The science behind the new FBA capacity management system

It’s September. You are a toy salesman who uses the fulfillment of Amazon (FBA) and most of your sales come just before the holidays. However, you overpowered this summer, breaking down the health of your inventory, resulting in lower remaining capacity when you were near the holiday season. In the meantime, you’ve just launched a new line of robot mounting sets that are becoming popular with customers.

How can you secure access to additional space in Amazon fulfillment centers to set yourself up for a good black friday and cyber Monday using FBA? And how does Amazon control capacity without the superstructure of space, potential imposing costs that do not benefit the customers, as well as without the support of space and restriction of product selection, available and fast delivery?

The journey to a package

Brands and dealers seeking to utilize Amazon’s science and technology-driven fulfillment and transport networks are about the fulfillment of Amazon (FBA) to outsource order fulfillment and customer service to Amazon.

The answer lies in an innovative new FBA capacity management system, powered by Amazon’s Supply Chain Optimization Technologies (SCOT) and Selling Partners Services (SPS) team. The new system gives most sellers greater capacity limits and also greater control to achieve additional capacity when they need it.

FBA is an optional service for sellers to outsource order fulfillment and customer service to Amazon. The sale of launch in 2006, the FBA has quickly expanded to sell partners ranging from family -owned snack manufacturers to multinational high -tech companies.

The FBA is an increasingly popular choice for both brands and dealers seeking to exploit Amazon’s science and technology-driven fulfillment and transport networks. The FBA provides sellers with significant cost savings and faster delivery speeds that together provide great value to customers and sellers. The accelerated delivery speed has led to significant increases in customers’ demand for products, while the wave in Dequest, especially during peak events such as Prime Day, Black Friday, Cyber ​​Monday and Holidays, has created so far so far for FBA capacity.

Amazon needed to find a way to match demand with accessible capacity while building additional capacity to meet the needs of the seller.

Capacity management challenge

Amazon’s prognosis capacity – driven by a combination of machine learning, simulation and optimization modeling – is well documented. Researchers have tackled research questions, such as how to manage fixtures for peak events, and how to best distribute products across Amazon fulfillment network for the fastest delivery.

Animation shows a map of the United States and each of the 8 individual regions that are the result of Amazon's regionalization

Related content

The pandemic turbo-charged retail growth teams for researchers at Amazon threw a path forward to deal with the scale.

But when it comes to managing capacity in the FBA program, even the Amazon’s prediction mechanic has limitations. The FBA serves a wide breadth of independent sellers who own their inventory and make independent decisions about what products to transport, how much inventory should be and what prices to offer.

“There is only so much we can really understand what the seller’s business plans are,” said Garrett van Ryzin, distinguished scientist at Scot. “It’s basically limiting to relying on our predictions only.”

In addition, Amazon does not have full visibility in these seller’s business needs or decision -making process. This became more recording as the seller’s use of FBA network capacity grew dramatically within a few years.

“Most of the time, our predictable models assign adequate capacity to each seller. Still, in some situations, we observed that sellers did not have sufficient capacity, which affects their ability to pursue certain growth options. Capacity for product that they are confident, even though we did not predict that they needed more distance to the data that Amazon had,” -Team.

To control the capacity of the FBA, Amazon initially created a system that scores sellers according to how well they manage their warehouse. A few weeks in advance of each allocation period, sellers are told how much capacity they will in the next three months – a number based on the seller’s planned sales, and earlier stock services (all equal, this number is rising for sellers using effectively).

Our original system worked well for capacity control. However, we needed a new approach that could handle the seller’s current and future plans, such as a new product introduction, marketing and advertising campaigns or other information that only sellers knew about.

“Our original system worked well for capacity management. However, we needed a new approach that could handle the seller’s current and future plans, such as a new product introduction, marketing and advertising campaigns or other information that only sellers knew about,” said Özalp Özer, who led the FBA Science Team in his capacity as director of research science.

The allocation of space based on previous performance and Amazon’s forecasts meant that sellers who had recently underpinned in terms of sale and layer efficiency could end up having lower remaining capacity to launch potential popular new products or campaigns. Voters could submit extraordinary requests for additional capacity; Amazon received many such requests every quarter, and each one had to go through manually.

