In common parlance, precision and accuracy are virtually synonymous and used almost interchangeably. But to a statistician, they have very distinct meanings.
The classic illustration:
An archer who is able to release a quiver of arrows and have each arrow hit in virtually the same place each time could be described as very precise. But if the spot where those arrows converge is an outer ring, you might question that archer’s accuracy.
Similarly, if a quiver of arrows consistently lands around the bulls-eye, but never actually hits the center, you might describe that archer as accurate, but not very precise.
So is it better be precise, or to be accurate?
A false choice, you might argue. Both are good. They are better together. And in most circumstances, we’d be hard-pressed to disagree. But from an Agile perspective, we would suggest that precision is, well, overrated.
Firstly, though, let’s give precision its props. Undeniably, in experimental circles, where the ability to replicate an experiment and get the same results each time is at a premium, precision is paramount. Without precision, there would be no advancing scientific theory and virtually no ability to define scientific “truths.”
In the world of market research, precision (when coupled with accuracy) is generally a good thing as well. The value of replicable findings that provide a complete, finely-tuned picture of an entire marketplace is hard to dispute. Particularly if time is not an issue. Or budget. Or if you operate in a market that is stable and determined to remain so.
None of our clients, however, operates in what anyone would describe as static markets. Upheaval and disruption, usually resulting from rapid advancements in technology and digitization, are the norm.
The need for consistency over time diminishes as the need to act immediately increases. Of course, that action must be the right action, in the right direction, toward the right goal. But immediacy and decisiveness become key. The bias shifts toward accurate action, rather than patient precision.
Which is where Agile Research diverges from conventional. Too much time continues to be spent in our industry designing exhaustive surveys, over-engineering sampling schemes, shooting for completion rates that are far in excess of what a marketers need to accurately determine a course of action.
The larger the sample, the better the study, goes the thinking. The more robust the numbers the more reliable the findings. The design of too much research—particularly quantitative research—falls prey to this notion, something we refer to as “the illusion of precision.”
Agility demands accuracy over precision. Pinpointing the right audience, isolating the right issue, determining the right action, of course. But keeping precision in perspective.
Given the volatility facing marketers today, actionable accuracy trumps replicable precision, every time.
Agile Research 