---
title: "Digital Empowerment, Reconstructing People-Goods-Places: Rebuilding Deep Distribution"
description: "At the 10th China FMCG Innovation Conference, Gao Bo, CFO of C&S Paper, delivered a keynote on how digital tools can rebuild deep distribution. He identified three reasons why deep distribution is struggling in 2024—slowing retail growth, high store closure rates, and product abundance—and proposed solutions: selecting the right stores using AI models, choosing the right product categories, and targeting specific consumer segments."
author: "高波"
publisher: "New Distribution"
email: "zhaobo258@gmail.com"
telephone: "+8615854817671"
published: "2025-03-30"
language: "en"
canonical: "https://xinjignxiao.com/en/articles/digital-empowerment-reconstructing-people-goods-places-rebuilding-deep-d-508f80f0/"
markdown: "https://xinjignxiao.com/en/articles/digital-empowerment-reconstructing-people-goods-places-rebuilding-deep-d-508f80f0.md"
original_source: "https://mp.weixin.qq.com/s/64EGniGjjmLgJOy7EneMRQ"
translation: "https://xinjignxiao.com/zh/articles/%E6%95%B0%E5%AD%97%E8%B5%8B%E8%83%BD-%E9%87%8D%E6%9E%84%E4%BA%BA%E8%B4%A7%E5%9C%BA-%E5%86%8D%E9%80%A0%E6%B7%B1%E5%BA%A6%E5%88%86%E9%94%80-508f80f0.md"
attribution: "New Distribution — https://xinjignxiao.com/en/articles/digital-empowerment-reconstructing-people-goods-places-rebuilding-deep-d-508f80f0/"
usage_policy: "https://xinjignxiao.com/ai-policy.txt"
---

# Digital Empowerment, Reconstructing People-Goods-Places: Rebuilding Deep Distribution

> At the 10th China FMCG Innovation Conference, Gao Bo, CFO of C&S Paper, delivered a keynote on how digital tools can rebuild deep distribution. He identified three reasons why deep distribution is struggling in 2024—slowing retail growth, high store closure rates, and product abundance—and proposed solutions: selecting the right stores using AI models, choosing the right product categories, and targeting specific consumer segments.

