Skip to content Skip to footer

Seven Considerations to Unlock More Value from ML/AI Planning Solutions


Share this post

Published: March 16, 2023

In today’s fast-paced business world, demand forecasting is critical for companies that want to stay ahead of the curve.

With consumer behavior becoming increasingly complex and volatile, many companies are turning to ML/AI-driven planning platforms for help. But how do you make the most of these solutions?

Here are seven considerations to unlock the most value from your ML/AI-driven planning solution.

Data Requirements

1. Start with Master and Fact Data

Typically, two types of data must be collected to begin training the predictive models of a machine-learning-driven planning solution. Master data, which refers to static structures such as product names, numbers, descriptions, and categories, and fact data, which represents events like date of sale, promotions, logistics, price, and seasonality.

2. Plug into External Data Sources

Historical sales data, a type of fact data, has long been the fuel for traditional forecasting engines. ML/AI-driven planning platforms add another dimension to improving demand forecast accuracy by ingesting data from sources such as weather APIs, competitor promotions, and social media, and product reviews.

Required Skill Sets

3. Combine Business Analysis and Data-Science Expertise

To turn data into demand forecasting insights, features like data collection and harmonization are a must. Equally are employees with sufficient data analysis and evaluation skills. Ideally, companies should have a data scientist and a business analyst working together to gather insights. “If you can have data science and business resources, that’s perfect,” said Simon Joiner, Director of Product Management at o9 Solutions. “But typically, you need data-science and business experts and they have to work together to get the best blend of data analysis and business benefit.”

4. Hire Planners Who (or Train Planners to) Understand the Language of Data-Science

In the digital world, planners need to react to the data that the ML/AI planning solution is providing them. Skills like data visualization, understanding variable importance, and having the ability to run scenarios based on the solution’s output are becoming increasingly important for planners. Equally important is having people that understand data usage, given that many traditional planning applications can’t load, calculate, or display external driver data. R/Python skills are also beneficial because open-source solutions are a growing trend. “Black-box” forecasting systems can only evolve in ways the solution provider understand. Open-source systems, on the other hand, can be adjusted whenever a planning breakthrough is developed, and planners should be co-creators in this process.

Effective Change Management

5. Ensure All Stakeholders—from Planners to Executives—Understand the Value

Due to the increased processing power of planning solutions, some demand planners may want to shift their planning cycles from monthly to weekly or even daily. However, this shift comes with its own set of challenges, one of which is the issue of trust in ML/AI solutions. “There will no doubt be resistance to change among planners who are used to relying on their instincts, but if the implementation of machine learning forecasting is well-scoped and its potential value is understood by stakeholders, this can be overcome,” said Geoffery.

6. Assess your Planning Maturity and Create a Roadmap

To take full advantage of these technological advancements, companies must assess their current level of planning maturity and create a roadmap to progress in their demand planning journey. This involves incorporating market intelligence, knowledge models, and ML/AI-based analytics into their business operations and leveraging real-time data to make more informed decisions.

7. Get an Executive Sponsor on Board to Articulate the Business Value

Success in an ML/AI project also depends on having a key business sponsor who can link the project to tangible business cases that leadership can understand and support. “You have to link it back to the money, not just to the data,” said Rahul Teotia, Executive General Manager & Group Chief Supply Chain Officer at Pact Group Holdings Ltd. “That’s where the success will come from. Sixty-five percent of the time needs to be invested up-front in getting people on board, and once that’s done, implementation can be quick.”

Recommended for you

Dennis Donelon

Former Sr. Director Cust. Supply Chain Integration

Cross-IndustryAI/ML AnalyticsBest practices

Why Visualizing Data Makes all the Difference

by Dennis Donelon
Pruks Itchayapruks

Discovery & Strategy Manager

Process ManufacturingAI/ML AnalyticsControl TowerSustainability

The State of the Oil and Gas Supply Chain

by Pruks Itchayapruks
Harshal Harkison

Sub Saharan Supply Chain Dir., South Africa

Cross-IndustryAI/ML AnalyticsBest practicesLeadership

Generating Supply Chain Value with AI

by Harshal Harkison
Cross-IndustryAI/ML AnalyticsBest practices

Turning Data into Insights

by aim10x