---
title: "Mu Wei of Zhoupu Data: Driving Unified Warehouse and Distribution Efficiency with Technology"
description: "At the 2018 China FMCG City Distribution Logistics Conference hosted by New Distribution on October 24, 2018, Dr. Mu Wei, CTO of Zhoupu Data, delivered a keynote speech on using technology to improve the efficiency of unified warehousing and distribution. He emphasized that without technology, unified warehousing and distribution cannot achieve cost savings, and introduced Zhoupu Data's full-process system and data-driven approach."
author: "慕巍"
publisher: "New Distribution"
email: "zhaobo258@gmail.com"
telephone: "+8615854817671"
published: "2018-11-07"
language: "en"
canonical: "https://xinjignxiao.com/en/articles/mu-wei-of-zhoupu-data-driving-unified-warehouse-and-distribution-efficie-fd80e198/"
markdown: "https://xinjignxiao.com/en/articles/mu-wei-of-zhoupu-data-driving-unified-warehouse-and-distribution-efficie-fd80e198.md"
original_source: "https://mp.weixin.qq.com/s/_esib-19Bs0sbHCeERKZcw"
translation: "https://xinjignxiao.com/zh/articles/%E8%88%9F%E8%B0%B1%E6%95%B0%E6%8D%AE%E6%85%95%E5%B7%8D-%E4%BB%A5%E6%8A%80%E6%9C%AF%E9%A9%B1%E5%8A%A8%E7%BB%9F%E4%BB%93%E7%BB%9F%E9%85%8D%E6%95%88%E7%8E%87%E6%8F%90%E5%8D%87-fd80e198.md"
attribution: "New Distribution — https://xinjignxiao.com/en/articles/mu-wei-of-zhoupu-data-driving-unified-warehouse-and-distribution-efficie-fd80e198/"
usage_policy: "https://xinjignxiao.com/ai-policy.txt"
---

# Mu Wei of Zhoupu Data: Driving Unified Warehouse and Distribution Efficiency with Technology

> At the 2018 China FMCG City Distribution Logistics Conference hosted by New Distribution on October 24, 2018, Dr. Mu Wei, CTO of Zhoupu Data, delivered a keynote speech on using technology to improve the efficiency of unified warehousing and distribution. He emphasized that without technology, unified warehousing and distribution cannot achieve cost savings, and introduced Zhoupu Data's full-process system and data-driven approach.

