Day 22 30DC - Playing in Traffic

Part two of Ed's magnificent symphony in four parts is all about traffic -- getting it, keeping it, watching what it does. To that end, the folks at Thirty Day Challenge have provided us with a statistics tracker. We know the now-old business adage that you can't manage what you don't measure. Tracking statistics with any business venture, whether online or offline, is vitally important.

The other aspect to tracking -- which is really testing and observing results -- is to make changes only under controlled conditions. Think of it in the same way as doing a laboratory experiment. In order to know for certain which factors affect the outcome of the experiment, you must set up your test design (your experiment design) in such a way that you can control as many of the variables as possible.

Then as you run your test or conduct the experiment, you observe and record -- record -- results. If the outcome is not what you expected (or hypothesized), or to explore the results further, you then make a change in one (only one, if you please) variable and run the experiment again. Observing and recording once more. In this way you build up a body of empirical data that informs your decision-making about what has been going on in your test or experiment, and why these results might be occurring. This is the nature of scientific inquiry.

Remember, too, that you must run the test over a statistically-significant sample, over a reasonable time frame to obtain accurate data, and continually be vigilant for other factors or variables you hadn't accounted for which may be affecting the outcome.

Many marketers skimp on the market research phase of their projects, and suffer the consequences. Even more marketers do not conduct adequate testing of their campaigns and therefore make poorly-informed decisions (guesses, really) and also cannot prove to their clients what works and why, and what doesn't. Many marketers would be flat out of business if they billed based on results.

Once you know how and why you are getting the results you are getting, you can duplicate the test -- it's repeatable. More than that, it's predictable. That is the key for experiments as well, and the crux of the scientific method. Given the same controlled situation and variables, any other researcher should be able to replicate the experiment and get the same results. In business, it means that you can achieve the same successes over and over again, and even predict the outcomes. It doesn't get much better than that.