FMCG companies' channel organizations are at a point where change is imperative. This feeling has grown increasingly strong recently. In late 2023 and early 2024, I shared this judgment in some articles. At that time, many readers commented, saying I was alarmist. Frankly, back then the industry hadn't reached today's urgency. Many companies felt slower growth, but overall they could still manage with old methods. Push the channel, promote policies, and short-term results could still be achieved. Recently, I compiled four years of systematic research into a volume, bringing together four reports on channel transformation. This allowed me to review the underlying logic of distribution over these years.

Why is retail changing?

Where is the distribution system heading?

Why can't distributors just be movers?

What drives growth in an era of surplus? —These questions have been addressed quite clearly. But looking ahead, another bigger variable has arrived. That is AI. The FMCG industry is already experiencing the impact of overcapacity, and with AI challenging organizational structures, these two forces combined make it urgent for traditional FMCG channel organizations to change. So I decided to write a series of articles to systematically share my thoughts on this issue. I also look forward to exchanging ideas with industry peers.

Traditional Channel Organizations

Were Tailored for the Distribution Era

Let me be clear: traditional channel organizations are not wrong. They have strong historical justification. In the growth era, the most important thing for FMCG companies was to distribute products. The market was growing, outlets were expanding, and consumer demand was not fully met. At that stage, whoever could cover more outlets, push deeper into channels, and execute in-store displays and promotions had the opportunity to capture growth. So, organizations naturally grew into a hierarchical transmission structure. Headquarters set targets, regions broke down tasks, branch offices monitored execution, sales office managers pushed down, distributors received goods, and sales reps visited stores. Everyone on this chain was essentially doing the same thing: pushing products down, transmitting tasks, and collecting information. This design made sense at that stage. Because the market was fragmented, terminal information was opaque, and system capabilities were weak, companies couldn't directly see the terminals. They had to rely on people to run, monitor, and report back layer by layer. Many middle management roles actually functioned as "human systems"—transmitting information down, reporting results up, and coordinating in between. Moreover, this organizational model had a built-in premise: the market would automatically absorb the goods in the channel. As long as you distributed and pushed, growth would likely follow. Problems could be masked by growth, and pressure could be absorbed by the market. But that premise no longer holds.

The Market Has Changed

Old Answers Don't Match New Questions

Today, many companies' pain, on the surface, is sales pressure, but deep down, it's a mismatch: The external environment has changed, but internal channel organizations are still stuck in the previous era. First, consumer purchase paths have fragmented. In the past, consumers went to supermarkets, small stores, and wet markets, with relatively stable paths. Brands only needed to enter these major terminals to be consistently seen and purchased. Now, the same consumer might buy in physical stores today, order via instant retail tomorrow, stock up at membership stores the day after, or grab an item at a fresh food store. The fragmentation of purchase paths means brands need to be present in more and more places, and these places have different stocking logics. Second, retail's bargaining power has strengthened. Retail is no longer just the end of product sales; it's increasingly the entry point to consumer demand. Membership stores, discount stores, and instant retail platforms are reorganizing consumer purchase scenarios in their own ways. They are now demanding that brands adapt their assortments, prices, specifications, and turnover requirements. In the past, brands were stronger, and channels and retail mostly carried brand supply. Now, this relationship is adjusting. Third, distributor capabilities are diverging. We can no longer broadly talk about "distributor channels." Among the distributors I've visited, some can still serve terminals, supply instant retail, and connect with emerging small stores; others only have receiving and warehousing functions, with almost zero terminal service capability. Even among distributors, the actual value they deliver is completely different. These three changes together point to one issue: The old organization answers the question "Is distribution coverage sufficient?" But the new era's question is "Are products reaching the right terminals through matching channels to meet matching scenario needs?" These two questions fundamentally don't align. We don't lack channels; we lack matching capability. But the design logic of traditional channel organizations was never for matching—it was for transmission and distribution. Using a distribution-oriented organization to solve matching problems is where many companies are truly stuck.

