Machine learning offers a broad spectrum of solutions to multiple business challenges, as a result, it is often difficult to highlight the benefits of machine learning without generalising.
Often the benefits are simply stated as ‘adopt machine learning to be competitive’ without any real detail, and that once adopted machine learning will help to automate your workflow, whilst providing data driven insights that can help your business. While this may be true, the challenge remains of articulating precisely how it will benefit them. The easiest way we have found to illustrate the benefits of machine learning is to provide customer scenarios we have provided a machine learning solution to, basically helping to unlock the magic in your data.
Customer scenarios
Asset Management – An asset management company had lost millions of dollars of revenue due to retail premises remaining vacant. We developed machine learning models that were able to predict when customers were going to vacate with over 90% accuracy. This provided the business with the capability to plan ahead and have new tenants prepared to fill these vacancies, thus significantly reducing lost revenue.
Risk Analysis – A large company had identified a critical issue where subcontractors were underquoting on bids, that would then go over budget. Using machine learning we established the level of risk for subcontractors of going over budget and were able to predict this with 85% accuracy. This allowed the business to make significant cost savings by choosing low risk subcontractors during the bidding process.
Customer Attrition - An online services company noticed their customer attrition was leading to a major loss of revenue. Using machine learning we identified customers at risk of churn with 95% accuracy. The benefits were two-fold, firstly providing a monthly list of ‘at risk’ customers, then our data science methods established what factors caused, or prevented, customers from leaving. This allowed these customers to be targeted with an offer that appealed to them specifically and prevented them from leaving.
Anomaly Detection - A company had particular financial transactions that needed to be flagged in order for them to get tax rebates. Due to the size of the data set this took multiple staff weeks to identify these transactions. Machine learning models we developed identified these transactions within minutes. This enabled staff to focus on value added tasks, and resulted in an increase in tax rebates.