While successful international businesses have long focused on customer experience as a major factor influencing a healthy return on investment, the same can unfortunately not be said for South African organisations who are letting this crucial business consideration slip – along with their potential profits.
This is according to Frans Cronje, MD of Cape Town-based tech start-up and machine learning specialist DataProphet, who notes that businesses like Amazon and Apple have great reputations for customer experience. “This is in stark contrast with the reputations of local businesses who may not realise the opportunities presented by their own data and how this can improve their insight into, and understanding of, their clientele.”
DataProphet has found that many South African companies don’t understand their customers at all, or even know which ones should be prioritised for support. “This leads businesses to marketing to all customers equally or building out-infrastructure without utilising their own customer information,” adds Cronje.
Having established that South African companies need to understand their customers better, Cronje suggests that the first step should be to perform customer segmentation in order to understand the different types of customers that exist in an organisation’s data. “Unfortunately this is where many companies stumble as they rely on old approaches developed when the data on customers was more limited. While these intuitions are often correct, there is more detail available in terms of different customer behaviours which can assist in forming a far better understanding of customers.
“Ultimately, it is not helpful to try and provide customers with personalised content if you only know where they shop. After all, each customer of any shop will go there for very different reasons – one may be buying a dress for a special occasion and would normally not accommodate this spend in their budget, while another may be buying a gift for someone else. Back then, the analytical techniques used did not need to be very sophisticated,” Cronje adds.
“However in recent years retailers have begun collecting far more data on their shoppers.”
With the data and IT infrastructure available to businesses today, he notes that it is tricky for those unfamiliar with the field to identify the best solution to get to know their customers and use this knowledge to provide great personalised content.
Cronje says: “A number of new mathematical approaches have been developed to handle the size of data collected on a regular basis. It is important to recognise that data collected today is very different from the data collected 10 years ago. This means that methods of gaining knowledge and information from data from 10 years ago, when techniques were developed to work on handfuls of inputs, no longer perform well in today’s reality where thousands of inputs from data are collected on a daily basis.”
DataProphet’s uniquely-developed customer segmentation solution combines state-of-the-art developments in machine learning and artificial intelligence to work specifically on usage data captured frequently.
DataProphet Commercial Director and Co-Founder, Daniel Schwartzkopff says, “Our solution determines the mathematically correct number of different types of users so that you know exactly how many types of customers you have while not relying on any prior assumptions.
“By applying this solution to customer usage data, we are able to identify different types of customers and provide insight into how they are using your company’s products, as well as how to serve them and which users to prioritise,” he adds.
Having applied this solution within a telecoms company, DataProphet was able to identify distinct user groups by their usage patterns – including where they were using the service and with what frequency. Identifying a distinct subset of the population which accounted for 18% of users – but over 40% of all upsells – DataProphet enabled the company to focus on that population through direct marketing.
In a separate case, the same solution was applied to an unsecured lending company which had proprietary information on their applicants. The solution was able to identify that nearly a third of applicants were being incorrectly served by the product and informed specific product development to better serve their needs. Usage patterns also noted which pool of applicants were most profitable, allowing the client to use this information in a direct marketing campaign to improve overall profits.
Cronje highlights that South African businesses can no longer ignore the importance of the customer experience in improving their bottom line – in addition to positively impacting customer loyalty and brand reputation.
“Industries that record customer usage data and who are well positioned to provide customers with personalised content – such as those in telecoms, banking, insurance and retail – have the tools and resources available to them to be more successful and competitive than they could ever realise,” he says.