Tamago-ya (meaning “Egg-shop”) was founded by Isatsugu Sugahara in 1975, as a small mom-and-pop bento (“boxed lunch”) shop. Under the entrepreneurial leadership of Chairman Isatsugu and his son President Yuichiro Sugahara, the company grew rapidly to become the largest lunch box producer in Tokyo, Japan. As of 2019, it produced and delivered 70,000 boxes a day, employed 650 people, and grew revenue to US$80 million. Tamago-ya produced and delivered high-quality lunch boxes at low price of US$4 to office workers in the Tokyo Metropolitan area.
The success of Tamago-ya depended on its unique hyper-efficient supply chain management. It received orders at 9 a.m. until 10:30 a.m., and delivered by noon. Thus, the delivery lead-time was very short, and on-time delivery was a challenge in busy Tokyo. However, Tamago-ya hardly ever missed a delivery deadline. While demand was large and fluctuated from day to day ranging between 60,000 and 75,000, Tamago-ya’s average “loss ratio” (the disposal ratio due to over-production or returns) was only 0.06 percent. This translated to only 42 lunch boxes left over from 70,000 total productions a day.
The combination of strong leadership with customer-focus orientation, empowered workers to achieve continuous improvement, and the unique economic and cultural environment of Tokyo led Tamago-ya to pull off its “magic.”
Bento – the Lunch Box
Tamago-ya offered only one menu per day, but it changes everyday. Each lunch box contained more than six items, most of which were made from organic and natural ingredients (Exhibit 1). The lunches are freshly made each morning and kept warm until delivered. Instead of disposable lunch boxes, Tamago-ya used reusable ones that could be used for up to one year. The overall cost of using reusable boxes was slightly higher than that of disposable boxes, since it took up to nine hours a day to wash and clean all the lunch boxes using specialised equipment and specially-treated water. But reusable boxes offered multiple benefits to Tamago-ya. First, they were more eco-friendly, either by saving trees or avoiding waste. Also, reusable boxes provided van drivers with more opportunities to talk to customers as they collected boxes after the lunches were finished. Valuable customer feedback could be obtained just after the meal. Last but not least, due to the pick-up, customers were obliged to consume the lunch by 1:30 p.m. in its best condiiton.
Exhibit 1
Van Drivers the Utility Players
Tamago-ya did not have salespeople, nor did it spend money on marketing. Instead, van drivers promoted sales when they delivered lunch boxes; they visited offices near their existing customers and introduced the company and its product. For customer retention, van drivers kept close communication with customers when they delivered lunches or picked up empty boxes.
All new customers needed to be interviewed by Tamago-ya and registered in its customer database before they placed daily orders. To reconcile the flexibility to customers and the commitment by Tamago-ya, it screened new customers both by their order size and their location – that is, from a delivery efficiency perspective. Tamago-ya did not accept all potential customers.
Demand Forecasts
Tamago-ya’s forecasting fully leveraged the information captured by van drivers. Every evening, Tamago-ya’s managers and van drivers in each region held a meeting to forecast the number of lunch boxes customers would order the next morning. The primary information source for forecasting was the van drivers’ daily reports. Every day, van drivers wrote a report including their own forecasts of the next day’s orders as well as customer feedback on today’s menu. When a van driver visited customers twice a day, he communicated with the person in charge in an informal and friendly manner, and listened to her evaluations of the day’s menu. The driver also asked her for an estimate of the number of orders for tomorrow. Customers talked to van drivers frankly: “Next week, many people will be away from the office attending a big trade show,” or “Some colleagues say beef steak was a little too salty.” Tamago-ya also used causal data, such as weather, day of the week and the month, and the menu, to forecast demand.
Production and Logistics
The key question about Tamago-ya was how it could produce and deliver an uncertain number of lunch boxes within a short delivery window to so many customers geographically dispersed in the heavily-trafficked Tokyo area. The answer can be summarised in the following directives:
• Dual-response production: First, build a stock of lunches up to a low-end forecast of the demand, and later, build more (if necessary) based on the up-to-date estimate as actual orders arrive. Tamago-ya counted on five key suppliers who were both nimble and flexible. Tamago-ya committed to the low-end forecasted quantity of ingredients on the previous day, and also carried an option to ask for more if necessary on the morning of production.
• Time-phased prepositioning of stocks: Divide the entire market into two regions by distance from the factory. Dispatch the first batch of vans early to the remotest region well before the order closing hour, with each van carrying an estimated quantity of lunch boxes. (This practice was often called “inventory on wheels.”) After order receipts were completed, dispatch the next batch carrying the exact amount of orders to the nearest region. Transfer stocks across vans to fill any demand-supply gap within and across regions. Use standby vans to adjust any remaining gaps. In fact, Tamago-ya applied this concept to three, instead of two, regions.
The company’s typical daily operational flow was:
• 17:00 – 18:00, the previous day: The total number of orders for the next day was estimated as an interval of low and high ends.
• 4:00 a.m. Tamago-ya received ingredients from 30 to 40 regular suppliers, and kitchen staff started cooking menu items.
• 6:30 a.m. Factory workers started assembling lunch boxes by putting the items together. Since the company had not yet received any orders, it produced lunch boxes based on the low-end forecast they made last evening.
• 8:30 a.m. The first batch of vans loaded the estimated number of lunch boxes for the farthest customers and departed for their designated routes. Tamago-ya has not yet received any actual orders, but production continued up to last night’s estimate.
