In this video, we will discuss how to assess the maturity of a demand planning process. We will explore five key areas of maturity:
- Driver data quality: How well does the organization understand the drivers of demand?
- Demand planning process structure: How well is the demand planning process designed?
- Collaborative demand planning process: How rigorous is the collaborative consensus process?
- Use of ML/AI and statistical models: How effectively is the organization using ML/AI and statistical models to forecast demand?
- Organizational capability: How well are the organization's people and systems capable of supporting demand planning?
We will also discuss how to use this assessment to identify areas for improvement and make a plan to achieve world-class demand planning.
Demand planning is a critical process for any organization that wants to ensure that it has the right amount of inventory to meet customer demand. However, demand planning can be a complex and challenging process.
One way to improve the effectiveness of demand planning is to assess the maturity of the demand planning process. This assessment can help to identify areas where the process can be improved.
There are five key areas of maturity that can be assessed:
1. Driver data quality: How well does the organization understand the drivers of demand? This includes understanding the factors that influence demand, such as customer behavior, competitive activity, and economic conditions.
2. Demand planning process structure: How well is the demand planning process designed? This includes the frequency of forecasting, the level of detail in the forecasts, and the way in which forecasts are communicated to stakeholders.
3. Collaborative demand planning process: How rigorous is the collaborative consensus process? This includes the way in which different stakeholders, such as sales, marketing, and operations, are involved in the forecasting process.
4. Use of ML/AI and statistical models: How effectively is the organization using ML/AI and statistical models to forecast demand? This includes the use of data analytics to identify trends and patterns in demand.
5. Organizational capability: How well are the organization's people and systems capable of supporting demand planning? This includes the availability of skilled personnel and the use of appropriate software.
Once the maturity of the demand planning process has been assessed, the organization can identify areas where the process can be improved. This can be done by implementing specific improvement initiatives, such as:
- Improving the quality of driver data
- Redesigning the demand planning process
- Increasing the rigor of the collaborative consensus process
- Investing in ML/AI and statistical modeling
- Developing the skills of demand planning personnel
- Investing in demand planning software
By taking these steps, organizations can improve the maturity of their demand planning process and ensure that they have the right amount of inventory to meet customer demand.
Chakri, So you've talked about the importance of Demand Planning. How can organisations assess the level of the maturity of their planning processes and where they can improve? That's a good question. Every company, of course, has a demand planning process in place to run their supply chains today.
The question is which parts of those processes are most broken or need most improvement. And that's important to understand to lay out a roadmap on how to drive incremental and attractive improvements. So I would classify broadly Demand Planning process evaluation into five broad areas. First, let's say on a monthly basis or weekly basis, you're getting your actuals versus the forecasts.
And there's a lot of surprises in there. You know, forecast accuracy is low, then the question is why is actuals deviating from the forecast? And a lot of times then you get answers or this happen that happened in the market. These are all leading indicators or drivers of demand.
So the knowledge of drivers of demand is one of the key areas of maturity. The drivers of the demand could be external drivers related to customers ,competition, the market drivers of demand could be internal drivers, which are commercial actions that your organization is taking that impact the demand and the drivers of demand could also be supply chain So there are a lot of drivers that are impacting whether the actual performance is meeting the forecast or not. And typically organisations have, you know, either they don't have the data, which is a big problem or a lot of times they do have the data, but the dots are not getting connected to really explain why the forecast is deviating from the actuals and what drivers are causing that.
So one of the levels of maturity is how good is your driver data, how well is it being collected and connected in order to explain the differences between forecast and actuals? So that is one category I would call assessing the maturity of driver quality data. Driver data quality. The second category is often in demand planning processes.
The three key elements of the structure of a good demand planning process. I call that the horizons, the cycle frequency and the granularit. Its very, very important to get the design of those things correct to set up a good demand planning process. We'll probably have to go into a lot more detail, explain that.
But at a high level, the horizon is how far out are you planning the demand? Right? Because your supply chain has lead times. The lead times dictate how far out your forecasting demand is.
