If you love beer and Data Analytics – and who doesn’t? – you may be excited to learn that the two pair up together in a system for bars and restaurants. These retail establishments use it to help them better understand and manage their on-tap inventory and improve their sales.
Twenty billion dollars’ worth of draft beer is sold every year in the U.S. The company thinks there’s an opportunity for outlets from local pubs to enterprise-owned restaurant chains to do even better – and to waste less product in the process, too.
The SteadyServ iKeg solution involves the company’s own Cloud database of virtually every beer ever brewed in the world, as well as the Internet of Things and RFID technology. The SteadyServ system is composed of SmartScale devices; each of these IoT devices sits in the cooler underneath each keg on tap, next to the beer line affiliated with that tap. Each SmartScale is numbered to coordinate with the beer line that it’s paired up with. Each beer line gets its own RFID identification tag that itself is paired up with a SmartScale as soon as that device is introduced to the system. Once done, the SmartScale knows, for instance, that it is Scale Number One and that it is affiliated with Tap Line Number One.
Once a new keg is put on the system for the first time, the SmartScale sensors recognize its weight. Then, in combination with the company’s database, the system does the initial work of determining what type of beer it believes is in the keg. It does this based on matching the keg weight information to its database of beer facts that includes various brands’ specific keg gravity and vessel sizes. Each beer keg, says SteadyServe CEO Steve Hershberger, has a specific gravity, and the system uses Machine Learning to narrow things down from there to one or two potential beer brands with a high degree of likelihood. The SteadyServe comprehensive beer database, in fact, has information for 62,000 unique beer SKUs, with 12 to 60 pieces of Metadata attached to each, providing a real time database of draft beer as well as on-premise retail activity.
After this first step is taken, “we push an alert to the establishment that it put a new keg on Tap Line 1 and ask for confirmation of beer type,” he says. Now the system knows that the keg affiliated with Tap Line Number One is Boston Beer Rebel IPA, for instance. That matters because the knowledge will help inform key business decisions for the restaurant or bar.
Making the Bar Business Work Smoothly
The sensor makes it possible to pinpoint-measure how much beer is left in a keg as drink orders are placed and filled, and matches that to how many pours that translates to. When it gets to the point where the keg is close to going empty, the system sends an alert to bar or restaurant managers and owners via a mobile device app. Or employees can scan the RFID tags directly for real-time updates.
On a busy night, when multiple alerts are issued for multiple kegs, staff members will have exact insight into how many pints are left in each so that they can send staff members out to change kegs starting with the one that has the least amount of beer left just before it’s drained. “If it goes all way empty that introduces air into the line, which means you have to re-prime the pump,” Hershberger explains.
Not only does that take time but it potentially risks a sale of a particular product. “If you blow a keg and tell the patron it will take 10 minutes because their beer choice isn’t available, the patron drinks one-third less,” says Hershberger, whose career has spanned the gamut from performance and Data Analytics to co-founding production craft brewery Flat 12 Bierwerks. Blowing a keg also introduces waste into the environment since it leads to the first ten or so beers drawn from the tap being “all foam,” as he puts it. To avoid this happening, many bars and restaurants take kegs off when there’s still plenty of beer left in them, losing money on these actions.
“We found that of every 10 kegs that are taken off, three have about 15% of beer left in them,” he says.
The Liberty Tap Room in Myrtle Beach, S.C., uses the solution. “It keeps bartenders on their toes if they know the inventory down to the ounce, and helps us not waste beer,” says GM Brittany Moricle. A keg could look low but there’s still ten pints left, she says. Now, “instead of just feeling it, you know with greater accuracy that there are ten vs. the four you thought were there.”
The Data Analytics system also applies Machine Learning in the Cloud, using the information it gathers about keg brand use at a particular restaurant and POS system sales data to build analytical models that provide insights of how much of each type of beer has been historically sold there. That provides actionable information to help the bar or restaurant better plan the right tap variety.
“Since we now track all they sell, the question is what’s selling, how well and what products are on the shelf,” Hershberger says. He notes that most bars and restaurants don’t have an appropriate product mix optimization when it comes to tap beer. Tracking what’s moving and what’s not lets the establishment put brands and types of beer on tap that are attractive to a wider variety of customers.
Another issue the system helps with is matching the amount of beer left in the keg to the beer’s freshness date, so that the business can take steps to sell it before it goes bad. Its visual analytics let owners or managers see how much longer the keg is fresh and provides pricing recommendations to move sales along. For this, it leverages the information its database collects for retail transactions across every SKU its retail base sells.
“The reality is that there is a price utility zone, meaning an optimal price for a pint of beer,” he says. “If you charge outside of it, whether you raise or lower the price too much, consumption actually drops.”
It also can auto-curate a business’ inventory information with a dynamic digital menu on bar or restaurant TVs that shows customers what’s on tap and how much is left of each beer. “When you put that monitor in with that information they sell between 5 and 10% more draft beer,” he says. For example, when patrons see there’s only ten pints left of their favorite beer, “they’re more likely to order two in quicker succession with dinner so they don’t run out.”
Roll Out the Barrels
The Data Analytics system became commercially available last summer but the company really began scaling it up in the fall. Today, it operates in 21 U.S. states and four countries, soon to be five. Customers include stadium, convention center, and airport dining sites, as well as franchise groups that manage brands including Applebees, Mellow Mushroom, and Tilted Kilt.
SteadyServ also currently supports keg wine and it’s preparing for keg coffee to make a big splash, too. “We are doing testing with a major brand on how to support them as they start to roll out keg coffee,” Hershberger says.
It sees further opportunity in extracting retailers’ POS sales data and analyzing that in context with its beer database information to help in creating suggested pairings between food and beer – teriyaki chicken wings with pale ale, for instance. SteadyServ has a pilot underway with US Foods, one of the largest institutional food providers, on this front, to help its retail partners build such pairing menus by understanding the top five or so beers they sell and the top five foods to pair with them.