“Digitalization” is very hot right now, but how hot is it? From December 3-5, 2020, New Distribution, together with teacher Liu Chunxiong, hosted the “Marketing New Year Kickoff Course” themed around digitalization, which attracted over 30 brand owners to attend in groups, making it one of the highlights of New Distribution's 2020 conferences. Bi Chaojiao, General Manager of China Resources Snow Breweries' Marketing Center, stated that the next five years will be a critical period for digital construction. The strength of digital capabilities directly determines a brand's market competitiveness. Marketing expert Liu Chunxiong defined 2020 as the year of digital infrastructure, and 2021 as the year of digital operations. In his article “2021 Marketing Digitalization Outlook: The Year of Digital Operations,” he systematically elaborates on this, focusing on CDP (Customer Data Platform) development and full-chain digitalization (BC integration). I believe Teacher Liu's judgment is too optimistic! In fact, the vast majority of brand owners and distributors are still in the primitive stage of channel digitalization, and the problem of “toxic” channel data is far from solved, let alone the advanced stage of digital operations. -01- Toxic Channel Digitalization If “digital infrastructure” refers to the development and implementation of systems, then the achievements are commendable. Many excellent marketing technology service providers, such as “Fenxiang Xiaoke” and “Waiqin 365,” saw significant growth in business from the FMCG industry in 2020. Fenxiang Xiaoke also received a new round of substantial investment. Many leading brands began developing and using digital systems years ago, with some even establishing their own system development companies to undertake this mission. In terms of technical capability alone, digital capabilities are indeed very impressive, and the adoption rate is quite high. But what about the actual results? Unfortunately, the performance of most companies is substandard, and their channel digitalization systems are operating with “toxins.” The biggest and most stubborn virus in channel digitalization is data fraud. One marketing manager said that the longer the system runs, the less trustworthy the data becomes. Basically, except for the final totals, no other data can be confidently considered real. In the end, the numbers in the system are just something one party says and the other listens to, with no one taking them seriously. But if we can only use the totals to prove authenticity, why spend so much manpower and resources to build and maintain a massive system? -02- Why Can't Channel Data Be Real? This problem not only puzzles business managers but also troubles the National Bureau of Statistics. Every year, local GDP figures never match the National Bureau of Statistics' GDP data. In one year, the sum of local GDPs exceeded the national figure by 4.6 trillion yuan, equivalent to the entire GDP of Shandong Province, causing embarrassment for all parties involved. Why is channel data fake? First, the data sources—distributors and salespeople—lack the will to seek truth but have the motivation to fabricate. Let's talk about distributors first. Distributors view terminal customers and order data as trade secrets; if they share them, they feel insecure. Therefore, distributors instinctively resist handing over data to upstream parties, and if they can't resist, they comply in appearance but act in opposition, resorting to fraud. Additionally, many brands have different sales policies for various terminal types. For example, in the beer industry, on-premise terminals often receive more investment, while off-premise terminals receive less. If you were a distributor, what would you do? Wouldn't you be tempted to input off-premise terminal orders as on-premise terminal orders? It's simply driven by profit! Now let's talk about salespeople. Every salesperson has KPIs, and the company not only assesses sales volume but also single-product share, single-product growth, and new product data. These data are all related to monthly bonuses, so what to do? Just do it! Manipulate the data to meet KPI requirements, making the data “flawless” to get the maximum bonus. To meet targets, many sales volumes that flow through “sewers” cannot be exposed to sunlight, so the data naturally needs to be “adjusted.” Some salespeople with good relationships even “borrow” performance from each other, creating a win-win for everyone. Second, the systems of various brands are incompatible, and data entry consumes enormous effort, making “data backfilling” a regular task for sales representatives. Currently, each brand's digital system is developed independently and is incompatible with others. Once a distributor represents multiple brands, all with data requirements, it becomes a nightmare. For example, if you are a distributor for Yinlu, Nongfu Spring, Uni-President, and Snow Beer, you must deal with four systems and four sets of requirements simultaneously. When a terminal places an order, you have to place orders through four different apps, and you also need four sets of sorting and delivery orders. What can you do? Do you still want to do business? The simplest solution is to not use them at all. So what if the brand needs data? The sales representatives of each brand can fill it in themselves. Thus, regularly “backfilling data” becomes a routine operation for brand sales reps. As for how much of it is