In our current digital age, insights into Artificial Intelligence (AI) applications like machine learning can help businesses target the right customers and deliver ‘big value’ from their data. With this innovation at our fingertips the question arises, are South African businesses optimising their data resources?
It is important to consider why companies started collecting data.
Historically, data has been collected to report back upon either for legal requirements such as fulfilling Audit requirements, or for their own internal monitoring purposes. This legacy still influences many companies in what data they collect and how they store it. The result is that much of the data held by companies is not ready for predictive modelling and machine learning.
Subsequently, it should not be surprising that much of the innovation in predictive analytics and machine learning is driven from younger companies that were built when cell phones, laptops and easy access to the internet were commonplace.
An important consideration when it comes to data is the wide range of opportunities it enables, particularly for service-related industries which can use it to identify consumer preferences and, in turn, help in detecting where products or services can be improved. Almost every industry can benefit from compiling and building data from the education, transportation and consumer products sectors, to businesses in electricity, oil and gas, healthcare and consumer finance like banking and insurance.
Ultimately, instead of relying on intuition, companies who handle their data correctly can embrace predictive decision-making approaches, which – when coupled with automation – can provide cost savings as well as profit gains.
So how can businesses get the most out of their data? Considering data from the predictive point of view can help businesses realise how to improve their data management. When taking this perspective, it is easy to identify the veracity of system log data is really valuable or overwriting data to provide a current snapshot of the data can mean the data is no longer valuable for predictive modelling.
The problem and solution should take the company’s unique environment and challenges into account. How fast does a result need to be returned, where does the data arrive from and at what frequency?
We prefer co-location modes of consulting with our clients which helps us understand each client’s domain and allows us to create a solution that can be delivered relatively quickly through a small team.
In terms of whether or not such capabilities can – or should – be out-sourced or developed internally, many organisations simply do not have the internal skills to implement machine learning applications. However, there have been developments locally to address this gap.
DataProphet is one of several collaborators involved in the design of the Postgraduate Diploma in Data Analytics and Business Intelligence which will be offered by the Faculty of Economic and Management Sciences at the University of the Western Cape from January 2017.
- Frans Cronje, Managing Director for DataProphet.