Scientific Method in Marketing

Ron Shevlin posted a great article a while back about getting your business into social media. This quote from Ron really caught my attention,

As “innovation mania” sweeps through the halls of marketing, many marketers are looking to experiment with social media, lest they get “left behind.” While I’ve got nothing against experimenting, I am against experimenting for the sake of experimenting.

I couldn’t agree with him more.

If there was one thing that I got out of chemistry class during high school, it was scientific method (and possibly a love of making things change color, burn, and explode).

Scientific method always calls for a control, a set of variables and the discipline to change one variable at a time to see the effect versus the control.

This doesn’t simply apply to chemistry, physics or biology, but any instance of experimentation.

The creation of a new idea is always exciting. It’s easy to get excited about social media and rush to put a blog, MySpace page, or Facebook page up for your credit union to “experiment” with social media.

Without the principals of scientific method, it isn’t really an experiment at all.

Create a control

Creating a control is extremely important in experimentation. When applied in science it means running an experiment once without any variable present, like growing a culture of bacteria without any substances present to measure its rate of growth.

In marketing, you can create a control by running a pilot. Test the original idea within a selected group and measure their reaction.

Before piloting, identify what the variables in the project are so you can measure the response accordingly.

The other important aspect of creating a control group is to make sure the people in that group fit the programs target audience. Identify the types of people you want to aim for with the finished product and find people that fit the bill. What’s the use of testing a product targeted towards youth with a group that includes a portion of people who are 35+. That would be like testing the reaction of a flu virus to penicillin by testing the penicillin on a mold culture; it just doesn’t make sense.

Change one variable at a time

Once you have your variables and target audience identified and have started piloting the project, measure how your control group is responding based on that set of variables.

Don’t expect, or try, to overhaul the whole project at once based on the feedback you get. Make small changes one variable at a time and measure the effect it has on the control group as a whole. If you change everything at once, you will never know what variable it was that caused peoples reaction to become more positive or negative.

The purpose of a pilot should be to collect actionable data, and the more things you change at once, the harder it becomes to identify the specific changes that caused the reaction in the control group.

Test, test, and retest

The advantage to running a pilot before a full launch is that you get the chance to play with those variables. If you launch a product to the public and then change it every few weeks, people are going to start getting annoyed.

Take the opportunity to change variables in a closed environment. Test people’s reaction to each change methodically and carefully. Analyze their responses to each change and track the changes that elicit a positive response.

By doing this, you can create a finished product that reflects the nuances of peoples preferences and create a product that resonates as completely as possible with the target audience that was reflected in your control group.

So if social media fits into your brand strategy, go ahead and experiment, but do it right.