alexa The state of big data in the Middle East and Africa - Daily Nation

The state of big data in the Middle East and Africa

Thursday February 6 2020

Digital economy

Big data concept. Digital economy encompasses a dizzying array of technologies and their application. ILLUSTRATION | NATION MEDIA GROUP 

FAUSTINE NGILA
By FAUSTINE NGILA
More by this Author

Data is becoming an integral driver of the Fourth Industrial Revolution as all emerging technologies – blockchain, data science, Artificial Intelligence (AI), Virtual Reality (VR), cloud computing and Internet of Things (IoT) – depend on it in the creation of solutions to socio-economic challenges.

A new report about the state of data and analytics in the Middle East and Africa (MEA) in 2020 explores the changing role of the chief data analytics officer (CDAO), how data leaders support business objectives, AI use and hurdles holding back organisations.

The report, An Exclusive Snapshot of the Opportunities and Challenges Facing Data and Analytics Leaders Today, reveals that having someone who is responsible for maximising the value of a company's data asset is essential for businesses operating in the digital age.

It indicated that more than 70 C-Suite data leaders have been hired in the MEA region since 2012.

"Going back five years ago, data and especially analytics were kind of a 'nice to have' for companies," says Louise de Beer, Head of BI and Data Science at analytics driven estate agent Leadhome, and one of three authors of the report.

"Then, if you look at reports from two years ago, innovators started playing with data as a differentiator. I think where we are now is that data is becoming a hygiene factor. If you don't have it, what are you doing with your business?"

Advertisement

The number of executives working in senior roles is a simple litmus test for assessing a region's data maturity. South Africa is a regional trailblazer in this respect, with the UAE in a distant second place.

The report suggests that a consensus is starting to form about where the chief data officer (CDO) sits in relation to the wider organisation. Gone are the days of the CDO reporting to the chief information officer (CIO).

Instead, 60 per cent of the CDAOs who responded to the survey answered to their company's chief operations officer (COO), 20 per cent report to the chief risk officer (CRO) and another 20 percent answer directly to the CEO.

As Mark Nasila, Chief Analytics Officer at FNB South Africa who also led the research work explains, data leaders need a mandate that spans the whole business in order to effectively drive organisational change.

"The business needs or use cases are very unique to each business unit. What I've done is to follow what is called a 'shared services' operating model. The strategy is centrally led, but it's executed in a decentralised way, at a business unit level," he told InDepth.

The role of a CDAO will naturally evolve with a company's data maturity. A company's first CDAO will initially be concerned with developing a data strategy and laying the foundations for data success.

But with this groundwork complete, their focus will shift towards enhancing the ways their organisation can extract value from its data asset.

The research shows that this evolution is well underway in organisations across MEA. A full 46 per cent of data leaders say their companies are in this transitionary phase, with 42.5 per cent reporting that their data teams are focused primarily on offensive initiatives.

Hartnell Ndungi, CDO at Absa Bank of Kenya (formerly Barclays), another key member of the research team, explains that people start visualising their data better, they start asking questions like, 'What will happen in the future?'

"In South Africa and East Africa, what I am seeing is that 90 per cent of CDOs are at 'level one'. They are focusing mostly on data management, data protection and data privacy," he reveals.

"Am also seeing a few CDO 2.0s, who are also able to talk about solutioning and coming up with dashboards and analysis to ensure that you're able to visualise and present data better, and very few CDO 3.0s," he adds.

The vast majority of the survey respondents said the purpose of their role is to drive business performance through advanced analytics.

According to the study, building a data science capability was the second most frequently cited 'purpose', tied with identifying opportunities for AI adoption.

Just 26 per cent of respondents said the purpose of their role is to implement sustainable data management and governance practices.

Data leaders are clearly looking to the future when it comes to articulating the purpose of their roles.

But some may be ignoring or downplaying the traditional data challenges that are still present in many organisations.

"They will need to balance their advanced initiatives with more foundational investments to give their data strategies the best chances of success."

Some executives may initially have resisted the idea that their instincts might not be the best tool for guiding their companies to prosperity.

But the Middle East and Africa's data leaders are successfully proving their worth and combating these old-fashioned ways of thinking.

The research shows that the region's data leaders are now laying the groundwork for data-driven business processes. Four in 10 now have a formal data strategy, and a further 32 per cent are developing one.

"The days are gone when someone could say 'I'm a very good leader' in finance or in business development just from experience," says Mr Ndungi.

"We are past the phase whereby we could only use our 'gut feel' and our experience to make decisions."

"Am not so sure if a lot of organisations have clearly defined data strategies which articulate where and how you're going to use data," adds Mr Nasila.

"They have to have a clear, defined data strategy." Developing a formal data strategy is a key step on the road to data maturity.