Is that set up correctly? The granularity at which you forecast is also very important. That's also dictated by the decisions you are trying to drive in the supply chain. A lot of times companies get that wrong and they are forecasting ta to low level of granularity too far out in the horizon.
It's fake details for no value. So getting the granularity right is extremely important in the shorter term. You might need at a more granular level, further out, you might need at a higher granularity depending on the decisions you are making. It has to be designed based on the the decisions you are driving the supply chain.
And then the third one is the frequency with which you're forecasting. How frequently do you forecast for the short term demand? How frequently do you forecast for the mid to long term demand? Getting that right as well is very, very important.
A lot of companies don't get the balance of those three things correct and it creates either a lot of either fake details and a lot of extra work that doesn't add value or you're getting surprised because you're not forecasting fast enough. So that we have to get into more detail. But that's a big area of maturity of the demand planning process. The third one I would call is basically the collaborative demand planning process.
As we discussed earlier, there are a lot of stakeholders in the demand planing process, You have supply chain, you have sales, you have finance, etc. and you need to bring together different perspectives on the forecast. But most importantly is the rigour with which the collaboration process is looking at the forecast. What has changed from last cycle?
What are the reasons for the changes? The assumptions and the reasons being challenged? Are multiple perspectives coming in? Is there a rich dialogue around why the forecast has changed?
A lot of organisations, especially when they put into statistical forecasting and are just pumping the numbers through and it's causing a lot of bullwhip in the supply chain, whereas there has to be a lot of collaboration and discussion around forecast changes, especially when they're changing from cycle to cycle. So the maturity of the collaborative consensus process is a big element of a good demand planning process. The fourth is the use of ML/AI and the statistical models. Clearly, given the breadth of the product portfolio, the breadth of end markets, your serving customer segments, your serving the supply chain locations that you are to ship from, the complexity of the demand planning process can be huge.
So a automation or forecast generation is actually a very, very important aspect. And are you using the right techniques for the right part of the product portfolio. Some products are imminently forecast to well, some products may not be. And you need to combine human intelligence plus tribal knowledge along with some data to forecast.
How good is the usage of ML/AI statistical forecasting? There's a fourth category of maturity that we need to look at. And finally, what I would call the maturity in terms of maturity is the capability of the organization in two dimensions system and the people. The outlook at the lowest level of maturity is all being done in spreadsheets.
The demand planning process is barely getting done in collecting the data, collating the data in spreadsheets and sending it to the supply chain. So demand planners have become more spreadsheet aggregators than really analyzing the demand plan. In the middle, somewhere it's more about they have put some system in place, but really because of the inflexibility of the systems of the demand planning system, most of the work is still being done in spreadsheets and reviews having done in PowerPoint, but there's a system to store the demand, so it feeds downstream systems like the supply chain, etc.
And this is in a lot of places where they have implemented first generation planning systems. They're in that state. And the more mature stage is where there's actually a full fledged demand planning system, where the demand planning process, the data, the analysis is all being done in a system. Then I would look at more the quality of the leading indicators, the quality of ML/AI forecasting.
That's where the improvements could be had. So that's the system side. On the people side or the capabilities of the organization. There are two aspects of good Demand Planning organization.
One is the forecasting function, which is how do I take all the data, the drivers and develop the algorithms, tune the algorithms so that they're creating the right forecasts and the right insights. Then there is a planning function which is should be much more about the analysis, managing the risk in the demand, explaining to business stakeholders why the demand has changed, what the risk is. And often those two roles get conflated. So one is more of a data science analytics algorithms.
They need to be business savvy. But it's a different persona and one is more of a business oriented risk management role. And so those are the skillsets that need to be heard. We have to assess how good the Demand Planning organization is in those skill sets.
So overall, in these five categories, we need to look at what the level of maturity is in the demand planning process, and then we can make a surgical design of how to improve the capability along each of these dimensions to get to a world class demand planning process.
Next-generation Dairy Planning
Download our Point of View for a break-down of the key planning capabilities needed for dairy companies to unleash growth, resilience, and sustainable success.