Last week’s Election Day here in the United States was a bad day for data modelers, not to mention data overall. One thing that interested me (and many people) very much was that almost EVERYONE got it wrong. As we know the data leading up to the election did influence both candidate (Hillary Clinton felt she had Wisconsin in the bag) and voter behavior (the support for President-Elect Trump was under-reported as was the lack of overall enthusiasm for the Clinton campaign).

Imagine if there had been little or no data regarding the electability prospects of the candidates. Do you think they would have campaigned differently? Haven’t you always been a bit questioning of exit polls and their validity?

As data-driven business strategists and marketers, we absolutely LOVE data. Flying blind is never a good business strategy. Yet at the same time we’re also careful to question the validity of decision-driving data. Is the sample size large enough? Were time frames truly long and comparable enough? Can we truly have confidence in the conclusions we are making on the basis of the data we’re collecting?

I’ve always maintained a certain amount of skepticism when it comes to data and statistics. After last week it’s even more the case (if that’s possible).   I don’t love data any less, it’s more that I now want to dig deeply into the question – “what if the data is wrong?”

Bad data is comparable to disinformation. It often leads to the wrong conclusions and outcomes. No data is just that. In the absence of data all that is left are hypotheses. Those hypotheses without supporting data then become the launch pad for a series of tests to try to determine the potential success of one over the other.   That’s inefficient, not easy and will generate a lower level of success.

It is often heard in financial circles that ‘past history is no guarantee of future results’. Of course we use past history to help guide our future decisions. That’s never going to change. But taking all data at face value is dangerous and maybe more importantly not doing the entire job of vetting the prospects of success.

Here’s the irony. I am counting on data reporting and correlation to continue to improve and will never turn away from knowing more about what’s going on in order to aid strategic business and marketing decisions. But after last week my eyebrows arch a bit higher than they did previously.

Do you have the same, more or less confidence in data reporting today than before the U.S. Election of last week?