It's essential for aligning data initiatives with wider business objectives and agreeing a framework for change. The data leaders who responded to the survey are responsible for delivering on a range of different goals.

These include increasing revenue and market share, generating operational efficiencies, uncovering customer insights, improving the customer experience and reducing operating costs.

But in order to do any of this efficiently, data and analytics leaders must first take the time to define their strategies and use them to secure the leadership buy-in they will need to execute their plans.

While some organisations are starting to make data-driven decisions, there's still a long way to go.

Challenges

The research identifies five core challenges that data and analytics leaders must overcome in order to become truly data-driven.

More than half of data and analytics leaders are grappling with poor organisational understanding of the value of data, making this the most frequently cited challenge in the region.

"The hype has made things worse," Mr Ndungi explains. "When you hear a competitor is hiring a data person you run to the market and start recruiting without exactly knowing why you need one."

Breaking down data silos and centralising data storage is the second most frequently cited data challenge, suggesting that many data leaders are still working to give their programs the right foundations.

The fact that 35.6 percent of the survey respondents say they struggle to access enough high-quality data supports this conclusion. Meanwhile, 21.8 per cent cite a lack of leadership buy-in or support as a key challenge.

"I have heard of cases where people train 20 or 30 data scientists and they sit in the office drinking tea, doing nothing," notes Mr Ndungi.  

Of course, regional pressures such as currency fluctuations create additional challenges when it comes to securing the budget for data and analytics initiatives and acquiring the necessary technologies.

"Being a start-up and being so new, we just cannot afford to spend upwards of Sh350,000 per month," explains Ms Louise de Beer.

"It is one thing to sign up for Sh10,000 today. But in a few months' time, the same amount might be exponentially more," she continues. "So, hedging against that foreign exchange rate is an issue."

Finally, 29.9 percent of survey respondents say that lacking a comprehensive strategy is holding them back – underlining just how difficult it is to deliver on business objectives without one.

"I think a lot of CDOs are spending time focusing on the defensive side of the data strategy," says Mr Nasila.

"The other aspect is to align what you want to do with the data to your holistic business strategy. If they are not aligned, then you're not able to drive measurable value."

AI

AI is now well into its 'early adoption' phase, with businesses throughout the Middle East and Africa clamouring to launch new initiatives.

Almost three quarters of the data and analytics leaders who responded say their organisation has appointed someone to be responsible for AI deployment.

Two thirds say their organisations are exploring potential uses for AI, with 40.2 per cent of them saying they've at least reached the 'early implementation' phase.

Just 15 per cent say they are not looking to develop AI capabilities, 18 per cent want to learn more about them and 26 per cent are doing feasibility studies.

The potential benefits AI may bring to a business have proven extremely enticing for business leaders in the region, as Mr Nasila explains.

"We have designed algorithms to try to offer the right product to the right customer at the right time. We enhance this with financial crime models to make sure we are not offering products to criminals."

Given the prevalence of financial crimes in the region, developing AI capabilities to combat fraud and money laundering is a priority for the banking industry's data leaders.

FNB is acutely aware that combatting cybercrime is essential for developing good relationships with its customers, and machines can check for suspicious activity far more effectively than humans can.

"The bank is now using an in-house developed AI system to optimise the due diligence forensic review process," Mr Nasila expounds.

"This AI system automatically creates a single consolidated report with all the information required, which includes a single view for financial crime risk management.  So, instead of the analyst doing all this work, we're having the analyst just doing quality assurance and making the decision."

Initiatives like these have the potential to revolutionise the ways businesses provide goods and services.

But with Gartner predicting that 80 per cent of AI projects "will remain alchemy, run by wizards whose talents will not scale" through 2020, it's not clear that all organisations are ready to deliver these projects.

The failure rates for AI and other advanced data projects are high. According to a 2019 study from the International Data Corporation, 25 percent of organisations report that half of their AI projects result in failure.

Survey respondents cite unrealistic expectations and a lack of skilled staff as the top reasons for these failures.

"Once you've decided to have a data strategy, the first thing most people do is buy hardware and software," says Mr Ndungi.

"So, you go and start buying 'best in class' Business Intelligence solutions before knowing exactly whether the tools are really going to solve your problems."

Only 40.7 per cent of respondents in the survey plan to invest in data governance or management in 2020, and just 29 perecnt expect to invest in data quality or enrichment improvements.

Yet, these architectures underpin every successful AI programme. Given that data leaders in the MEA region are grappling with data silos and working to instill data culture in their organisations, the importance of these investments should not be overlooked.

The business may be excited about the potential of AI and advanced analytics, but organisations must make sure they have the right foundations in place first in order to succeed with these ambitious new initiatives.