Can you really trust your data?

Can you really trust your data?
Christopher Tredger
By Christopher Tredger, Portals editor
, 02 Feb 2017
...

South African and African business leaders perceive value in using data and analytics (D & A) but are challenged by a mistrust of analytics used as well as a lack of measurement tools and relevant skills.

This is according to Frank Rizzo, Data and Analytics leader at KPMG in South Africa.

The auditing and services firm released research in the form of a report Building Trust in Analytics, which is focused on the extent to which businesses lack trust in their D & A models and processes to drive decision making.

Amid the ongoing hype around big data and its ability to solve problems, as well as global attention to the issue of fake news, KPMG's view is that clever analytics is all good and well, but it is critical to look closer at the source of data that is being used.

"As we all know no data is perfect, but it's all a question of how good or accurate is that data for us to then draw some insights and conclusions from," says Rizzo.

Recommended

The survey was conducted to test whether or not businesses can trust the analytics that teams are producing.

"Our view is that while the analytics may be fantastic and robust, the data quality is potentially lacking and this survey shows that," says Rizzo.

South African respondents tended to be less trusting of data quality than their global counterparts.

From a survey of 2, 165 respondents from 10 countries, 51% of businesses in South Africa use D & A tools to analyse existing customers, 42% use these tools to find new customers and 49% to develop new products and services.

"Yet executives do not trust that they are managing their D & A processes effectively to generate desired outcomes and lack the necessary measures to assess the efficacy of those models," states KPMG in its report.

Africa analysis

In the case of Africa multiple sources of data, most notably social media, is driving the uptake of data and analytics technology.

"I am in close contact with our colleagues in Nigeria and Kenya, specifically. There are two contrasting things, there is an incredible appetite for data and analytics ... and I have been saying for a few years, in Africa we have an opportunity to leapfrog a lot of other developed economies because we don't have legacy systems per se. If you are engineering new solutions, you build it around data. On the opposite side, there is a real scarcity of data. I think that's due to the collection mechanisms," says Rizzo.

Chief among these collection mechanisms is social media, which, according to Rizzo will only gain prominence as a result of increased smartphone penetration.

Companies in South Africa are rerouting the power of data analytics fuelling the establishment of internal data (master data, sales data) from the organisation as a data supply channel, Rizzo adds.

"The third one is what we call 'sensor data' or data travelling from the field based on whatever business you are in," he continues. "For example, if I am operating shopping malls in various African cities, I can use the sensors to track traffic patterns in my shopping malls. These things are becoming more sophisticated because you can use customer's mobile phones as a proxy of the customer to see how they move in a shopping centre."

However, this type of data sourcing and application is not yet mature in the rest of Africa, mainly because of infrastructure issues says Rizzo. "I am generalising, but you are missing that sensor data input."

Sensor data would be able to validate what is being seen on social media, and that is still in short supply, he continues.

Rizzo predicts that M2M and sensory connectivity is likely to explode in Africa, and organisations that will win will be those that have set up their channels or platforms to collect and leverage this.

However, with that rapid growth comes the issue of trust in the quality of the data itself, as well as access to available relevant and sought after technical skill sets.

The proliferation of cloud computing, with shared platforms and systems, is an enabler and making it easier for companies to collect and analyse data without having to set up infrastructure.

Share

Read more