What if you could know what your customers were going to purchase before they did? This is the promise of predictive analytics when applied to demand planning. Food and beverage manufacturers and distributors have approached this challenge by creating reports that show past purchases over time. While this can be a useful tool in the hands of an experienced purchaser it is often meaningless to someone new to the role. Understanding seasonality, the impact of holidays, weather and promotions can have a significant impact on a food and beverage distributor’s ability to prevent out-of-stock scenarios while not carrying too much inventory that results in waste. Trying to keep track of this at a macro level is one thing, but applying it at an individual store level can be a daunting task even for the most experienced purchaser – there are just too many variables. With machine learning allowing for this level of predictive analytics, we can now let the machine figure out the patterns and important variables.
How can you streamline this machine learning and predictive analytics as a food and beverage manufacturer or distributor? We have six steps to help transition these ideas into a repeatable process for your organization.
- Identify the Business Problem: The value in a predictive analytics solution is in being able to be hyper-focused on a specific business challenge. Maybe you’re a distributor of goods with a short shelf life or a long lead time and having a better sense of demand would improve your bottom line. It is critical to define not only the broad business challenge, but the specific problem that can be addressed through predictive analytics.
- Assess the Data: When looking at a business problem, it’s important to first determine whether data is available to address the specific problem. For example, if sales history at an SKU level for each store by day/hour for the last 2-3 years isn’t available, then it may be challenging to predict demand. When working with an organization we assess the impact of a business problem, but also the ability for that organization to solve for it with relevant data.
- Transform and Refine: With a clearly defined business problem and relevant data available, you can then work to transform and refine the data so that it is more useable and consumable by predictive analytics tools. This often involves data profiling to ensure we have a clear understanding of the data in question, as well as data enrichment to supplement with additional, related data that may not have been part of the original data set.
- Build the Model: This is where the heavy lifting of data science comes into play. Whether working on classification, regression, clustering or recommendation models, data scientists can use tools such as Azure Machine Learning Studio, R Studio and Cortana Analytics tools to create a model that exemplifies the business problem in a predictive fashion.
- Validate and Deploy: Once you have a working model, it is important to validate the results in a limited test to ensure that the model is working as intended. This typically takes place through a parallel process where we continue using the current method of planning or forecasting, but run our model alongside. Through this process, you can identify any potential opportunities for improvement in the model. Once any improvements are made we deploy the model through a web API.
- Incorporate and Evaluate: Once you have an effective model in place with a web accessible API, we can incorporate the model into our business process. This can be accomplished very simply by connecting the API to something like Excel for simple data entry or by incorporating it into a line of business application for real-time decision making.
Demand planning is not a new concept within the food and beverage industry and is perhaps something your organization already does well. However, our experience suggests that there may be additional profitability to be achieved from the demand planning process using some of the advanced predictive analytics capabilities available today. With recent innovations and technology, we can now take a human forecasting problem and turn it into a machine learning solution. Learn how BlumShapiro Consulting can help you achieve these goals.
Learn more about Predictive Analytics and Machine Learning