"Push" And "Pull" Combine To Create Efficient Supply Chain

Push production not only shortens the delivery cycle, but also makes full use of production capacity. For example, when you are not busy, the waiter can help pick beans, so that you do not have to prepare the production capacity according to the peak demand, and the overall cost is low. Of course, the risk of push production is what to do if there is no demand or if it does not come in time. So, "push" to a certain level, it must become "pull", that is, the order from customers to pull to control inventory backlog risk. For example, if no one orders, the small restaurant does not stir fry the beans into dried beans or the small wonton. Push production has Stock Risk, but simple pull production is also not perfect: orders have, although there is no risk of demand, but because goods are urgent, supply risk increases. Cheng Xiaohua, an expert in inventory control, said, "MRP is nervous: for example, the delivery period of a customer is three weeks, and your components need five weeks to arrive. The processing and assembling time is not yet calculated. The rush cost is always higher.

Therefore, despite the advantages of pull driving, few supply chains are pure "pull"; almost all supply chains are combined with "push" and "pull". Where should the combination of "push" and "pull" be located? Recently read the article of Marshall Fisher of Walton business school, and found the answer: it can be predicted to use "push", and the part that can not be predicted is "pulled" by Fisher. The higher the customization level, the lower the prediction accuracy, the farther the "push" and "pull" points are from the final customers. For example, in a small number of equipment industries, product configuration is diversified. Manufacturers rely mainly on customer orders to drive production and assembly. The combination of "push" and "pull" is in the procurement of parts and components. For general parts, manufacturers are driven by predictive drive suppliers ("push"), waiting for customer orders to come, and then assembling the final products. For low grade parts, manufacturers often wait for orders after customers, and then push the "pull" point of integration farther away from customers. On the contrary, the higher the degree of standardization, the higher the accuracy of prediction, and the combination of "push" and "pull" is closer to the final consumption point. For example, the consumption of tap water is relatively accurate. The combination of "push" and "pull" is closer to the end consumers, that is, when we open the faucet.

In addition, the higher the requirement of the product's timeliness, the closer the "push" and "pull" point is from the consumption point. For example, the key spare parts of large equipment will be lost when they are shut down. Supply chain The combination of "push" and "pull" is near the customer's production facilities. For example, when I manage a global spare part plan for a semiconductor equipment manufacturer, the spare parts inventory of US $more than 100000000 has more than 20 databases built in every major area of global chip manufacturing, and more than 80 consignment stock points are directly built on the chip production site. So once the customer stops to wait, more than 95% of the spare parts will be delivered to the scene within four hours. There are similar indicators in the aviation industry. Once the aircraft is not in place, it will not be able to take off. The delivery time of spare parts will also be measured in hours. This violates the accuracy requirement of demand forecasting, because the demand for non expendable spare parts is generally hard to predict, even when you can roughly predict the average demand, but when exactly do you need it? The cost of doing so is high inventory level and low inventory turnover. For example, in many large scale equipment industries, the turnover rate of spare parts is always one or two times, which means that the average turnover of spare parts is six months to a year. The defense industry is even lower, for example, the spare parts turnover rate of the US air force is even less than 1.

In the field of supply chain, many best practices are related to the effective positioning of "push" and "pull". For example, the direct selling mode of DELL is a typical "pull" supply chain at the finished product level, and the order is not assembled. However, at the component stage, it is a typical "push" type supply chain, because these raw materials are like the snap beans of a small restaurant, with high versatility and predictable usage. HP The famous "Postponement" is also the same: the common parts of products are well predicted, and the "push" production is used to achieve economies of scale and reduce costs. The differential part is "pull" production to reduce the risk of inventory changes caused by demand changes. The success of direct selling mode and postponement strategy lies in the perfect combination of "push" and "pull". {page_break}

The combination of "push" and "pull" is not static. Even in the same industry or company, this combination may change with the life cycle of the product. For example, when DELL just started the direct selling mode, the computer is an innovative product, with many types of machines, a fast price reduction for finished products, low prediction accuracy at the finished product level, and high cost of inventory risk. The direct selling mode is the lowest total supply chain mode. However, over the years, computers have become popular products, and their configuration is becoming more standardized. The accuracy of prediction is getting higher and higher. The best combination of "push" and "pull" moves from the manufacturer to the retailer, that is, the producers are expected to mass produce, bulk supply to retailers, and become "pull" at the zero sale outlets. This way maximally benefits the scale, so the overall supply chain costs the lowest. On the contrary, DELL's direct selling mode is one-piece production and one-piece delivery. When the risk cost of computer inventory is significantly reduced, the high operating cost can no longer be offset from the cost of inventory risk cost, and the total cost of direct selling mode is no longer the lowest. One of the fundamental reasons why DELL has been on the decline for several years is that it can not effectively transform the supply chain and achieve the best combination of "push" and "pull". Of course, there are many reasons behind this.

The appropriate combination of "push" and "pull" will effectively balance the response speed, cost and service level of the supply chain. On the contrary, improper selection of "pull" and "pull" points will cause many problems and increase the total cost of the supply chain. For example, in the fashion industry's common order mode, the combination of "push" and "pull" is at the distributor's office. The brand manufacturer asks the distributor to order about two quarters ahead of time, and the entire supply chain is driven by the dealer's order. The problem here is "pull" production for the brand, and the "push" purchase for the dealer. That is, their orders depend on the forecast. For example, in the big summer, those dealers, especially small shopkeepers, can predict what will happen in winter and how much they can sell. So this prediction is no doubt gambling. The result of gambling is either shortage or backlog. The backlog seems to be the source of the dealer's capital and the dealer's capital. In fact, the brand business has to pay the bill together: the dealer's capital is backlog, there is no money for the next quarter's goods, where does the brand business come from? The backlog of inventory has to be sold at a discount rate of 80%; often discount, can the damage to the brand be small? That's why branding seems to be a profitable business, but so many brands are in trouble.

So where should the ideal combination of "pull" and "push" in the fashion industry be? brand Manufacturer. Because brand manufacturers design fashion, they have the deepest understanding of fashion, and can integrate different distributors' needs and predict higher accuracy. This is not difficult at all. The fashion industry, which is deeply suffering from ordering, knows it. The reason for the existence of orders is that dealers and brands are different business entities and have different interest demands. Orders are dealers' commitment to brands, and they are the result of game between the two sides. However, from the perspective of supply chain, the result of this game is sub optimal, resulting in the lose lose situation of brand dealers and distributors. There are two solutions, one is the vertical integration of brand players, and the other is to enter the retail sector. For example, what Lining did in the past two years is to strengthen the control of supply chain and better combine "push" and "pull". This is also a major reason for the success of Zara. Heavy assets direct operation is a great move. It requires not only to invest a lot of capital, but also to put forward new requirements for operation and management. The second is to straighten out the relationship with the channel, enhance cooperation, and improve the response speed of the supply chain by improving information flow. For example, sharing early retail sales data in a timely manner and adjusting the brand's production plan. This is shown in the study of Marshall Fisher that the prediction accuracy of the fashion industry is very low, but correcting the forecast on the basis of early sales data can greatly improve the prediction accuracy. Plan two requires higher supply chain management capabilities. When the company does not have such management capability, it chooses vertical integration and enters the mode of scheme 1.