Hacking the TOM Beer Challenge
Even with its rapid growth, the beer industry is one of the only “just in time” industries left. It’s an industry weighted down with archaic inventory management and ordering systems, prime for IoT.
“Imagine if HP guessed at what Best Buy needed to fill its shelves on any given month without talking to the company or looking at any hard data,”
-Steve Hershberger, Chairman & CEO of SteadyServ Technologies.
While we are all well acquainted with the TOM Beer Challenge and the famous Bullwhip Effect, I bet that most of you did not know that this game was developed at MIT back in the early 1960’s – shocking! Similarly, I also bet that most of you aren’t aware of SteadyServ Technologies, the latest Internet of Things (IoT) startup attempting to solve this real-life problem.
Inefficiencies in the Beer Industry
“Our inventory management systems are very lacking. In many instances, we don’t know what we have, we don’t know how much we sell, and we don’t know when we sell it. We lose sales because our customers want what we don’t have.”
15 percent to 20 percent of every typical keg delivered to retailers ends up as waste. Considering that draft beer is by far the largest contributor to a bar’s margins and that a total of $23 billion worth of draft beer is sold each year at 340,000 locations in the U.S., a significant incentive exists to improve operational efficiencies.
Studies have also demonstrated the importance of maintaining beer inventory in bars level – when a patron’s beer choice is not available, they tend to drink 1/3 less beer.
If you were a bar manager, how would you know how much beer to order from your distributor? The current practice requires the bar manager to head to the refrigerator to count kegs. Since he is unable to see how much beer is left in each keg, he must lift each keg to see how full it is—essentially making the ordering process a guessing game. In some instances, this back-breaking task of assessing inventory in order to place their orders takes a minimum of two and a half to two hours a week, per location. In practice, if the beer rep doesn’t get the order in time, the rep will place it for them by guessing.
Introducing the SteadyServ iKeg
How does this IoT technology work?
Kegs are first tagged with a radio-frequency identification tag that has information on the keg’s contents (brewery, type of beer, volume, age). Once the keg is delivered to a bar, it is placed on a special sensor which measures the keg’s weight. As the beer gets depleted, real-time data is automatically aggregated and sent to the cloud where it is accessible to bar managers and distributors via the iKeg app interface. This technology takes the effort and guesswork out of assessing inventory, and by knowing exactly when a keg is going to run out, this eliminates the dual problem of unhappy customers and scrambling bartenders when a keg “kicks”. Customers pay $30 a month and up per location for access, plus about $4 a keg for the hardware.
Click here for a promotional video featuring the SteadyServ iKeg.
In addition, when combined with sales data, anonymized competitor data, and consumer demographic data, this technology provides powerful analytics capabilities:
- Enhanced order accuracy by forecasting beer demand based on current consumption levels
- Understanding timing around beer consumption (different styles popular at different times and days) and when to launch new products and promotions
- How beers are selling compared to competitors
- Predicting combinations of beers that appeal to a target demographic
From the discussion above, it is clear that retailers (bars and restaurants) are the big winners here, with the patrons benefitting as well but to a lesser extent. The SteadyServ website claims that breweries and distributors are big beneficiaries as well but I remain somewhat skeptical. If a real-life supply chain resembled that of the TOM exercise, with just one entity in each part of the supply chain, this would indeed solve the problem since all upstream entities have complete information on total retail demand. However, since retailers vastly outnumber breweries and distributors, and given the limited penetration of the iKeg technology, distributors and breweries are unable to plan accordingly given the small, non-representative glimpse into true retail demand.
In order for the whole supply chain to benefit from this technology, penetration must be increased to capture a significant proportion of retailers. Encouraging uptake at any scale is challenging and early successes must be demonstrated ASAP. One important first step, in my opinion, is to experiment on a small US region with a local ecosystem of breweries, distributors, and retailers (possibly Vermont, Oregon, or Colorado), encouraging uptake through subsidies. Armed with some success stories, SteadyServ can pitch the iKeg to larger distributors and breweries, players with deep pockets who stand to gain most by supply chain optimization.
Student comments on Hacking the TOM Beer Challenge
Interesting article and technology. I would be interested to see if a major player in the large variety beer scene, such as World of Beer, doesn’t eventually turn to technologies even on a pilot basis. Given the rapid growth of craft beers the quantity of beers locations need to keep on tap or in bottle seems to have grown significantly over the last 5 years and appears to still be growing.
As a person whose family member owns a pub, I really enjoyed this article. I remember taking inventory as part of a summer job once and it was pretty miserable and inaccurate. I do agree with your point that it’s still fairly unclear how much distributors and brewers benefit from SteadyServ at this point. You do almost need a closed system to show incremental improvements over the current process. It’s going to be absolutely crucial to prove a true ROI for smaller bars to invest in a system like this; I’m not sure how StedyServ handles the hardware, but perhaps they can lease it to the bar to cut down on the initial costs? Also, with the infusion of other technologies into bars i.e. advanced POS systems, gauges on liquor handles to measure pour volumes, etc. you start to wonder if all of this technology becomes a burden given typical high turnover of bar staff and the subsequent need to constantly train employees.
I actually do think that there is some value here for the breweries. For large breweries it is always useful to understand how their beer brands are being consumed across geographies and time. These things change over the years as well – a certain kind of beer could be out of vogue and then suddenly have an uptake by urban young professionals. It can also help them match their sports marketing to actual consumption – i.e. if they sponsor a certain sports team, how does that play out in the actual bars where the fans are watching? For microbreweries, this can help them determine the brand identity of new beers that they produce. So all in all I do see some marketing value here for the breweries.
