When dealing with organization-wide strategy, the impact of an error can affect many lives. The use of the Scientific Method helps to ground strategy decisions in real-world customer experiences, and increase the likelihood of uncovering potential risks and crafting contingency plans for them before they become actual problems.

​The Scientific Method is generically defined as a standardized set of rules that apply to an experiment that aims to uncover the reality behind a given situation. 

• WARNING: The experimenter may have a desired outcome from the experiment, but the results should be irrelevant to and unbiased by that outcome. 

• Rule #1: Before beginning the experiment, the experimenter must define a hypothesis: a statement of what the experimenter expects to happen to the subject of the experiment (e.g. a person, animal, or object) during the experiment.

• Rule #2: The experimenter must use a process called “operationalization” to translate and document all variables within the experiment in a way that is objectively measurable. A “variable” in this context can be defined as any property of the test subjects, or their surrounding environment, that may vary from subject to subject.

  • CAUTION: Try to minimize the number of variables between test subjects to increase the accuracy and repeatability of the experiment. 

• Rule #3: The experimenter then designs a set of procedures to prove that the opposite of their hypothesis is correct. If they succeed, their hypothesis is proven wrong, or irrelevant. If they fail to prove the opposite, then their hypothesis is validated as true.

  • WARNING: It is insufficient to simply prove that your hypothesis is correct. It is possible that your hypothesis could prove true, and the opposite of your hypothesis could also prove true. In such cases, the variable that you are measuring is proven to be irrelevant to the outcome. 

  • TIP: It is an important part of the Scientific Mindset to try to prove yourself wrong. Only through searching for the points of failure in your ideas do you find the opportunity to be right. 

• Rule #4: The procedures of the experiment must be outlined and documented in a way that it can be repeated independently by anyone else without the involvement of the originating experimenter. Only when an experiment can be independently verified by other skilled scientists can the hypothesis be assumed to be true.

​An experiment is said to be “causational” if it can be shown that a direct change to one variable in the experiment (or a particular combination of variables) can be shown to “cause” another variable to move proportionally in response. If the chain of cause and effect between variables cannot be operationalized, measured, and documented, then an experiment is said to be “correlational” and it can only be proved that one variable moves in “relation” to the movement of another variable (or combination of variables).

• CAUTION: A causational relationship often has much greater validity, but is also much harder to achieve than a correlational one. 

• CAUTION: Correlational relationships between variables are often the best that can be achieved in an experiment, but they are also subject to interpretation and thus introduce “noise” that could result in false predictions about future behavior. 

​Data is often collected in two forms: “qualitative” and “quantitative”. 

• Qualitative data is of very high quality and reveals in-depth insights into the exact experiences of the subjects of the experiment, but requires a large amount of time to collect and process effectively. As a result it is often employed for only a small subset of the customer base, which leads to a risk that the information found does not translate to other customers not studied.

• Quantitative data collects information about a large, representative portion of the customer base to identify commonalities and trends, but is not able to accurately determine the reasons behind the choices of individual customers. 

• TIP: An effective strategy usually involves early qualitative research with a few potential customers to identify a hypothesis about larger trends in customer behavior. This research is then followed by a quantitative research to see if the behaviors found in qualitative research hold for a larger audience in the same ways. 

​The outcome of scientific experiments in business is not intended for use in marketing, but internally to help build an overall strategy.

• WARNING: As discovered by Nobel Prize winners Daniel Kahneman and Vernon L. Smith, people do not make rational decisions in their best interests based on data (a formerly-held position of economics), which makes logic only tangentially helpful in marketing.