Jump to content
  • Data Driven Decision Making

    In the FMCG sector, our company is rated among the top 5 organisations both by the consumers and industry. Much of the success is attributed to the increased consumer confidence and round-the-clock shelf SKU availability of our product line. The role of our Purchase Division has always been at the core for achieving this mark. Over the years our team under your guidance has been able to keep the inventory at optimum levels thereby helping the Production Unit to make delivery on time. It is due to the careful execution of our fundamental function - 'Material Management' which ensures that raw material is procured from the suppliers at the right time and in right quantities. This is the key to success in our sector and thus any minuscule improvement towards this can play an important role in raising our organization's bar.

    Our well maintained database of daily Material Management activities can help us in this direction for making more accurate estimations. An approach-'Data Driven Decision Making' can be applied in which assessment data and background information can be used to take decisions related to planning activities. According to a study by MIT Center for Digital Business, organizations driven most by data-based decision making had 4% higher productivity rates and 6% higher profits. With data driven decision making we can deploy Just-In-Time as approach and Material Resource Planning as the method for waste minimization in the purchase cycle.


    To demonstrate the strategic, operational and financial advantages of DDDM approach, we can consider the following conventional process being carried out at our unit. Under usual circumstances, we have to keep a safety stock of inventory items so as to counter the market and supply chain uncertainties. These may include logistics delay, plant failures, supply side variability, demand side variability and others. This leads to estimations made with the help of brainstorming methods within the team and experiential knowledge that cumulatively determine the amount of inventory to be ordered. With the help of estimations, performance is delivered in terms of fulfilling various conditions like on time delivery, safety stock, safety time related to inventory items. But there could occur estimation errors during the process due to which stores or warehouses can have lower or excessive amounts of inventory. This can affect the overall process efficiency both in monetary and operational terms. To improve this scenario, Data driven decision making mode can be deployed. Considering the situation that we faced last month. The production unit had given a demand forecast of 100 packs for one of our products and keeping safety stock levels in mind, we had ordered 120 packs. But we received only 115 packs from the supplier due to production based variability at his end which had not been a factor of consideration in our process. Also, the actual demand turned out to be 118 packs which was closer to our safety stock levels. So, overall we faced a shortage of 3 packs. From this we can identify that there is a chance of around 4% variability in delivery from supplier's end. This data if incorporated in taking decisions before ordering again from the same supplier can help us in achieving more accurate outcomes. This was an example of countering lower inventory received from the supplier where DDDM could have helped otherwise. Considering another situation that occurred 6 months back where we had ordered around 200 packs where as the production had ordered for 120 packs only but finally picked up around 150 packs from the warehouse. We had a rough idea of up-scaling from demand side as well as in the costs from the supplier's end due to which we increased the order. Though a more accurate estimation using DDDM would have given an additional gain by reducing the wastage of the remaining underutilized 50 packs as well.


    Using DDDM as one of our modus operandi would be convenient and beneficial too as its output depends on the quality of data gathered and a well managed database already in place can help us in reaping maximum benefits out of this investment. Further its effectiveness depends on defining the questions to be considered before analyzing the data and with your experience level in this area, we can easily frame pertinent questions to get relevant results from the data. Data driven decision making would give us added advantages of faster processing, refocusing our resources to increase the yield, relevant data backing to explain our rationale behind purchase decisions to the management, foreseeing the opportunities and threats in the market and overall supply chain. Over a period of time, this approach can also lead to building a reliable group of suppliers giving us a competitive advantage gained by adopting a data backed strategic purchasing model. An early adoption of Data driven decision making would bring maturity to our supply chain infrastructure and resilience towards unforeseen circumstances so that we can quickly respond to them without compromising on financial and operational aspects.


    Note - Visitors shall not be able to comment on this article until they are logged in.

    User Feedback

    Recommended Comments

    Good article Swatee!! 

    I have but one question . As DDDM is based on past data , how will it bring resilience towards unforeseen circumstances ?


    Hi Saumyadeep.

    Thanks for your interest in this article. Below I have tried to answer your query that how DDDM can help in unforeseen circumstances.


