Marketing Research provides the greatest benefit to internal stakeholders when it helps them “predict the future.” Predicting the future can involve answering several very different kinds of questions. For example, “How should I develop this asset?” “Should I acquire a competitor?” “Which message or campaign will achieve the highest market share?” and “Who in my target audience is most likely to purchase my new product?” all ask for different types of predictions.
When stakeholders ask us to predict the future, they generally are seeking guidance on making the right choice between a set of options. In reality, we provide decision support by helping our customers or clients decide which set of options is likely to provide the largest return. This entails three types of outputs:
- Probability of success. Decision makers care less about a single revenue or share number than they do about the probability that a given action will achieve a target level. When a client says, “Assure me that I’m not overpaying,” she wants to know the probability that she is not spending too much money on the purchase, not the number that predicts on the average how much her company will profit.
- Logic behind the estimate. Virtually every senior executive I have ever met believes his instincts are excellent. If you provide an answer he does not trust, they will tend to discount it. The way to avoid this (particularly when delivering bad news) is to pinpoint a few critical market factors that drive your estimate. If the factors make sense, he may be more willing to reconsider his initial perspective.
- Levers that can improve the outcome. For a decision maker, knowing the levers that will improve the probability for success provides two potential benefits. For a “go” decision, this guidance will drive actions that improve the probability for success. For a “no go” decision, this guidance may ease the path to deciding that the action is not likely to achieve even minimal success targets.
Implicit in all this is the importance of “KISSing” your work (keeping it simple). Data geeks see a positive correlation between the number of data points in the model and the ability to convert it into a credible finding. I have heard more stories than I care to remember about powerful 1,700-line spreadsheets, comprehensive two-hour telephone interviews with migraine sufferers, and Monte Carlo models controlling for 125 separate distributions. In every case (certainly these three), the results that came out the end were difficult to interpret due to their complexity, but, I was assured, “right” because they were based on so much detail.
This approach is an algorithm for obscurity, not proof of insight. A limited number of easy-to-interpret data points will provide far more clarity, and clarity is the key to high-value decision support.
We can get into ways to present data that enhance impact in a different article…
Decision makers: what kinds of data help you most? Are there specific ways you like to receive or review it?
Consultants: what have you learned about enhancing the impact of your work? How do you like to frame decisions for your clients?