At the 2018 China FMCG City Distribution Logistics Conference hosted by New Distribution on October 24, 2018, Dr. Mu Wei, CTO of Zhoupu Data, delivered a keynote speech on using technology to improve the efficiency of unified warehousing and distribution.
First, I would like to thank New Distribution for the invitation to share with you a technical perspective.
When I first received the invitation from Mr. Zhao, I saw the list of distinguished speakers, and I was the only technical person. The issues I address may differ from others' starting points, but I thought this could enrich today's presentations. We will discuss the current state of city distribution logistics and unified warehousing and distribution from multiple angles.
First, how big can a distributor's business grow? What limits the cost of a distributor's business? The most important point is that when you scale up, especially in the FMCG sector, you find that your revenue growth trend begins to decline. Simply because many customers have already been served. However, your cost scale rises rapidly.
Profit maximization is not revenue maximization. If you earn more, you also spend more. What can we do? Can we implement unified warehousing and distribution? But this is not easy. A simple question: have distributors ever tried joint unified warehousing and distribution? If there are ten distributors, each with three warehouse staff, would the total be more or less than 30? I believe experienced people know that if you simply do unified warehousing and distribution, the total will be more than 30. Why? Because when your warehousing costs drop, management and distribution costs rise faster, so your costs do not actually decline. Here comes the paradox: **Without scale, you cannot achieve this cost-saving model, but with scale, complexity makes costs uncontrollable.** So from the start, we firmly believed that without technology, unified warehousing and distribution cannot be operated.
When we founded Zhoupu Data, we held two beliefs. First, to build a full-process system covering all scenarios of a distributor's business. Second, to use technology as a core driver. Moreover, we not only build systems; we also build our own warehouses and have our own professional distribution team and fleet. Only by personally iterating on efficiency can we provide distributors with the best and most optimal solutions.
This is a system chain that covers all business processes of distributors. As you can see, starting from the left, our system serves distributors' inventory and sales management, helping you connect people and stores, facilitating business flow, and providing cost and financial accounting, as well as future customer service and sales analysis. In the middle, we have order management systems for unified warehousing and distribution, and our warehouse management system provides all value-added services within the warehouse. On the right, we also have an independent driver operations management system. Zhoupu's unified warehousing and distribution and city distribution not only deliver goods but also handle shelf restocking, customer service maintenance, and account settlement ahead of distributors. We will help you build a good reputation and brand.
Why do we build a complete system ourselves rather than purchasing components from the market? The most important reason is to create our own operational support system. We believe that digital integration and optimization must be full-chain. If we collect all digital points from start to finish, optimization becomes global and maximizes its power, combined with a big data platform.
We have both breadth and depth in technology. For example, to make the system more flexible for distributors, we adopt a microservices architecture; to ensure data security, we use multi-tenant, multi-redundant storage architectures. But here I want to emphasize how we test our cloud warehouse system and why it is reliable. We establish a virtual environment where all data running is real data. If a self-operated order appears, the virtual environment generates three orders; if there are three warehouses' data, we consolidate into one warehouse. This way, before release, all systems undergo higher-intensity and richer business validation. Only when these validations pass can we push to production. This is not just about testing; it is also about efficiency iteration and strategy optimization. In the virtual environment, we can precisely tell you how good our optimization is and how much cost it saves. This is what an automated virtual environment can tell us. And it is automated, requiring minimal human intervention.
Data—I have always separated technology and data. Why? Technology is often discussed; simply put, technology is electronification. Common technology means providing a system: previously paper and pen, now computers; previously PC, now mobile; warehouses operate paperless and fully mobile. The benefit of technology is that it makes it easier for people to decide things, but decisions are still made by humans. What is data? When I collect all information and present it to people to decide what to do, that is the first level of data. The second level is when I collect it and tell you how to do better; data manages people, data can decide for people, data can control processes. That is the power of data. If people decide data, it is just replacing paper with a phone. But data has a broader, more accurate, and more precise view than humans. So I always emphasize digging it out. This is the big data platform we have built.
It is hard to explain each technical component of the big data platform here. As I mentioned, I have built big data teams at Google and Facebook and led AI teams. I can confidently say that even from a Silicon Valley high-tech perspective, this is a first-class data platform. For example, if you currently generate reports or your business processes rely on manual data sorting, or if you need technicians to pull data from databases and copy to Excel to generate reports, if you still do this, you can talk to me. We can achieve integrated data processing across the entire system and chain, real-time analysis and display, and instant generation of various charts and reports without switching between systems. All task flows are automatically scheduled. Once people decide the task and requirements, all tasks are automatically arranged. Moreover, this platform can integrate with most open-source AI learning packages. This demonstrates our technical strength.
As for our product managers, the requirement is that they must be data analysts. That means your product manager must not only understand customer groups and find needs but also read customer needs from data and extract information from data.