    DDDM is essentially a field associated with data analytics in which past data is gathered and used in predictive models. These models take into consideration various environmental variables and based on past data give a probabilistic numbers to unforeseen situations. These probabilistic algorithms give us some confidence level and find out what can go wrong. Thus, with data a manager can achieve a considerable insight into futuristic situations. Though there is always scope for inefficiency but that could also help in long term as that data would also be captured and would become a part of DDDM bringing greater accuracy the next time. 



    Link to comment
    Share on other sites

    Hi Swatee,


    Its a wonderful article. Could you please give examples of situations where Data Driven Decision Making should not be applied.



    Kunal Nichani


    Hi Kunal,

    Thanks for appreciating the article. Below I have tried to answer your query about the scenarios where D3M finds less relevance:



     When there is time constraint and delivery takes precedence over analysis then D3M approach can not be applied.



    When the manager has to carry out operational tasks which are of day to day nature, then D3M might delay the process as it takes time in analysis.



    When the resources involved in the process are not available easily then the manager has to take a call and proceed with the available means and not apply D3M.


    There could be more such situations where trade off needs to be done between time and delivery, then whether D3M should be applied or not largely depends on the factor which holds more relevance.



    Link to comment
    Share on other sites

    Hello Swatee,


    Could you give some examples of software packages which could be used for DDDM purpose? E3 ( Automatic Warehouse Replenishment System ) is used in many FMCG companies in US for Demand Forecasting. Does it fall under DDDM ?


    Hi Jayati,

    Thanks for your query. Below I have tried to answer the same.


    D3M process is performed largely with the help of data analytics software packages. In these systems, data is fed and they deploy predictive models such as Bayesian Plots to predict future demands and supplies.

    Example: JDA provides supply chain softwares for Data Forecasting and Replenishment. It considers the historical data,current balance on hand, on order to provide forecasts for placing orders. Its underlying mechanism is D3M only.


    Also, the example stated by you of E3 systems which are automatic warehouse replenishment systems are also a good example of D3M as they perform the calculations on the basis of data.



    Link to comment
    Share on other sites

    Does DDDM prescribes forecasting and production decisions to be automated based on past sales, production and procurement data? If not then there is a possibility that even after having all the data the production team takes there own decision based on their experience and gut feeling. Also it is not that current ERP systems are not data driven , what is new in this DDDM concept?  

    Link to comment
    Share on other sites

    Does DDDM prescribes forecasting and production decisions to be automated based on past sales, production and procurement data? If not then there is a possibility that even after having all the data the production team takes there own decision based on their experience and gut feeling. Also it is not that current ERP systems are not data driven , what is new in this DDDM concept?  


    Hi Soubhik,

    Thanks for your interest in the submission.


    The points mentioned above that DDDM is based on past sales, production and procurement data is true and its benefits are immense. The article highlights the advantages that could have been achieved in the two cases occurred in past where in the manager had taken decisions based on gut feeling instead of D3M approach. 

    Experience along with data backed analysis gives much more efficient results and reduces uncertainties to minimum. So, D3M approach should be deployed as far as possible.



    Link to comment
    Share on other sites

    I have a question as well. Can DDDM be utilized for small scale family business too? Are there any relevant softwares available which I can buy online?

    Hi Sonhal,

    Thanks for bringing forth an interesting perspective that whether DDDM can be applied for small scale family business or not.


    The decision for applyiing DDDM is based on various reaqsons. Some of them are:

    1. Increase the scale- Current Coverage and Potential capacity

    2. Huge data to analyse

    3. Earlier insights into future situations

    4. Increase efficiency and thus drive profits

    5. Sound backing for decisions taken


    If these are the reasons present in any case then DDDM can be applied in the case as these can be eaily fulfilled. For a small scale business, DDDM can even help further to increase its scale to medium and large at later stages. The software packages available for data analytics can be used for any amount of data and would help in predicting outcomes accordingly. One can go for customized software packages so as to suit the requirements and save on investment as well.



    Link to comment
    Share on other sites

    This is now closed for further comments

  • Create New...