Two Tips for Making Good Use of Marketing Experiments

Strategic chaos is common with small businesses. Even medium sized businesses are often unfocused with their growth strategies. But I understand why. A small business owner never stops thinking about how to grow, and ideas on how to squeeze more out of the turnip come in a random, ceaseless stream flecked from a journal article she read yesterday, or a podcast she listened to on the way into the office this morning. This constant head chatter leads to excitement, which can lead to trying anything and everything whenever the mood strikes you.

I will never fault someone for experimenting with their marketing. I just want to encourage people to do it in a disciplined way.

1. Re-read the annual strategic plan

I advise my clients to experiment within the confines of the annual strategic plan. You set goals for the year, marketing experiments can help you determine the best tactics to reach the goals you've set. So make sure your experiments are aimed squarely at achieving a specific annual goal. This may seem obvious, but if you're not careful, it's easy to chase the latest flashy object in the marketing swimming pool.

CEO: "I just heard that Snapchat is the newest tool in to build a brand new customer base. How long will it take to get our company account created."

CEO's Son: "Dad, you sell health insurance. Snapchat is only popular with teenagers and people who don't care about their health insurance."

CEO: "True. In our annual planning session we decided that people who don't care about their health insurance were not going to be our target market this year. Perhaps it's better to spend our resources elsewhere."

2. Create a campaign scorecard

Every experiment should have it's own scorecard. Even if some of the metrics are the same as the ones you're using month to month on a different dashboard, take the time to set up a separate dashboard for the campaign.

You're running an experiment to test a set of variables. In most cases, it's best to isolate the experiment from the rest of your marketing efforts. That way, if you see a change in output you'll have some confidence in attributing the change to inputs you're tinkering with. So with an isolated experiment, you should have an isolated scorecard to track the results.

This will be really helpful in 13 months when you're marketing manager gets the same idea, and you have to tell her that you tried that and it didn't work. If she is any good, she'll ask for some evidence. You'll be able to hand her the scorecard for the experiment you ran a year ago, and neither of you will have to waste any more time on running the same campaign.

Good luck and happy (disciplined) experimenting!