Decision making and business analytics for supply chain
Within supply chains, a variety of decisions have to be made. Think of production decisions that include the number and location of factories to be used.
Or what about distributions decisions, such as what distributions channel to use, where to locate warehouses, what to organize in-house and what to outsource. Procurement decisions include make-or-buy and what mix of suppliers to use. Material flow decisions include what inventory to put where and in what quantities?
A decision making process contains a number of phases:
- Recognize the problem
- Define the problem
- Generate alternative solutions
- Develop a Model
- Analyze the alternatives
- Choose an alternative
- Implement the solution
For relatively simple and operational decisions these phases are carried out quickly and frequently. Think of a decision on what to do when an order delivery is delayed. As this may happen often, the problem is quickly recognized and defined and alternatives may be known upfront.
The model to evaluate alternatives could already be in place or built into the planning software. This makes analyzing and selecting alternatives relatively easy. For more complex decisions that happen less frequently and involve several companies in the supply chain the same steps need to be executed but will need more time and typically team work. Think of setting up a supply chain to produce and deliver to new markets. The problem needs to be carefully defined and information needs to be collected. A model can be built using simulation software systems to evaluated alternative solutions. Implementing such a decision can take months to even years.
ICT systems are available to support each step of the decision making process. ICT systems have also enabled that decision making can become a true team process and make use of different views and multi-disciplinary expertise. For example, in Sales and Operations planning (S&OP) the production plan is set and adjusted using current levels of sales and sales forecasts. In S&OP decisions have to be made that require marketing, supply chain, production, financial and human resource perspectives. ICT support can be very helpful to integrate various views and ideas and arrive at a balanced decision.
You can use ICT systems like performance dashboards monitoring customer satisfaction, financial performance, operations to recognize problems. Today, real time ‘big data’ coming from sensors, websites and social media can help in identifying problems early. You can use electronic meeting systems, group support systems, brain storm tools like mind-maps and concept maps to support the problem definition phase. These systems are increasingly offered as Software as a Service (SAAS) and support anytime anywhere collaboration. To model and evaluate solutions you can make use of simulation tools. These tools offer visualizations of the alternative scenarios. For complex problems that require innovative new ways of working simulation games can be developed as part of the evaluation phase. Making a choice can be supported by Negotiation support systems that enable debates (anonymous is preferred) and include voting functionality.
The possibilities of ICT to analyze problems, understand the factors that are at play and eventually build models that predict what will happen have progressed rapidly. In memory computing, powerful hardware and software has brought advanced statistical techniques and data visualizations within the reach of small and medium sized businesses.
In business analytics and business intelligence (BI), large internal and external data sets are integrated and analyzed to address the following questions:
- What happened (Descriptive analytics)
- Why did it happen (Diagnostic analytics)
- What will happen (Predictive analytics)
- How we can make it happen (Prescriptive analytics)
In addition to the decision support functionality provided by ERP systems and the world wide popularity of spreadsheets like MS-Excel, over the last few years we have seen a wealth of specialized business intelligence software tools emerge in the market. They offer increasingly advanced features for:
- data import, transformation, cleaning and integration (offering adaptors and links to include all kind of data sets, offer simple ways to clean and process data)
- data storage (in memory storage of large data sets for fast processing)
- data analysis (offer a wealth of descriptive, diagnostic and predictive techniques. Provide a rich set of interactive visualizations)
- model building (offer support for fitting models to data using for example machine learning)
- decision support and collaboration (offer support to build and discuss results as a team, e.g. through storytelling, offering chat and negotiation functionality)
In your next supply chain decision, think of the phases that make up the decision cycle. Think of how you can use ICT support to analyze and weight alternatives and make the decision more collaborative by inviting colleagues and external experts to join the decision making process.
© University of Twente