In the last two weeks my department has brought on a new hire. This new hire is fresh out of college, no prior advertising, marketing or media planning/buying experience. As the fellow responsible for both strategic development and the day-to-day operations of this department, it has fallen to me to teach her… well… everything!
Among the set of "everything" is handling data reports. It's funny to review these reports with her because she asks questions about them that, after so many years of looking at them, make me remember that not everyone knows what they mean.
But it isn't just the new media assistant who doesn't have an inkling of what online media reporting is for, or what the data mean. Plenty of clients don't either. And it is easy for lots of us, with so many years of experience, to forget that the reasons for things are not always clear or that the articulation of those reasons is necessary.
Many clients know that the data being generated by their online media plan can be used to demonstrate the success or failure of a particular online marketing effort.
So what do you do if after running a campaign for a few weeks the data reveal that the effort is not meeting expectations?
Data must be turned into information
The first thing to do before altering your media plan is to turn the data into information and hear what it is telling you about both the campaign and the site's weaknesses. Using one of the major third-party adservers (e.g., Atlas or Doubleclick's DART or MediaPlex, et cetera), you can track basic activity on each page of the advertiser's site on which you want to note traffic.
What is the creative telling you?
Perhaps the sites selected for your plan were not appropriate. The way to see this would be the rate of response to multiple creatives run against a single placement, followed by the conversion rate for that placement. If I'm not converting anyone, regardless of the creative unit, it's a good bet the placement is bad.
Or maybe I have multiple content category placements within a single site. If this is the case, treat each of those intra-site placements as if they were unique unto themselves, and read the data this way. You might find that a site overall is falling out of bed, but a specific area is doing well.
What is your own site telling you?
What if the data tells you that you're getting plenty of audience to the site and you are getting them there efficiently, and their quality seems to be right, demonstrated by their moving past that first page and going deeper into the site, but the conversions are still dismal?
This would suggest something is wrong with the site. If I can track a user through a sequence of pages, it may turn out that something else is amiss. An example would be a multiphasic registration process or sales path.
I have to say, it was fun explaining to the new media assistant what "impressions" are and what "clicks" are and what the relation of the two might be able to communicate about the media, the message being carried, and the audience being carried to.
But the most important thing I think I taught her is that a test online cannot fail. It is just as important (and valuable) to learn what doesn't work well as it is to learn what does work well.
And then I threw in a quote from Churchill, "Success is moving from failure to failure without loss of enthusiasm."
This is a good lesson for my new media assistant. And it is a good lesson for clients.
Media Strategies Editor Jim Meskauskas is vice president and director of online media for ICON International, Inc., an Omnicom Company. Read full bio.
