Why business must adopt data science

Chalenge Masekera
BIG data, data science, data mining, machine learning and artificial intelligence are currently buzzwords in the world of technology.

So what is the hype all about and what does this mean for business in Zimbabwe?

By adopting data science businesses stand a chance to be able to understand and predict customer behaviour and system processes in simple and faster ways.

Broadly, data science refers to the use and conversion of data into knowledge and actionable insights.

Data science applications cut across all major industries of retail, banking, telecommunications, health, insurance and manufacturing.

All these industries that embraced ICTs some years ago and record their transactional data daily in computer systems are prime candidates for data science.

As a business, figure out a strategic and key business initiative that is critical to immediate term performance, define KPIs and metrics that impact it, collect the data from your system and build a system of intelligence around it.

For example, as a retailer with the current liquidity crunch, you might want to optimise inventory levels.

Using purchases and sales information, you can build an intelligent system that can predict demand for thousands of products for any time period which can save your company money.

You can even go further and do market basket analysis to understand products that get bought together, do smarter shelf allocations and increase cross and up-sells.

Fraud detection is a typical use case in insurance.

In telecoms, by analysing factors such as number of calls and texts made and received, patterns and probabilities can be established for those who are likely to leave the network.

Besides this typical use case in telecoms, cell phone usage data can be used as a proxy to predict other things such as a user’s economic status or likelihood to default on a loan.

From a business standpoint, a bank like Steward can use data from Econet to make better loan decisions.

Manufacturers, on the other hand, using data on how long equipment has been running and when it was replaced, can predict when equipment is likely to fail; thereby saving on expensive downtimes.

Data science is not restricted to major businesses.

There are multitudes of opportunities for SMEs, NGOs and Government as well.

This debunks a key common misconception that data science needs big data or huge amounts of data and massive computational power to implement.

A lot of data science applications don’t have huge amounts of data.

No matter how small the application, by using data science even with little data, computers tend to learn faster and be better at making predictions and recommendations than human experts.

In terms of cost, data science is one of the few tech fields where most major software components are available free through open-source communities.

With the right people, all you have to do is download, build and customise for your use case.

Whilst we may not ultimately rely on machine intelligence for all decisions in the near future, research has shown that some combination of human and machine intelligence often yields better results compared to just just humans.

The real beauty in the Information Age is not in fancy gadgets or plain automation but the use of technology to process massive amounts of data and build smarter systems for smarter decisions that improve business performance.

People have to be smart and resort to technologies that can easily improve lives.