I’ve worked with several software companies that develop planning software. Some are based on mathematical statistics, offering a variety of forecasting models for optimal selection; others leverage artificial intelligence and machine learning to provide unique solutions for demand forecasting and inventory planning. These software programs are far from perfect, but overall, they’re better than a bunch of planners working independently based on their own intuition. However, many companies are reluctant to adopt these tools, or even if they do implement them, they still rely on manual planning in Excel.

There are several main reasons.

First, planning software can’t effectively address significant errors , such as low forecast accuracy caused by new product launches, promotions, and fluctuating demand. We know that a good forecast “starts with data and ends with judgment.” Good software systems can help us achieve continuous improvement, create baseline forecasts, and address the “three-point technical” aspect, effectively starting with data. However, software can’t address the “seven-point management” aspect, such as sales and operations coordination, requiring judgment from sales, marketing, and product teams to better adjust baseline forecasts. Therefore, this “judgment-based” approach doesn’t exist.

The strength of planning software isn’t to break down cross-functional barriers and facilitate cross-functional collaboration; it’s to help you achieve continuous improvement and improve data-driven plans . Using the wrong metrics to measure planning software will inevitably lead to erroneous conclusions. For software developers, if they don’t have a clear focus and spend a lot of energy solving organizational and process issues, they’re likely wasting resources. Software developers should focus on software functionality and handle data analysis; organizational and process issues are the responsibility of implementation consultants.

For some companies, because their overall operations are too extensive and haven’t yet reached the stage of continuous improvement, they simply can’t afford to implement such planning software. Instead, they’re advised to focus more on organization and processes, fostering collaboration between sales and operations, which may yield a higher return on investment. In other words, organizational administration is the first thing we need to solve , addressing the “70% management” issue; then, address software systems, addressing the “30% technical” issue. Of course, once “70% management” is achieved and drastic mistakes are avoided, the next step is to pursue continuous improvement, and software systems are essential.

Second, software often fails to significantly reduce overall inventory levels. Many planning software touts inventory reduction as a key selling point, but this isn’t true: good planning software generally reduces inventory, but the overall reduction is often modest. The greater value of planning software lies in improving inventory structure, placing the right inventory on the right products. While appropriately reducing inventory, the system improves service levels. Let’s use fitness and weight loss as an example.

Like weight loss through fitness, inventory control has three stages. In the initial stage, inventory is high, but customers don’t have what they want. It’s like a body full of fat in the wrong place—around the belly. The result is high inventory and low availability, a starting point for many companies. Then, you start exercising, running and lifting weights all day. After a while, you step on the scale and are disappointed to find that your weight hasn’t changed much, or may even have increased. Think about it: with constant exercise, a better appetite, and more food, can you lose weight? However, you’ll also notice that your body is more balanced, with the fat growing in the right places. This is the high inventory and high availability stage. So, continue training, be more “hard on yourself,” change your diet, and live a more restrained lifestyle. Little by little, you’ll lose that fat, build more muscle, and lose weight. This is the third stage: low inventory and high availability.

In this three-stage “fitness and weight loss” process, planning software improves forecast accuracy, sets more appropriate safety stocks, and allocates the right inventory to the right products, thus alleviating the coexistence of oversupply and shortages under extensive management. It can be said that planning software primarily addresses the second stage: achieving high availability despite high inventory levels. So how can we reach the third stage? This requires returning to the three main causes of inventory: turnover cycle, uncertainty, and organizational behavior. We can reduce turnover inventory by shortening turnover cycle, reduce safety stocks by reducing demand and supply uncertainty, and reduce excess inventory by changing organizational behavior. I have written extensively about these issues. This requires comprehensive collaboration across marketing, production, procurement, and the supply chain. It is more of an execution issue than something that planning and planning software can address. Planning software is more about accurately quantifying, but it cannot address the root causes of these inventories.

Third, there’s a disconnect between the system and the organization, and planners are reluctant to relinquish control . This is somewhat puzzling. Simply put, while the software may be excellent, it’s difficult to use due to various factors, making the planning results unreliable. Or, the logic is complex, making it difficult for planners to understand. Without understanding, they lack trust, and without trust, they naturally don’t use it. Both of these factors contribute to a disconnect between the system and the organization. Even after implementing planning software, planners often continue to work on their own, working with Excel spreadsheets. The planning software becomes little more than an interface for uploading planning results and connecting to the ERP system.

You should know that any planning software, to produce reliable results, requires data cleansing, such as removing non-repeated or erroneous data. This is laborious and, if done properly, takes a considerable amount of time. Many companies lack the resolve and resources to do this well. Naturally, the planning team will not trust the planning software’s suggestions and will naturally not adopt them, rendering the software system useless. This is one reason why the system and the organization are two separate entities.

Another reason is that planners simply don’t understand much of the software’s logic, such as the principles and optimization of forecasting models. Mathematical statistics, while objective, are cold and impersonal, and few truly grasp their understanding. Naturally, we can’t trust something we don’t understand, and we feel a loss of control. Humans crave certainty—not just of the outcome, but more importantly, of the process. This all contributes to software disuse.

Furthermore, the people who design these logics often have an IT background. I don’t want to offend IT professionals, but this isn’t their expertise, so they can only mechanically apply formulas. Planners, while unfamiliar with these logics, can easily tell that the software’s plans are unreliable and, of course, won’t use them.

Mathematical statistics, while objective, are cold and impersonal, and few truly grasp their understanding. Naturally, we can’t trust something we don’t understand, and we feel a loss of control. Humans crave certainty—not just of the outcome, but more importantly, of the process. This all contributes to software disuse. Furthermore, the people who design these logics often have an IT background. I don’t want to offend IT professionals, but this isn’t their expertise, so they can only mechanically apply formulas. Planners, while unfamiliar with these logics, can easily tell that the software’s plans are unreliable and, of course, won’t use them. In light of these complexities, it is imperative to bridge the gap between statistical rigor and practical decision-making for effective planning strategies.