To tackle this customer pain point, the Amazon FBA team created a new way for sellers to give a single, objective measure of how much value they could create if they were given additional capacity. The new tool used financial principles to strengthen salespeople to communicate credible need for additional capacity.

MMID-GIF-2.GIF

Related content

Why multimodal identification is an important step in automation of object identification in the Amazon scale.

Here’s how it works: In each capacity distribution period, sellers can place requests to increase their capacity limit. These sellers submit a “reservation fee” – the maximum fee they would be a willingness to pay per. Cubic foot to reserve additional capacity along with the desired maximum love of additional capacity.

Requests are supplements objective, starting with the highest reservation fee per Cubic foot unit is all available capacity under this program awarded. When additional capacity is a subsidy, sellers pay no more than their selected maximum reservation fee. In addition, sellers receive performance credit from the sale, they generate the extra capacity – up to 100% of the reservation fee.

Another approval of capacity management

The FBA team designed the tool to optimize the value of customers using additional capacity effectively – and not to maximize the proceeds for Amazon. In fact, sellers who win more space can earn credits from the extra sales they generate, and these credits can equalize the entered to the reservation fee.

“The mechanism is designed to reward sellers who use their extra capacity productively to generate sales,” explained Van Ryzin. “We use financial market design principles to manage supply chain capacity, and this is the first time we have really done it in this scale of the FBA.”

This new tool is inspired by previous financial research in the sale of securities, where payments are based on the value generated from the asset after it is assigned, but the details constituted unique challenges.

“You also need something that is simple for sellers and has a level of justice in his decisions,” Tolga Seyhan explained a main scientist at the FBA team.

“We make sure that if you arrive early and you paid a high fee, you will ultimataly get the same fee as someone who paid less in a later one,” said Alexandre Belloni, an Amazon Learn from Duke University’s Fuqua School of Business, who has led projects. “We will reward early requests that help with our capacity planning and give sellers a better opportunity to utilize the extra capacity.

An Amazon -Employee is seen loading boxes in a delivery vehicle

Related content

How customer order and network density optimize (Condor) has led to improved delivery routes.

Take a theoretical example: say that a seller has a Halloween store on Amazon and needs space for a new costume. They will have 5,000 additional cubic feet (“dice”) on top of their usual allocation of 10,000 ft3. They place a reservation fee of $ 6 per Cube in an early round of the quarter of a total potential cost of $ 30,000. Their request is a subsidy, but then the low reservation fee falls, which in subsequent rounds falls to $ 4 per day. Cube. The new tool reduces the Halloween store’s reservation fee to $ 4 per day. Cube that matches the lowest supplements for reservation fees in the later round. Therefore, the seller’s total reserve fee would be a maximum of $ 20,000.

As is often the case with these capacity reservations, the seller in this theoretical example correctly predicts that their new costume voice justifies the space. The additional space generals $ 300,000 in step -by -step sales. The Amazon Awards have 15% performance credit on the step -through sale, up to a maximum of $ 20,000 total commitment. In this case, the seller has read $ 45,000 performance credit, which has fully offset the entire reservation fee. The seller ends up paying nothing for the extra space and because they used the space effectively, their capacity distribution will also probably rise in the future.

The system was successfully launched for a small number of sellers by the beginning of 2022.

Amazon Science Forecasting Timeline.gif

Related content

The story of a decade plus long journey towards a unified forecast model.

“When we launched the first experience, we were glad to see that it was using this rationally and that it helped address information asymmetry. Sellers had enough information to certainly predict both the need and the price they would be willing to pay for additional space. Most of the sellers got their reservation fee back and ended up paying nothing for the extra space,” tireless.

“The productivity of the space assigned through this mechanism is higher than what we would get the bear allocation, and it benefits everyone: Amazon, our salespeople and our customers,” said Özer. “Building this new system required our researchers to use advanced tools from optimization, simulation and finances and work closely with product conductors and engineers. Together we are innovating on behalf of our customers and sellers around the world.”

Leave a Comment