Today’s CPG retailers and manufacturers have several variables to contend with in order to drive sales and meet company goals. As retailers, manufacturers and consumers do not exist in a vacuum, maintaining and gaining market share relies on implementing mechanisms to account for circumstances beyond the control of the various parties involved. Recent research showed […]
The great English poet and romantic Lord Byron, usually the composer of flamboyant verse and intricate alliteration, when faced with the eternal question that comes before us all, chose to answer it with analytic simplicity: “What’s drinking? A mere pause from thinking!” It is with such a noble principle that I choose to explain my frequenting of the local public houses. I am an analyst after all, and I never truly stop analyzing.
It was one such thought-provoking evening when a friend asked me to join him at a snazzy hotel lounge in San Francisco. The aforementioned gentleman is actually working towards a Master’s of Wine and seeking out different wines for tasting is mandatory for the course. Or so he keeps telling everyone. This particular evening, however, he was looking to break from type and work on mis-identifying single malts. I needed little convincing and was out there with him every step of the way, contributing to his education and my thinking process.
Except, that day was different. We enjoyed single malts at the hotel and then, embracing the true wonder that is San Francisco, decided to follow that up with a visit to a dive bar and meet up with some more boisterous friends. It was at the latter establishment, while jostling for position to place my order at the bar, that I suddenly had my moment of inspiration. Working on beer, wine and spirits (BWS) analytics for IRI over the last 6 years has been extremely challenging, especially since syndicated data sources only cover the off-premise consumption. So, most of the work we have done so far is around analyzing how purchasing dynamics play out in grocery or liquor stores. It is based around picking up a product off the shelf and paying for it at the counter. Essentially, it’s straight-forward pricing, occasional merchandising, and basic price promo, mix, assortment modeling with a custom BWS angle. Obviously, this assumes a fair amount of homogeneity in consumer behavior even though price sensitivity for a single malt scotch may be different in a liquor store versus a grocery store, but it is not radically distinct.
But that logic goes hurtling out the bar window when we move our focus to consumption of alcohol in bars, restaurants and clubs – the entire universe known in the food and beverage world as “on-premise”. The same consumer that during the early part of the evening didn’t think much of shelling out $12 for a glass of Glenmorangie, just ended up going to another establishment whose star attraction was a $2 a pint of Anchor Steam. The packed bar was testament to the effectiveness of that particular campaign. And, when the hipster to my left ordered a round of tequila shots and just handed his credit card to the barkeep without even really hearing how much it was going to cost him, it all started coming together. Two things about that order were so unique: the apparent lack of price sensitivity and the generic nature of the request. It was enough to cause psychological warfare within a brand manager’s mind: “My consumer is not price sensitive, great! But wait a minute, he doesn’t really remember my brand by name, not so great. “
Extremes aside, price sensitivity, promotional incrementality, mix effectiveness and other sacred paradigms of advanced analytics are still relevant, but they need to incorporate a certain finesse and nuance that warrants a different approach – and almost a different data structure. For instance, if we were able to track beer purchases on the same check through the night, would be see a change in price sensitivity as the night goes on? Are more shots ordered right before last call, or are they merely a function of the amount of alcohol consumed? Or both? Was the $2 beer really worth it for Anchor Steam, or did they just help drive overall bar sales? The questions are numerous and of extreme relevance to suppliers and retailers.
And, IRI now has a major player that actually collects on-premise data at a scale which enables answering these questions – GuestMetrics. This firm has already set up highly granular data sources for on-premise data and is currently working with IRI on exciting data and analytics opportunities. The data comes in at the check level and, given that granularity, is also able to shed light on what was purchased and at what time. They also integrate various happy hours and other promotions that are run in these establishments. Retailers are no longer subject to the black hole that is bar sales without resorting to depletions, and suppliers and distributors are able to better spend their time and effort supplying, pricing and promoting the right mix of these wonderful offerings. Not to mention what it offers to the world of advanced analytics! The observant reader will not fail to see the similarities between this and IRI’s store level databases and analytics, and the near-revolutionary impact it had on analysis and business decision-making within CPG. This could do the same for BWS.
Every generation has its own frontier. Lewis and Clark might have had the West; Kirk, Spock and co. have Space. But for thinking folks like me and my fellow analysts, exploring the on-premise frontier will be the journey of a lifetime!