From March 17 to 19, the 10th China FMCG Innovation Conference, the 4th China FMCG Hard Discount Conference, and the 4th China FMCG Distributor Conference, under the theme "New Order · Symbiosis," grandly opened in Chengdu.
Gao Bo, CFO of C&S Paper, delivered a keynote titled "Digital Empowerment, Reconstructing People-Goods-Places: Rebuilding Deep Distribution." Below is an edited excerpt of his speech, compiled for our readers.
**Why is deep distribution difficult today?**
In 2024, increasing volume is hard, increasing profit is harder, and deep distribution is also very difficult.
In the past, all profit growth came from deep distribution. Deep distribution required entering over 300 supermarkets and over 7 million retail outlets offline. But by 2024, it seems to have stopped working. Why?
I believe there are three reasons.
The first reason is total retail sales of consumer goods. In 2024, total retail sales were 48.24 trillion yuan. This huge figure indicates that business still exists. However, looking at growth, the 2024 growth rate was 3.5%, lower than GDP growth of 5%, dragging down GDP.
**The economic situation is indeed difficult, but not all sectors are struggling.** We see that grain, oil, and food still grew by 9.9%, home appliances and audio equipment grew by 12%, while automobiles and cosmetics saw negative growth.
**The second reason, and the most critical indicator for deep distribution, is the store closure rate.**
If we continue large-scale, disorderly deep distribution under such high closure rates, what will it bring? Only bad debts.
**A large number of bad debts have made distributors afraid to engage in deep distribution.** How can we do deep distribution under such bad debts? One failed store wipes out the profits of ten successful stores. In some cities, the closure rate is as high as 16% or 15%. This is why deep distribution is no longer effective.
**The third reason is product abundance.** The emergence of Douyin has allowed a large number of products to sell quickly, and product innovation continues.
Online platforms like Tmall, Taobao, and Douyin have over 20,000 categories, more than 50,000 brands, and over 800,000 products. New products are emerging, but we are still promoting familiar products. Will consumers still buy them? This is a very important question.
What kind of products do consumers really like? Do the products they like offer good value for money and quality? This is the third key question. If we sell products that consumers don't truly like, then any distribution will be useless.
Facing these three problems, how should we solve them?
**Deep distribution is difficult, so how to solve it?**
First, solve how to do deep distribution precisely (i.e., how to choose the right "place").
Add a scope to deep distribution, unlike the past when we did everything like sprinkling water. The underlying logic remains people-goods-place: people looking for goods in places is the core offline logic—what kind of people look for what kind of goods in what kind of place.
In an era of scarce resources, there were only so many places; as long as you went to a place with goods, you could buy. Now there are many places, with various schools, such as category-rich stores and category-curated stores like COSTCO, which does very well. So precise matching of people-goods-place is the most important point in deep distribution. The first and most important thing is the quality of the place.
How to choose the right place among 300 stores? We built many models using DNN and random forests.
The first model is called high-risk stores. Don't think about how to do business well; first think about how to avoid pitfalls. If you avoid pitfalls, you've already won half the battle. Verified, the closure rate for high-risk stores is 50%.
Now look at ordinary stores: verified closure rate is 13.45%; for high-potential stores, the closure rate is only 3.24%, basically no loss.
But is it certain to win if you only focus on high-potential stores? I think there is a chance.
Opportunity comes from practice. We chose a city, selected over 800 high-potential stores from 16,000 stores, and tried to see if focusing on high-potential stores could improve sales. After one year, growth in that city exceeded 60%.
This model is useful, and in this model, the closure rate for high-potential stores is only 5%. So from the perspective of focusing on high-potential stores, if you can focus correctly, business growth is not particularly difficult.
Second—choose the right products.
Today we are facing distributors. Distributors have a characteristic: they don't produce products but help consumers choose.
We see hundreds of thousands of products and tens of thousands of categories. Not every category is good. At the beginning of the year, we did an analysis: from 1,400 categories, we selected those with year-on-year growth exceeding 100%, and found 96. There are still over 100 categories with year-on-year decline exceeding 40%. If the category you operate in is not particularly good, then growth itself will face huge difficulties.
Let me share a particularly interesting category that emerged from Douyin: female intimate wash. It is growing very fast. One brand, Baidi Bio, sells two bottles for 59.9 yuan with over 80% gross margin, growing from 4.17 million to 240 million in one year.
Sometimes it's not that business is bad; maybe your business is bad, because consumer needs are constantly changing and being stimulated.
Consumer profiles are becoming increasingly distinct. In this context, more new categories are being stimulated, and more old categories are shrinking or changing.
Another example: in recent years, a new product called "explosion salt" (oxygen bleach) has become very popular, squeezing the traditional laundry detergent category, but the total volume hasn't changed.
The 48 trillion yuan social retail sales won't change. Therefore, facing future stock competition, it doesn't mean your category will maintain its current stock, nor does it mean your chosen category won't have explosive opportunities. The important thing is how to use big data to choose categories.
Third—consumer profiles.
In today's era, the population has undergone huge changes, wealth distribution has changed, and consumer profiles are diverse.
In recent years, online consumer profiles have become very rich, and offline also has rich profiles, but the complexity is much higher than online. We cleaned the consumer profiles and accumulated 1.405 billion profiles. Here are some labels we picked.
There are 208 million adult consumers without children. What kind of emotionally valuable products can we provide to make their lives better? This is a very important question and also an opportunity.
For feminine hygiene products, sanitary pads, there are 320 million women aged 14-45. To sell sanitary pads offline, you must know where they are distributed. If the community is predominantly male, selling this product is difficult. If the density of potential customers is very high, then promotion is relatively easy to succeed.
So, we must shift from the past mindset of "go where there are many people" to the future mindset of "target the target population." **Places with many people are not necessarily good for business; places with target populations are where business thrives.**
**What about local data in Chengdu?**
We came to Chengdu to look at some interesting data in the local database.
We see that Chengdu's total retail sales in the third quarter only grew by 3.3%, with total retail sales of goods exceeding 800 billion yuan, up 2.6%, and catering revenue up 6.2% year-on-year.
The total population is 24 million, with 46,800 retail outlets, 1,191 industrial parks, 4,051 commercial buildings, etc. This is Chengdu's business landscape.
All opportunities are here, but so much data doesn't directly help us; it only gives us a direction.
This is data for a 2-square-kilometer area: 7,000 permanent residents, 7,294 main consumers, 11,000 permanent residents, 4,600 office workers, one-third are entrepreneurs, 4.4 freelancers, and the surrounding consumption level is rated high.
**Only by precisely targeting the consumer profile of each area can we truly achieve new people-goods-place matching and the reconstruction of deep distribution.**
All of this comes from DNN and random forests. We turned it into products, hoping these products help distributors get data support and digital product support more easily in future business.
We divided the country into 56 regions, each averaging 3 square kilometers. We tell you which stores are there, which are high-potential or high-risk, what the consumer profiles are, their spending power, age composition, and how many business opportunities and business population each city has.
With this, combined with market experience, you'll find that people-goods-place matching becomes easier and better judged.
Additionally, in the system, we have accumulated 3,700 categories, 40,000 brands, and 700,000 products. Only by finding the right category among many can you catch the trend; standing at the trend increases your chances of winning.
Of course, will mastering so much data guarantee winning? Not necessarily. **Strategy is the final key.** So we also condensed 10 strategies to help distributors use big data products well.
Most distributors have heard about big data for a long time and have been exposed to it, but they haven't really applied it for long. Big data application means that after you get the data, in many cases, you must learn a thing called insight. Only after insight can you have the right strategy, and only after execution can you truly get the right results.
Data will be a particularly important tool in the future. If you don't master data and rely only on experience, the probability of failure is high. The closure rate alone will cause losses you cannot bear.
We hope distributor friends can use big data to do better business in the future.


---

## 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