“You don’t know what works until you test it, so test everything.”
Over the years, my marketing mentors have instilled in me the understanding that marketing is 10% creativity and 90% testing. Examined through this paradigm, marketing takes on an almost scientific nature, where every piece of data is scrutinized until the truth becomes undeniable from the weight of empirical evidence.
As you can probably guess, this involves more than a little knowledge of statistics, which may or may not be your specialty. Of course, analysis is pointless if you can’t understand it. Ask your marketing analyst these seven questions, though, and you’ll have a pretty good picture of your test results, regardless of their complexity.
- What do the results indicate?
Here, you’re asking for the objective findings. The results might indicate, for example, that the first ad variation did significantly better than the second. The answer to this question should include cold, hard facts supported by statistical measurements.
- What conclusions would you draw?
In contrast to the first question, here you are asking for subjective results. Over time, many marketing analysts develop a sense of what a data trend looks like. Thus, while there might be no empirically provable findings, your analyst will likely be able to offer his or her gut instinct as to what can be learned.
- Did the results meet with your expectations?
Results that fall in line with expectations probably don’t require any additional scrutiny. After all, you just confirmed what you already suspected to be true. Results that fly in the face of expectations, however, often demand further questioning, especially if they go against the expectations of an experienced marketing analyst.
- How were these results produced?
Assuming there is any real knowledge to glean from all the data, it’s worthwhile to question that data’s integrity. What sampling method was used? How large was the sample size? What sort of statistical comparison was performed? By asking your marketing analyst about the methods used to obtain the data, you can often get a sense of its reliability.
- How much do you trust the results?
Much like the first two questions, it’s always worthwhile to switch back and forth from hard facts to gut instinct. After all, the research methods and results may sound reasonable even while your marketing analyst has shortcomings about them. Always be sure to ask his or her opinion on the data’s legitimacy.
- What next actions would you recommend?
Based on the conclusions and how much trust you place on them, you probably have some ideas of how to move forward. Then again, your marketing analyst probably has some good ideas, too, especially if further testing is called for.
- What do you predict will happen if we take these actions?
As the scientific method dictates, research is cyclical. Now that you’ve figured out what the data indicates, how much you trust the findings, and what you want to do with that knowledge, it’s worthwhile to speculate what those actions might produce. This feeds back into question three when you’re going over the next round of results.