We Like a Year End list but LOVE Accurate Ones More

Minds like ours tend to get very excited about Year in Review lists and articles - especially when they are in our wheelhouse. A December 28th article about alcohol consumption in Ontario, posted by a ‘celestial’ Toronto news source was just what the data doctor ordered, albeit hard to substantiate what they wrote.

If the goal was street cred, the article was another great example of why organizations should always consider an extra set of QUALIFIED eyes to review or validate the numbers. If it was not street cred, as is the case for many publications made to look like anything but advertisements, then all we can say other is “ditto” to the qualified eyes comment.

We know there are lots ‘experts’ creating niche ways to slice the data just to be able to say, “We’re #1!” However, we know this data like the back of our hand, and yet we could not figure out how this particular data was sliced. Our intent is not to call out the authenticity and motives for the piece, or, reprimand authors for omitting vital footnotes that could provide explain methodology), instead we will just move to what should have been included with the graphics they chose to use ignoring most of the words that went along with the pictures.

For truth in our field of work, we always look for signs to let us know if results are trustworthy, biased or even the result of bad data grouping.?Going with the latter, the first error in judgement was a common trip wire in the field of analytics:?a Unit does not equal a SELLING Unit.?A Single Serve anything that retails around $4 is different from one Selling Unit of a 4, 6 or 12 Pack that usually retails north of $10.?The same can be said when equivalizing serving sizes in a bottle to a Single Serve can leave many wines and liquors out of the list. We will give them a bit of grace for that lapse in expertise (or even basic math), but we still could not match the results they reported, even when trying to replicate the perceived error in methodology.?The reported Unit winner was never the actual overall unit winner.

Moving to retail sales Dollars should make this easy to get right.?Unfortunately, the way the data is reported breaks the easiest of measures and now with questions surrounding the two most basic facts. In our 15+ years of analytics, including over a decade in Wine, Spirits and Beer, we have not seen the identified winners win anywhere as near as much as reported.?The products shown in the full graphic (not just our slice) may have won over the reported timeframes in their respective segments, but we cannot find a single period where they beat all comers. That means the graphic that shows White Claw as the best seller in 19 of the 20 periods is not the correct takeaway. Hennessey is a leader in Cognac but claiming that it outsold everything else in dollars is another motivator for us to write this.?For a Cognac to outsell a Vodka, Whisky or Beer (even inside the LCBO) would take a lot of marketing money and probably some form of planetary alignment to likes never seen before.

Everyone deserves to know where they stand on their own leaderboard, to know what has succeeded so navigation going forward is easier. The earnest ones know how to avoid the missteps that could just as easily put them on wrong road.

CMI Insights Inc. is a specialist in making sense of big data and at the forefront of analytic innovation and data science for 15+ Years. Happy New Year and we would be more than happy to offer a second set of eyes for your lists.