How MTN became a data-driven organisation

Read time 3min 30sec
Michael Bell, GM of BI and analytics at MTN.
Michael Bell, GM of BI and analytics at MTN.

Before MTN embarked on a journey to implement its unified business intelligence (BI) roadmap initiative, the mobile operator was drowning in the volume and velocity of data.

This is according to Michael Bell, GM of BI and analytics at MTN, addressing delegates at ITWeb Business Intelligence and Analytics Summit 2019, in Johannesburg yesterday.

Discussing MTN's journey to becoming a data-driven organisation, Bell pointed out the telco's various voice, data, mobile money and music streaming service offerings aimed at connecting people and communities across the Middle East and Africa resulted in the telco dealing with a consistent "big data headache".

"With 233 million mobile subscribers, over 27 million mobile money subscribers and a growing enterprise and digital business across Africa and the Middle East, MTN was dealing with a massive amount of data. All these projects contributed to our 'biggest headache' in terms of big data.

"With over 10 billion call data records stored a day, data plays an important role in us [MTN] driving agility and growth, through our connectivity, communication and collaboration solutions. As our business grew, so did the data and we were forced to re-assess and look at how we could better handle it and prepare for the challenges of scaling."

One of the operator's fastest growing business units, according to Bell, is the enterprise business.

Inflexible legacy system

Bell explained that in the last couple of years, MTN's business and customers grew quickly and the telco's previous IT systems were no longer sufficient to handle large amounts of data, forcing the operator to implement a research programme that would see it deploy a new data management platform and other initiatives.

"Our legacy BI system started falling over as it was rigid, inflexible and not scalable.

"So we introduced a unified business intelligence roadmap initiative to understand the current MTN BI environment and its challenges. Through the study, we got a holistic view of where we are and how we could scale as the business grew and as data grew exponentially."

The research, conducted through workshops, interviews, cross-departmental verifications with various business units in various regions, recommended the telco had to drive transformation in four key areas, noted Bell.

"Our vision was to enable MTN to become a data-driven organisation. First, we had to create a BI competency centre consisting of a cross-functional organisational team that has defined tasks for supporting and promoting the effective use of BI across the organisation.

"Second, we had to implement a data platform to structure our data lake. Third, we needed a sound data governance strategy, and fourth, we had to implement an agile approach for quick action on the operating company function feedback and also implement effective transformation management mechanisms."

The big data platform, named EVA (Enterprise Value Analytics), was segmented according to various features, capabilities and real-time decision-making requirements. Algorithms and machine learning tools are embedded on the platform to draw and standardise data from different elements of the business quickly and efficiently, enabling powerful analytics.

"Some of our data programmes entail data-as-a-service, the massive call centre systems which drive insights through live dashboards and reports, self-service BI initiatives, and campaign management segmentation.

"In future the platform will be able to provide us with a 360-degree view of the customer, where we will track their activity across various digital platforms, provide profile tariff optimisation data and eventually integrate artificial intelligence, machine learning and Internet of things tools to draw insights."

EVA is live in four countries, with the remaining countries being rolled out over the next two years.

"This is not a project; it's really a business transformation process," concluded Bell.

Login with