The Data Analytics Blog

Our news and views relating to Data Analytics, Big Data, Machine Learning, and the world of Credit.

All Posts

How To Apply Machine Learning In Your Business To Improve Your Revenue

January 16, 2018 at 8:17 AM

Machine learning is a hot topic. Although we are interested in the inner workings of machine learning and how to improve our models, we are as interested in how to apply machine learning to improve business’ revenue and ROI. The positive impact on ROI through the application of machine learning techniques is well known to us. Based on our experiences, here are the most influential business applications of machine learning. These are the areas where we would recommend you start introducing ML in your consumer-focused business.

Identifying consumer marketing leads

If you have a large lead list, you’ll know that some of the leads are never going to respond, while others will have a much higher propensity to respond. And not all of these leads would turn out to be ideal customers. By developing machine learning algorithms from historical campaigns and running your current data through these algorithms, you’ll be able to generate leads lists to target your most desirable consumers and those profiled as most likely to respond to your campaign. This information will allow you to prioritise leads, and prevent effort spent on leads who will never take up your offer. Machine learning algorithms can also provide recommended action and execution options for each contact to ensure you’re not only contacting the right people but also communicating with them in the most effective way and via the right channel.

Our past experience has shown that marketing leads that have been through a selection process are often more than twice as likely to respond when compared to more traditional approaches. Not only will your campaign response rates increase, but your cost of acquisition and bad debt costs will decrease, as you will only be targeting your most desirable customer profiles.

Improve customer retention and re-activation

It costs you quite a bit to on-board a new customer and you really don’t want to lose them to your competitors (especially not your good customers).  Focusing efforts on retaining customers is a common machine learning application.  You can use ML to identify which customers are most likely to churn and out of these, which ones are worth spending time, money (e.g. vouchers or special discounts) and effort on to retain or re-activate. One of the caveats with retention strategies is to make sure that an intervention on your part is not the trigger that results in them leaving you.  That would result in your retention strategy backfiring.

However, machine learning can certainly increase your marketing ROI and retention rates through improved targeting and optimised prioritisation.

Increase customer growth

All businesses are faced with the ongoing challenge of how to identify customers with the best potential for growth, what to offer them and when to do so. Machine learning can help you identify these customers and recommend which offer they are most likely to respond to. This helps you retain and grow your profitable customers, maximise your share of wallet with customers, improves your upsell and cross-sell rates and, ultimately, increases your marketing ROI through improved targeting.

Identify the best time to call customers

You can use ML to predict the best time and the best number on which to contact customers for collections, upsell or cross-sell opportunities or to prevent churn. Then you can prioritise calls by time of day when a customer is most likely to answer. This will not only reduce your telephone costs but will increase connection and conversion rates, by improving your call centre agent effectiveness. Dependent on the purpose of your calls, this will result in improved retention rates, higher collection yields and increased share of wallet. The only requirement is that you have historical dialler information at a number level that can be fed into the model to make time of day predictions.

Identify propensity to pay or roll

Signing up new customers or retaining current customers doesn’t help if they default on payments. But classifying risk is another great machine learning application. You can identify the accounts most likely to roll forward to a worse delinquency state and those most likely to pay at any given time and prioritise calls based on their probability to pay or roll. In addition, this will allow you to determine the best treatment for each account, achieve higher collection yields and decrease your operational costs by increasing your call centre agent efficiency and effectiveness.

These are the five machine learning business applications that will have a big impact to your bottom line quickly and are great ‘starter’ machine learning applications. Using these in combination optimises the effectiveness, but even implementing one application in your business today will have a large impact. Read more on Machine Learning as a Service if you are interested in deploying ML without needing to grow your own data science team and without the need to invest heavily in additional infrastructure costs.

contact us

Robin Davies
Robin Davies
Robin Davies is the Head of Product Development at Principa. Robin’s team packages complex concepts into easy-to-use products that help our clients to lift their business in often unexpected ways.

Latest Posts

Our Top Credit Risk Management Blogs From The Past Year

In our third roundup of 2018, we list the most popular blogs from our Credit Risk Management topic. Our credit risk blogs have mainly revolved around IFRS 9 and legislation changes this year, and have aimed to provide expert advice and insights into the current landscape.

Best 2018 Data Science Blogs

For the second roundup of our most popular 2018 blogs, we cover the Data Analytics topic. The blogs that our readers love are thought-provoking and aim to inspire and teach – and we hope these articles do just that!

Our Top Debt Collection Blogs From 2018

We end off every year with a roundup of our most popular blogs – the blogs that you should not miss out on! This year is no different, and we'll be putting together four roundups this year, one for each of our main topics.