The chief data officer's revival: Demand rises for data leadership in the Middle East and Africa

At every turn, we are told that artificial intelligence (AI) is "the future". While the technology has certainly evolved in leaps and bounds and is likely to become a much bigger part of daily life, the multitude of conversations often ignores an important fact: to be truly valuable for businesses, any AI implementation first requires clean, articulated data and well-ordered architecture.

Despite the world's continually growing mountain of data, accessing and using it effectively remains a major challenge. To stand a chance of winning in the AI game — and indeed in other areas of driving value — organizations need to manage their data effectively and make it more than an abstraction. This inevitably calls for skilled data leadership in the form of a chief data officer (CDO) or equivalent. 

"Companies are struggling to generate business impact through data analytics and AI at a scale comparable to data and digital 'natives' — the likes of Amazon, Google and Uber," explained Vedran Karamani, group director of data analytics and AI at Alghanim Industries. "There are many reasons for this but one that stands out is a lack of ability to integrate analytical outputs and AI into operations."

If we study the chief data officer role, we find it is far from new. As long ago as 2002, Capital One became the first company to appoint a CDO and the role now exists at a reported 80% of large organizations globally. Yet its level of development and implementation is patchy. In the Middle East and Africa, the position is largely misunderstood and under-invested in. This should sound alarm bells for businesses in a fast-developing region that is under pressure to stay on top of global best practices and keep up with the rapidly increasing pace of technological change.

The history of the languishing CDO

Over the past decade, we have seen large companies hire CDOs to build out the data function on the premise that data is the "new oil" of the digital economy. A limited number of sectors, such as aviation and financial services, have committed to integrating the role into the organization and supporting its evolution. Despite good instincts, many others have failed to follow through, with the result that at a certain point, the CDO becomes devalued, downgraded or merged into a peer role within the technology function.

There are three main reasons for this lapse. One is perception. CDOs are by nature intrusive. To have an impact, the CDO must exert a level of disruption across functions and business lines to ensure that they are adopting the necessary data practices. It might mean a new obligation to log activities on a system as a sales leader. It could also mean dual accounting or data entry within the finance function. This can prompt resistance because it affects how people work. Depending on the level of pushback, the CDO can become a toothless tiger.

A further problem is uncertainty as to the positioning of the role within an organization. A Gartner study into the evolution of the role found that just 22% of data and analytics leaders reported to the CEO, while more than 40% reported to the CIO. In some cases, the CDO can find themselves reporting to the COO or CFO. Unclear reporting lines breed confusion around the role's objectives and can prevent the CDO from taking ownership and driving tangible impact. 

Then there is the issue of mismatched expectations. Effective data management is a long-term endeavor and investment, and a company cannot expect the CDO function to demonstrate success overnight. Setting up sufficient data warehousing, running data efficacy exercises and fine-tuning processes all take time. A company's current infrastructure will determine how quickly some of these processes can be established and actioned. If a company lacks the right infrastructure, it falls on the data leader to put it in place before a company begins to see the fruits of their labor. Without this understanding, the absence of quick-win opportunities can jeopardize the CDO's position.

How to set the CDO up for success

For companies looking to reignite the CDO role or even bring a CDO on board for the first time, it is important to recognize that the role has evolved considerably in recent years. In addition to having strong advanced software knowledge, CDOs today must also be adept at stakeholder management and influence. They must be able to secure constant buy-in across the organization, and this means explaining complex technology models in a language that others understand. They must also be able to interact comfortably at board level as boards place increased scrutiny on how their companies are using data as a competitive advantage. 

Reflecting on his experience working at Dubai Airports, Karamani noted: "CDOs need to be able to assess operational readiness for adoption and then prioritize initiatives and guide the boardroom in generating steady, gradually increasing business impact from data. They also need to partner with transformation officers in driving such change across the board.” 

Encouragingly, some boards are applying data to their own activities. Abu Dhabi-based AI company G42 Group is developing an AI-powered board member to analyze decades-worth of business data, financial information, market trends and global economic indicators. At a former company in the telco space, Karamani oversaw a CEO-led analytics initiative whereby the executive committee (exco) monitored performance and customer experience through almost real-time data scorecards. It was apparently "phenomenal to see the trickle-down effect and rapid process change it created across the group" as data worked to resolve speculation.

Engagement with and support for the CDO's activities at the topmost level is vital to their success and to give them a license to be "intrusive" without being shut down. Karamani added: "The organization needs to be able to champion change and be able to define and drive business strategy that clearly incorporates analytics and AI outputs into processes, policies and operating models needed to achieve business ambitions. This approach should naturally lead to funding and other enablement necessary for the execution of the CDO program."

In practical terms, an organization needs to make clear that “data is everyone’s problem, and everyone has a role to play”, said Mohamed Abdel Hamid, chief information officer (CIO) of Mashreq Bank in Dubai. But he stressed the importance of clarity in roles and responsibilities when it comes to data management. Hamid sees the roles of the CDO and CIO as interlinked, with the CDO requiring specific data management applications, data governance tools and data retention capabilities among other forms of infrastructural support. “The level of data maturity and roadmap must be agreed and the drive towards higher data management and maturity needs clear ownership.”

Miguel Rio Tinto, chief digital and information officer at Emirates NBD said that two factors have been central to positive outcomes. The first is the creation of a center of excellence that is blended into vertical teams and comprises data scientists, data governance teams and data visualizations. The second is a clear process to prioritize projects and determine impact. He added that teams must be "set up well to work together".

Even the most highly qualified CDO cannot hope to make an impact if the organization does not treat the success of the role as a collective effort. As well as being able to extend team members across each department, the data leader requires empowerment, vocal and practical support from the exco, and ultimately the patience of their company. The sooner that an organization embarks on the journey, the better their prospects for being able to implement and run the next generation of AI-based applications and solutions. For those without an active CDO, the clock is ticking.