target.
I’m not saying that web-based audience research is useless. To go back to the pet food example earlier, I would likely make use of @Plan to determine the current and potential size of the market by crossing people who purchased pet food online in the past 30 days against the general web audience. This definitely helps in the formulation of marketing and media strategy. However, I would not then cross people who purchased against specific web sites the survey won’t tell me the specific areas within those web sites that will be best for reaching my target.
Web-based audience research can serve as a terrific resource, if used in the right way. Just remember, it’s not the only thing that should be steering your buy. Context is king.