One important point: if you only have numbers, they are meaningless. You must bring value to customers. Truly useful data that generates value is good data. For example, this table is a regional sales analysis done by Cloud Steward for a provincial distributor. Our data must generate customer value to be meaningful.
Having said so much about concepts, I want to share some specific cases and examples that can help distributors.
The first efficiency improvement example is about cloud warehouse strategy optimization. Strategy means when you have many products in a warehouse—for instance, nearly 10,000 SKUs—where do you place what? What to pick, how to take down, how to arrange? These are strategy issues. During operations, we need complex scheduling. In early systems, these schedules relied on warehouse managers' personal experience. Their experience was coded into the system, but it was still people deciding the system. Our new system has achieved engine-based dynamic scheduling. All systems and strategies are dynamic, can be combined and searched, and can be virtually simulated. As mentioned, we have a virtual simulation platform where different strategy combinations can be precisely displayed, and in a visual way. Whether warehouse staff or developers see direct effects, they can intuitively follow and act, which also brings efficiency gains.
Another efficiency example is about sorting and rechecking. When you start unified warehousing and distribution or joint operations, you find that the number of SKUs rises sharply. What to do? Sorting and rechecking. If you rely on a technical system but human decisions, you need full-item sorting and rechecking. Errors are handled offline, and you don't know if the recheck is correct. After integrating a digital system, we first solve the problem of automatically collecting the correctness of all goods upon delivery at the driver end: whether the batch is correct, expiration date, quantity, and appearance. These determine the work results of sorting and rechecking staff. We not only determine their results but also precisely analyze each staff member's efficiency at different times and intensities, and different items yield different results. With these models, we can intelligently assign more reasonable tasks: which tasks need rechecking and which do not. We ensure a driver delivery rate of 99.9%, and the workload of sorting and rechecking drops by over 60%.
Besides efficiency, another pain point of unified warehousing and distribution is cost. Efficiency can be improved when everyone works together and there are more goods, but how do you calculate costs? I wonder if distributors have tried joint warehouses. How do you allocate costs? Some items are easy to handle, some are not. Not to mention break-bulk and repacking. If we go out for delivery together, delivering to ten stores, some near, some far, some need shelf restocking, some do not—how do you allocate costs? It is unclear. When it is unclear, you rely on trust, but that business cannot last long. In our early business discussions about billing for unified warehousing and distribution and cloud warehousing, we encountered problems. When we discussed, we listed items sparsely: one item had 15 to 16 details, each negotiated separately. Is the business clear? Not really. What is our more advanced model? We can precisely track the time and cost of each job on each order, and we can calculate clearly.
We can then summarize for each item and each distributor entering the warehouse whether the accounting is profitable or not. This gives guidance to our business team, certainly, because it shows the cost distribution for different items. More importantly, it guides our operations team on where to optimize efficiency. That is crucial. But I still say this is a relatively old model. Why? Because when I talk to distributor friends, we sign a contract, tell you we charge a 3% fee, and the rest is Zhoupu's business. If Zhoupu does well, it earns more; if not, it earns less. We attract more merchants, use better smart hardware, and optimize better—you know nothing about it.
What else can we do? We will launch a machine learning-based cost prediction model. That is, you tell me some basics: what items you represent, in what region and scale, and what customer service requirements you have. Our system model will predict your cost model and which billing method is more reasonable. Why do this? An important point: suppose your self-operated warehousing and distribution costs 4% (4% of cost). Zhoupu's professional operations can achieve 2.5%. I will precisely show you how that 2.5% arises, where each order occurs. I hope you find the unified warehousing and distribution statement simpler than your mobile phone bill. If I can do that, do I still need to discuss the 15-item cost table? Do I need to go item by item? No. I only need to talk about this: you do 4%, Zhoupu does 2.5%. How much profit do you share with Zhoupu? If Zhoupu improves city distribution in the future, or makes profits better, introduces hardware, brings in other merchants, and rationally matches items among different merchants to make unified warehousing and distribution more efficient, you benefit because Zhoupu only charges a fixed profit fee. This is not a baseless claim; I actually did this in Silicon Valley.
I want to say that unified warehousing and distribution not only saves money for distributors but, most importantly, frees up your energy. It allows you to spend more time, as Mr. Zhao said at the beginning, on expanding sales—your most focused and professional area—expanding sales channels, product lines, and customer groups.
Finally, I must emphasize that Zhoupu's cloud warehouse system not only serves unified warehousing and distribution. It can also serve a single distributor's self-operated warehouse model. If you are a large distributor, Zhoupu's Cloud Steward + Cloud Warehouse system model can perfectly solve your business needs. The efficiency and cost optimization examples I gave earlier are also of great help to this model.
To summarize, as Zhoupu Data, we adhere to two principles: one is full data flow, using data to drive continuous efficiency iteration; we not only do cloud warehousing but also full-process ERP systems. We firmly believe that only by connecting data can data iteration truly improve warehouse efficiency, allowing our service model for distributors to continue.
Zhoupu Data is still a very young startup. We sincerely hope to cooperate with everyone, learn together, and progress together, striving to do well in city distribution logistics and unified warehousing and distribution. Thank you!
Click **Read Original** to see more highlights of the 2018 China FMCG City Distribution Logistics Conference...
-END-


---

## Copyright and AI use

This article is sourced from New Distribution. Search, quotation, summarization, and model training are permitted, but every use must credit New Distribution and retain the canonical source URL.

Contact: zhaobo258@gmail.com · +86 158 5481 7671