Thanks for the article! I think the SteadyServ technology is fascinating. I agree with your view that penetration of the technology is critical for it to have any impact on overall supply chain efficiency (though it the individual bars may benefit from the analytics on their own, if nothing else). I agree that testing with retailers is important, but I challenge you that earning the buy-in of the breweries is even more important. I believe this because each brewery serves a substantial amount of retailers; should the breweries believe in the technology and choose to advocate that their retailers use it, the technology could gain massive network effects very quickly.
First, I had no idea that this was a current problem with beer inventory (thank you for sharing!). Second, from reading your article, it seems the major selling point is the tracking of inventory and its accuracy; however, I think an equally if not more important selling point of the technology is optimizing the types of beer to sell and recommendations of new beer. I’d imagine that the company is now collecting data of beer consumption and can help bars and restaurants compete by offering the “right” beer to their consumers. I’d be curious if this is a robust part of the business and the success rate SteadyServ is in improving a bars beer sales.
Thanks for the piece, and great way to tie it in with a common reference point (i.e. our previous participation in the beer game)! I agree with most of what you said, but tend to also think Ross’ comment holds some value that might not have been captured in your article, namely, that the breweries themselves could benefit from this. Who is drinking what? And where? How does the behavior of our customers typically correlate with the behavior of customers in other regions of the country? These are all valuable insights that the business savvy brewery owner (maybe you, one day?) might be able to profit from and drive a superior business model to competition.
Interesting article; thanks, Alex! I agree with the comment above that this is something that should be pushed down the supply chain. As we observed in the beer simulation, the bullwhip effect is most pronounced at the brewers’ level, and therefore significant value must accrue to them through the implementation of this system.
Thanks for the post Alex! While I find the $30 monthly cost to access reasonable, I worry that the $4 per keg for the hardware is too high. It’s possible this could represent a 10% price increase for some bar owners depending on the type of beer, which could make uptake more challenging especially as it will take a few weeks, and a lot of kegs, in order to see the value here. Perhaps SteadyServ could start with some sort of trial period to engage usage, or charge a higher subscription fee instead of the fee / keg? If not, SteadyServ should market heavily the re-usability of its kegs and attempt to convince bar owners to spread this cost over a longer period in their minds.
Very interesting. I wonder who really benefits the most from this and therefore who should pay for it. To me, it seems like the distributor and the manufacturer may save the most amount of money by having a much leaner in stock inventory and being better able to manage their production cycles. Therefore, if I were selling this product, I would consider going straight to the distributor and integrating into their options – the distributors could then push the technology down to the retailers to improve their operations too.
Alex – this was really interesting, thanks for the info! Based on my work with beverage manufacturers, I do actually think that they would both benefit from and be willing to pay for the data that comes from the installation of iKegs at various retailers. To manufacturers, data is king. They will often have to pay large sums of money to gain point-of-sale data regarding their products. If they could get access to real-time consumption data by location, the benefits to purchasing, production planning and distribution could be huge. Future applications might also involve testing out the effectiveness of promotional spend in the market (a huge cost for beverage companies). As mentioned above, this data would be most useful if the technology was adopted by many retailers. Manufacturers and distributors might be willing to split the cost (or pay for it entirely) if enough retailers showed interest. Retailers would have to be willing to share the data with their upstream partners (something, from my experience, that they can be quite reluctant to do).
Interesting read, thanks for the article. I think the value of the product is two folds, one is direct labor cost saving for end customer (bars), one is potential savings from better supply chain planning. Data collected by these iKegs can be used for customers to better project demand to order. If this data is shared with distributers and manufacturers from more and more end users, it will be valuable in inventory cost reduction. However, it’s interesting to know who actually owns the data, because i could see manufacturers paying for the end user data to better understand their own distribution chain, their end customers (bars) by region, and so on.
Loved this post. At the risk of sounding naive, couldn’t most of this problem be solved in a very analog way? What if kegs had a transparent strip like many medicine bottles so you could see inventory levels at a glance? This would not allow retailers, distributors and manufacturers to share data but it would also be much cheaper for the retailers to manage.
Alex, thanks so much for sharing this cool technology! I agree with you that there are still some additional steps to take in order for this to be truly beneficial to the breweries/distributors. As we saw in the Beer Challenge exercise, the amount of time it takes for retailers to request and then to receive new inventory can be problematic. My concern is the large volume of sales that a more popular or more frequented retailer would experience on a particular night. Even if a large/popular retailer is able to see from this app that a keg is low, the retailer could potentially sell out of numerous kegs of the same variety in any particular night, and given the lag time in re-ordering, there could still be supply/inventory issues. I see this technology being most beneficial in either inventory control for smaller retailers or for large retailers to aggregate more specific data of demand points rather than inventory control/re-ordering.
I really enjoyed this article! I think a vast majority of the problems can be solved if iKeg achieves a level of scale. Unlike the Barilla and IKEA case, Steadyserv is not a participant of a traditional beer supply chain and will have to work hard to get the whole supply chain to get involved in their project. But when Steadyserv gets a hold of a significant number bar owners, it can leverage its position in the market to add brewers and other players in the supply chain and ultimately will be able to manage information of the entire supply chain. I hate it when I go to a bar and the bartender says that they are out of the beer I was craving for. It happens quite often. I hope this solves the problem.