The Center for Data Innovation spoke with Steve Hershberger, CEO of SteadyServ Technologies, an Indiana-based inventory management system for the beer industry. Hershberger discussed how the Internet of Things is disrupting the beer industry and what bar owners, distributors, and brewers can learn from real-time data.
This interview has been edited.
Joshua New: What problems in the beer industry does your company solve?
Steve Hershberger: With our Internet of Things toolset, we are setting the standard for near-real time beer inventory, order management, and business intelligence systems. SteadyServ’s iKeg and iQ systems are end-to-end mobile, software-as-a-service based enterprise systems. Each member of the three tier system—retailer, distributor, and brewer—benefits by eliminating inventory guesses and ensuring that their customers always have the right beer. The consumer benefits from knowing exactly where their favorite beers are on tap, how much a specific retailer has on hand, how fresh the product is, and which is proving to be the most popular by consumption. For business users, SteadyServ’s iQ analytics engine compares performance data in the market as well as integrated unstructured data, such as local weather and social media, and it provides recommendations designed to increase sales, customer loyalty, and product freshness.
New: How does better real-time data influence the whole supply chain, from bars to brewers to distributors?
Hershberger: The beer industry operates in two categories of retail. The first, “off-premise”, is comprised of package liquor, grocery, and convenience stores. Both “sell in” and “sell through” information—referring to tracking sales to the retailer and then through the retailer to the customer, respectively—in “off-premise” is relatively robust due to the dominance of well-established and sophisticated companies such as Walmart, Costco, Kroger, Safeway, Walgreens, and Target. Third-party services purchase and manipulate much of this data and sell it back to large distributors and retailers. Inventory management software suppliers also collect their own “sell-in” data, aggregate it, and resell it to industry.
The second, the “on-premise’ segment, is made up of bars, restaurants, taverns, sporting and entertainment venues, and the hospitality industry, and it has little to no reliable “sell through” data. Like with “off-premise”, there is some “sell-in” data but this is typically limited to the information each distributor or brewer has from its own facility. Data aggregators attempt to cull information from tax records and other such sources but this information is inconsistent and full of holes. Worse yet, it is on average 9 to 13 weeks old. Essentially, the $22 billion dollar draft beer industry operates in a complete vacuum when it comes to operational intelligence. In essence, everybody is guessing.
SteadyServ’s near real-time system—accurate down to 5 seconds and single pints of beer—bridges this gap, providing the members of the three tier system robust, highly accurate descriptive, predictive, and prescriptive intelligence to better optimize their position and market share in this very large and important segment.
New: The core of the iKeg system is data from a host of networked sensors. Why do you think nobody else has taken an Internet of Things approach to beer distribution until now?
Hershberger: Very good question. Let me illustrate why this is with a quote. Max Planck, the father of quantum physics said, “If you change the way you look at things, the things you look at change.” The beer industry is a very old school, closed industry. Beer people know beer with a specialty on either brand management and brewing, distribution, or retail optimization. IoT and big data are about as far away from these disciplines as you can get. They don’t speak each other’s language and technologists surely didn’t get the subtle nuances of the beer industry’s problem. Moreover, the beer industry had no way to articulate the problem and their need to IoT experts. Furthermore, they wouldn’t even know where to look to find an IoT expert, so the problem perpetuated for years. If you ask a beer executive, they’d simply tell you that “it has always been this way…it is what it is.” The problem is so severe and longstanding that MIT’s Sloan School of Management created a simulation for their graduate students called the “Beer Distribution Challenge” which is nearly impossible to win. The simulation mirrors the real life realities of the beer industry. What it took to solve this problem was someone with experience and skills in both industries, with an entrepreneurial background, a desire to solve problems, and really good timing to allow technology, cost of solutions, and need to merge at the right time. That is what happened here.
New: With all the data from your customers, have you learned anything new about how the beer industry operates? Anything particularly insightful?
Hershberger: Yes, we’ve learned a host of things. For example, most retailers think they waste the majority of their beer by shrinkage (bar tenders giving away product to their friends). While this happens, it is only a small percentage of the hidden waste. Three out of 10 kegs of beer are taken offline with on average 10-15% of the saleable beer left in it. Literally, on an annualized basis here in the United States, tens of millions of dollars of beer that retailers purchase goes unsold. This problem vanishes with near real time predictive and prescriptive intelligence that tell retailer exactly when to take the keg offline. Traditionally, the retailer would shake a keg to determine if it is “mostly empty” to avoid “blowing the keg” (completely depleting a keg during a pour, which can waste beer and upset customers). This waste has traditionally been an acceptable cost of doing business.
Additionally, the retail price of beer is highly elastic to a specific point that is based on popularity, availability, time of year, temperature, and number of competitors. Once it crosses that point, it becomes highly inelastic. The trick is knowing what drives the elasticity of that beer’s price and where that inflection point is.
Another thing we’ve learned has to do with how Stock Keeping Unit (SKU)—a distinct item for sale— growth is highly reactionary for beer. New SKUs follow the growth of category trends, such as with India Pale Ale (IPA). When IPA’s are hot, more SKUs are added to the category, creating a dilutive effect to the brands and SKUs seeking to profit from the category’s growth. To counter this, most move upstream such as a higher IBU (a measurement of bitterness) or higher alcohol content. Essentially, when the room gets crowded, beers have to be louder, more aggressive, or more over the top to stand out. Data would lead you in non-intuitive directions from where everyone else is going. Instead of brewing and promoting one of 500 double IPA’s, brew or sell a robust Saison where there are five options, knowing that the trend will swing different ways, including toward Saisons.
New: Can SteadyServ’s platform eventually be expanded to other industries, or are the problems it addresses unique to the beer industry?
Hershberger: The solution can be applied to a host of applications from wine, to soft drinks, to agri-chemicals, to certain types of pharmaceuticals, and even industrial gasses. We are focused right now on getting beer right and providing the gold standard for this industry though.