AI Adds In

Exposing Old Organizations' Problems Faster

Just as the FMCG industry enters an overcapacity cycle, AI is rapidly entering enterprises. Many people initially understood AI as a tool—faster copywriting, faster spreadsheets, faster PPTs. That's useful, but it's not the most important change. AI's real impact is changing "who processes information and who makes judgments." A core reason traditional channel organizations need so many middle layers is information opacity. Headquarters can't see terminals, so regions are needed; regions can't see every county, so sales offices are needed; sales offices can't see every store, so sales reps and distributors are needed. Information must be passed up layer by layer, and judgments pushed down layer by layer through experience. But AI is breaking this logic. Which outlets have long-term no-sales, which distributors have abnormal inventory, which expenses don't bring real conversion, which regions have high coverage but low output—these pieces of information, which previously required manual collection, aggregation, and analysis, are becoming outputs that systems and AI can process. This doesn't mean AI replaces human judgment. Trade-offs still require humans. Which distributor to keep or replace; which format to invest heavily in or pilot; whether to allocate expenses to channels or terminals—AI can provide analysis and insights, but judgment remains human. This means AI has changed one thing: the core value of middle layers was information relay, and now that value is being taken over by systems. If a role's main work is relaying messages, chasing reports, aggregating, and forwarding, its reason for existence will be questioned. Not because the person isn't hardworking, but because such work itself is being redefined. The pressure from market overcapacity and the organizational restructuring brought by AI are two different forces. But they converge in the same time window, making the challenge for traditional channel organizations more complex than any single variable.

From Distribution Organization to Matching Organization

Channel transformation has reached a point where the bottleneck is not external but internal. So I want to propose a framework to describe the direction of this change: From distribution-oriented organization to matching-oriented organization. Distribution-oriented organizations ask: How many outlets? How many distributed? Have distributors paid? Are tasks completed? Matching-oriented organizations ask: Which consumer segments? What needs in what scenarios? What products, delivered through what methods? A one-word difference corresponds to completely different underlying logic.

  • Distribution-oriented organizations focus on transmission—pushing tasks down, collecting information up, and pressing goods into channels. Their performance metrics are coverage rate, task completion rate, and distributor payment amounts.
  • Matching-oriented organizations focus on judgment—what products suit what scenarios, what channels suit what terminals, what distributors suit what tasks, and where expenses yield real sales. Their performance metrics are sales efficiency, terminal output, and regional business quality. This shift is not simply layoffs, not compressing layers, and not installing an AI system. Layoffs won't reduce costs, flattening won't improve judgment quality, and if AI is connected to old processes, it just makes the old organization run faster. What truly needs to change is the core function of every role in the organization—from transmission to judgment; from pushing goods to operating. This is the direction of my further research. How to restructure regional manager roles, how to re-tier distributors, and where AI is truly useful in specific scenarios—this series of articles will elaborate one by one. But discussion through articles alone is clearly insufficient. Because restructuring channel organizations isn't something one person can figure out sitting in an office. It requires brand owners, distributors, retailers, and technology service providers to sit together, lay out the problems, break down the scenarios, and truly validate tools in the business chain. So, at the upcoming "2026 China FMCG Conference AI Application Forum," we are also making this topic a core direction. This conference isn't just about AI trends; it focuses on a more practical question: How exactly does AI enter the growth chain of FMCG companies? From the brand owner's perspective, AI shouldn't just stay in copywriting, PPTs, and meeting minutes; it should enter market insights, channel matching, expense allocation, store operations, distributor management, and organizational collaboration, helping companies re-see the market, allocate resources, and improve sales efficiency. From the distributor's perspective, AI is not a distant concept. It can help distributors monitor inventory, sales, risks, and profits, and even assist in judging which stores deserve deeper cultivation, which products should be adjusted, and which expenses don't produce real results. So, we've also set up a "Closed-Door Course on AI Application and Growth Implementation for FMCG Enterprises" and a "Hands-On AI Practice Course for Distributors." One is for brand enterprise decision-makers, discussing how AI restructures organization, growth, and channel management; the other is for distributors, explaining how to truly nurture and use AI tools after installation, turning them into advisors and assistants in daily operations. For conference details, scan the QR code to add the WeChat enterprise account and contact the organizing committee.