How Can Data Science Help to Take the Best Decisions in a Company?
Nowadays the companies live a constant need to reinvent, the competition is very hard. So the executive has to take better a quickly decision every day.
In a company, the executive takes decision every day. In virtually every decision they make, executives today consider some kind of forecast to understand how will be the sales in the next months or years.
The principal function of a forecast is to give visibility to the company.
But with this revolution of data science is not the only tool to get visibility of the company. The executive needs visibility to take a decision but that not mean just a forecast. Also, they need to know the strengths and weakness of the company and practical diagnose of the situation in all department.
Although we know that is very important information. To get the information and create a meeting to talk about the situation for areas to take along the time that the executive doesn´t have. In the ’80s born the S&OP.
This process promises to solve situations like the previous one by aligning demand and supply. However, as the process has evolved, it has been understood that its scope is much broader and that well executed it becomes a central process to the integral management of the business.
The form to take a decision in the business and they agree about the forecast and the operation management. It is to have an S&OP.
Sales & Operation Plan (S&OP)
¿What is an S&OP?
“The integral process of management and business decision making to balance the demand and supply, align the commercial, operational and financial plans with the business strategy in an adequate time horizon”.
The key concept is “integral management process and business decision making”: The one we follow to make a company assign its resources and execute daily actions that are consistent with the plans and objectives. The one that facilitates the decision making that marks the course of said actions.
Remember that, the decisions must be made with respect to the benefits that it generates to the company, the executive must have his recognition that all action oriented to an end entails a cost. Every action uses scarce resources that can not satisfy the next higher value order.
The problem that is not the only decision-maker must also understand that there are other agents that are going to make their own decisions regarding their own incentives (benefits).
That means, prioritize decisions or investment that agree with the goal of the company. To carry out that decisions are very important the metrics we use to follow the plans of the company.
The metrics which are finally governing the behavior of people (they do, what the metrics say) that point is very crucial because If they have great metrics, they have great results.
That is the first point to participate Data Science in process of S&OP. General, there are already metrics to follow objectives and plans. But not always are the correct.
In the Sales and Operation Plan is the place to diagnose the situation of the company.
In this process, the basic premise of the S&OP is that it must be a plan created to unify strategies and information in the areas of Sales, Supply Chain, Operations and Finance.
Data Science helps to get a better analyzed and find areas of opportunity that another area they can´t see. In the S&OP, there is a coordinator or S&OP team follow the process and manages the information, such participation is not limited to collecting information, but requires dynamic interaction and intelligent discussion with all the areas involved.
The principal fail is no integrated metrics. The S&OP is the process where it put together the different department of the company. So, the departments of the company should have metrics that check the objectives of the different department.
That means, we cannot require sales of the 20% when we do not have the capacity to support the sales for a supply chain. Also, we cannot say to commercial to get the most margin possible occasionally force the inventory management to get GMROI down.
We must align the metrics of the different departments, that means, the metrics of a supply chain are the same for operations, commercial and vice versa.
Data Science participates to separate information and classification of the metrics of the company. In most companies, we have a lot of items to do, coordinate and sell. In some companies we have a classification of these items, the general is for its sales or the frequency of his sales calls it “The ABC Code”. How can you see? it is a very simple general classification. That generates that although help us to focus on some items. The classification is very general and may is not the correct.
For example, we have classification the sales for ABC codes depends on the frequency of the sales. A is the most frequency sales and C is the fewer sales. The items with a code A may are items that, yes, we sell a lot of that item but may e his margin is not great. Do you need to focus on this item?
Now, data science allows classification in a better way. We can classification on variability, margin and sales frequency. That means, products are constants, frequency and generate money.
With a simple classification, we can see a better form of the importance of the items and we can get better actions to respond.
For these reasons, It´s very necessary that The coordinator or the S&OP team should be additionally skilled:
Communication with data users, the S&OP team is able to learn the needs and preferences of data users, is able to translate the technicalities of statistics and computing in terms understandable by users.
Understand how data will move between all relevant systems and people.
Knowledge of how the data could be represented, which describes how the data could be sorted.
Data transformation and analysis, when data becomes available to decision-makers, must be transformed by data scientists, summarized, made inferences and be able to communicate those results to those who need them.
Visualization and presentation, that is, choosing between precision and detail versus a good representation more effective to communicate the results to the data users.
The recommendation is that the coordinator or S&OP team should be the knowledge of data science or they should be a team with data scientists.
Originally published at https://medium.com on July 29, 2018.