• 9:00 a.m. Tamago-ya’s operators began taking orders for the day by fax, by phone, or through the Internet. Production continued based on the previous night’s estimates, but was gradually adjusted as actual orders were received.
• 10:00 a.m. – 10:30 a.m. The first batch of vans arrived at their first customer drop-off. The drivers parked their vans in front of the customers’ buildings and waited for a call from headquarters. As soon as headquarters (HQ) received an additional order from a customer, it called the driver for that route to inform him or her of the new order. Finally, the HQ gave a “go” signal to bring the lunch boxes to the specified office. Some customers received their lunch boxes only a couple of minutes after they placed an order.
• 10:25 a.m. The second batch of vans departed the factory for middle-distance customers just before 10:30 a.m. At this point, Tamago-ya had received almost all of the actual orders for the day. They loaded the lunch boxes specified by actual orders for those who had ordered, plus an estimated number of lunch boxes for the customers who had not yet placed an order. Production continued in parallel, with the final number of orders constantly being adjusted. Tamago-ya finally stopped taking orders for the day sometime after 10:30 a.m. (e.g., 10:32 a.m.).
• 11:00 a.m. The third batch of vans loaded the final lunch boxes to fulfill actual orders and departed for short-distance customers. If the first or second batches over-carried lunch boxes, the third batch carried fewer lunch boxes according to the difference. If the previous batches under-carried boxes, the third batch carried more lunch boxes accordingly.
• 11:00 a.m. – 12:00 a.m. If actual orders were different from the estimate, the dispatch centre instructed van drivers via cell phones to move lunch boxes from one van to another, so that the vans had enough boxes and traveled along optimised routes. This dynamic load transfer was done at any place where vans could be parked together (e.g., a public parking lot). Locations and load transfer were determined between van drivers and an area manager. Decisions were based on their experience and intuition; no IT was used.
• 11:20 a.m. Tamago-ya also had 30 extra vans called “adjusters.” Often than not, Tamago-ya could not fill the difference between its forecast and actual orders. Also, accidents happened. The adjusters were dispatched at the last minute when the vans on the field could not adjust the number of lunch boxes on their own.
Portfolio of Supplies – Fixed and Options
Suppliers were a critical part of Tamago-ya’s core competency. Tamago-ya created a portfolio of two supply sources – fixed and options. The following example illustrates how it worked. On a certain day, Tamago-ya’s initial forecast was 67,000 with its low-end forecast 58,000. These two numbers were shared with all supply chain partners. Tamago-ya used its low-end forecast 58,000 to order ingredients from its regular suppliers on the previous evening. This is the fixed portion of its supply. After it started receiving orders, Tamago-ya updated its database every 15 minutes and shared it with all its partners including its five agile suppliers. These agile suppliers, strategically located near Tamago-ya, were involved in the last-hour demand fulfillment process, serving as options. They brought ingredients to Tamago-ya every 15 minutes in response to the updated orders. Tamago-ya cooked the ingredients to produce menu items, which took another 15 minutes. Once food items were ready, Tamago-ya would assemble 500 lunch boxes every minute. By the time orders were closed at 10:30 a.m., the total number of orders turned out to be 66,880 due to the heavy rainfall in the morning, and the gap of 8,880 between the low-end forecast and actual orders was completely filled when the last batch of vans left the factory at 11 a.m. Tamago-ya had its production facilities close to its suppliers, not to its customers, so that it had more flexibility in procurement. It also maintained relatively small warehouses, since it believed larger warehouses tended to lead to larger inventories. Tamago-ya only kept condiments (e.g., soy sauce) in stock for one week. All other fresh ingredients were delivered on demand, and were discarded if left unused for the day.
Tamago-ya’s procurement team spent Monday and Tuesday meeting existing suppliers. They reviewed the performance and discussed future plans in pursuit of continuous improvement. On Wednesday, Thursday and Friday, however, they met new supplier candidates who could bring in new values to the supply chain.
Growth, Social Responsibility and the Future
President Sugahara believed that Tamago-ya might already have reached the optimal size at 70,000 boxes per day. If they grew further, he was not sure how the company would source the quality ingredients (e.g., fresh mackerel) without breaking the current sourcing paradigm. He did not see a compelling reason for growing bigger, and preferred to keep Tamago-ya as a medium-sized private company wholly owned by the Sugahara family. In addition, he believed that the company should not be too greedy. Tamago-ya targeted 5 percent net income, with any excess profit reinvested in equipment, or used for better quality of food and higher employee compensations. It recommended the same attitude to its suppliers. In the same vein of social responsibility, Tamago-ya promoted healthy food (e.g., broiled fish and vegetables) ahead of popular food (e.g., hamburgers and fried meat). Chairman Sugahara said, “Of course, the lunch prepared by the wife is the best, and ours is the second.” When asked, “Why don’t you expand to other regions?” Sugahara replied, “It is not easy to replicate the Tamago-ya model elsewhere.” And he added, “We are happy and proud to be in Tokyo.”
About the Author
Seungjin Whang is Jagdeep and Roshni Singh Professor of Operations, Information and Technology at Stanford Business School. His research interest is in supply chain management and the economics of information systems. Outside, he serves on the advisory boards at large manufacturing companies and a venture capital firm.
