Our account service lead for one of our clients had a question from the client. “How come your performance is so good?” On this account we are actually pitted against another agency. There is complete transparency between the accounts and the client. So, everything is open to everyone.

However, the question was still posed. The reason for the question is the apparent disconnect between commonly accepted SEM “wisdom” and what the client sees in the account.

Excessive long tail campaigns, minutely segmented campaigns, absurdly geo-targeted campaigns are generally accepted as the necessary basis for SEM. I have been given every excuse for this: The long tail gives efficiency, the minute level of campaign setup provides exact control, the level of market focus lets us account for geo-variations in buying habits…

The problem (primarily) with using this approach out of the box is that it breaks one of the basic tenets of marketing: Understand your audience.

When you introduce sever complexity to an SEM campaign at the very beginning, you are assuming knowledge about your market that you don’t have. We talk about the audience of SEM in terms of product segments, geo, age, etc. However, we seem to ignore the one common factor among them…

they are using a Search Engine.

This very fact, that we are targeting them with SEM, and they are targeting us, introduces an overriding similarity to all users.

We talk about the differences between users, and how we will exploit that. But, these differences are often not as grand as SEM program complexity would suggest. Will you find long tail conversions? Of course. But, you need to let the data tell you where these are likely to originate.

The only way to truly do this is to start with simplicity; let the SEM data tell you where the opportunities are. This data is the overriding commonality among all the users, yet we distort it and artificially manipulate it by introducing complexity, based on ignorance, from the very beginning.

This is not an advocacy for throwing everything into one ad group or one campaign. It is, however, a voice for a little common sense, a bit of humility and a recognition that patience is important in understanding the SEM landscape.

SEM programs may end up complex. This complexity should be introduced gradually, as the data indicates. The manager needs to understand the interplay between the campaigns – something that can only happen if you introduce them at a sensible rate, over time and as the data leads.

Overly complex campaigns divert the focus from productive activity, causing managers to jump around chasing the last campaign metric for the 100th & something campaign in the program. We have become so enamored with the idea that we can get the CPA cut in half on a particular campaign, that we ignore the fact that the same campaign generates 0.002% of the leads and will have absolutely no material impact on the total campaign performance. We continually choose to ignore the fact that 90% of our productivity on a campaign will come from less than 20% of the keywords.

The sources of this misguided approach come from across the spectrum. I see this on the agency side, where they sell complexity and confusion to create a dependency by the clients. I have also seen this on the client side, where they assess program progress based on the number of keywords being managed, or the number of campaigns being optimized. Simultaneously, they hound an agency about performance while insisting that they focus on furthering campaign complexity rather than optimizing the program. We have collectively accepted this fallacy of complexity.

Its time to rethink the common wisdom, admit we are working with imperfect knowledge, and exercise some patience to allow the market (through the data) to show us the path to smart (perhaps complex) campaign expansion. In some cases, the smart path may be a very short one.