Consider a survey item that has 50% favorability, 40% uncertainty, and 10% unfavorability. And another item that has 50% favorability and 50% unfavorability. Despite favorability being the exact same for both items, the first item has more uncertainty. The second item is polarized between favorable and unfavorable perceptions. These differences suggest that focusing on the entire spectrum of ratings is more important than focusing just on one metric.
That point is well-reasoned and technically accurate because one metric can never show the full complexity of results. However, it also overlooks three important aspects of focusing just on favorability:
- Simplicity and consistency are crucial for having effective discussions about survey results. Focusing on one metric (favorability) not only reduces complexity, but doing so also makes it easier to interpret and increases the chances that more employees will be able to contribute to discussions about survey results. Some teams may prefer more complexity, but these decisions are at the discretion of each team leader.
- Responses that are less than favorable are less than ideal. The two example items noted above have 50% “less-than-ideal” responses, regardless of the varying levels of uncertainty and unfavorability. Uncertainty and unfavorability become irrelevant because the goal is to maximize favorability. That goal, by necessity, means minimizing uncertainty and unfavorability.
- It is very easy to over-analyze results and get stuck in "analysis paralysis". A lot of time and effort can be spent (and potentially wasted) trying to figure out...
- why the first example item has more uncertainty
- why the second item is so polarized
- which item takes priority
- how best to address those items because of their uncertainty and polarity
...all before talking with employees. This analytic strategy is certainly well-intentioned, but it is a distraction from the most important activity of all: creating an open and consistent dialogue with employees about survey items that have lower levels of favorability.