real, we're all just workers, so why care so much? The channel system data of most brands is untrustworthy, turning the massive system project into a tool for the marketing system to fool itself. This cruel truth is like the emperor's new clothes; no one wants to be the child who points it out. -03- How to Detoxify Channel Digitalization? Is there a way to solve the toxic data problem in channel digitalization? Yes, there is. First, make data generate in real-time in real scenarios as much as possible; this should not be difficult technically. For example, orders must be matched or verified based on LBS information, and LBS information cannot be customized. If a large number of orders have LBS information inconsistent with the customer's LBS, it is highly likely that “fake orders” or “backfilled orders” exist. Additionally, you can bind the salesperson's ID to their phone's device identifier. If there are frequent cases of changing phones to place orders, problems may have occurred. Technical measures cannot prevent problems, but they can at least help us detect and audit fake data issues in a timely manner. Second, design business models or organizational structures to separate the flow of goods, information, and logistics as much as possible, increasing the difficulty and complexity of “collusion for fraud.” For example, outsource logistics to a third-party service. Midea Group's “warehouse and distribution integration” service provider, Annto Zhilian, has already provided fully independent third-party warehouse and distribution integration services to over a hundred FMCG companies and distributors. Once logistics is independent, the third party will only operate precisely according to system instructions. In such cases, data fraud becomes almost impossible. But what if many brands cannot yet achieve third-party warehousing and distribution? First, the authority to audit system data should be centralized at headquarters, with regional marketing organizations having only viewing rights. This makes the system's operation independent of regional marketing management, establishing a big data middle platform. Data flows directly from the front line to headquarters, and data review and audit also go directly from the headquarters system to regions and individuals. By stripping regional marketing organizations of the possibility of participating in data production, fraud can be maximally avoided. It is understood that in Fenjiu Group's digitalization construction, its marketing system data flows directly from front-line terminals to Fenjiu headquarters. Terminal information, promotional activities, expense reimbursement, and reward calculations are all reviewed and approved by headquarters according to standards. Regional organizations have no authority to intervene in system data, thus minimizing organizational data embellishment. In the early stages of Fenjiu's business system implementation, there were some impacts on business flexibility, and some salespeople did not receive deserved rewards because they did not comply with system standards. But in the long run, the results have been quite good, with the proportion of data distortion decreasing. Second, the authority to enter business data related to orders should be strictly bound to salespeople. Distributors can only operate inventory and order processing data, and delivery data can only be entered by delivery drivers. Separating permissions can maximize the prevention of data collusion and fraud. Third, since the source of business data is the salesperson, on one hand, we must infinitely increase the difficulty of data fraud; on the other hand, we must eliminate the motivation for fraud through incentive innovation, making salespeople willing to enter real data. Many big data systems are already mature, and cross-validation of data can detect some not-so-sophisticated fraud. One brand innovated its reward system in practice by making the reward algorithm opaque. The system only tells salespeople what basic actions to take and what the assessment criteria are, but it does not disclose the weights and calculation formulas for each action. Business rewards are input and determined by headquarters system personnel based on regional business requirements, and no one else can intervene or view the calculation method. As a result, distributors and salespeople can see the reward amounts in the system but do not understand the corresponding algorithm. Since the data is calculated by the system and cannot be manually manipulated, it can be trusted as fair. At the same time, because they do not understand the specific algorithm, they do not know the system's “preferences,” making fraud impossible. The best choice is to do every job well and enter data truthfully. This brand's incentive innovation turned the traditional KPI assessment game into one with clear goals and vague rewards, and because the system automatically calculates, it ensures fairness, to some extent eliminating salespeople's motivation for data fraud. In practical application, it has achieved good results. To perform surgery to detoxify toxic channel digitalization, much more is needed to integrate technology into management innovation and business model innovation. If we cannot ensure that data is clean and real, then no matter how much we spend on building a digital system, it is just a digital castle built on sand—impressive but useless. If a tip is adopted, a reward of 400-2000 yuan will be paid.