How finance leaders can drive better data analytics

How finance leaders can drive better data analytics

Many organisations have spent millions on data analytics technologies but often achieved disappointing returns. This presents a major challenge for finance professionals as they often play a key role in data transformation projects with many CFOs sponsoring them and taking the lead.

One way finance leaders can improve their impact on data analytics is expanding their knowledge in the area. Most finance managers believe data analytics will be essential for business survival, but they lack the necessary knowledge that underpins it, according to UK research by MHR Analytics.

In a poll of 500 senior finance and technology managers, 55% said data analytics will be essential in the next ten years, but 40% said skills gaps are a barrier to advanced analytics. Meanwhile, 43% of senior professionals say they will need to learn data science or analytics skills to progress their role in the next five years.

Florian Mueller, head of management consultant Bain & Co.’s advanced analytics team in Europe, said unless the company has a chief data or analytics officer, CFOs will often sponsor data analytics projects and provide senior leadership as part of their business strategy role.

“They’re often gatekeepers of data transformation and have an important role due to their independent stance in the organisation and ability to provide the right environment by freeing up budget,” said Mueller. “But we see many CFOs struggling to cope with their responsibilities related to analytics.

“The reason is often a disconnect between data analytics initiatives and strategies. When setting the analytics agenda, businesses need to identify their immediate needs and priorities, future vision, what assets they might need in the future, and where their plans might be disrupted.”

Mueller said many companies make large investments in analytics that lack focus on data quality and data integration. “Without those basics, analytics at scale with tangible business results are not possible,” Mueller said.

But the biggest issue CFOs face is getting buy-in from the people who will use the systems, Mueller said.

“Ensuring that people can work with the analytics outcomes you’re generating is hard,” he said. “It requires a strong sponsorship spine and a broad change management approach.”

Alex Vayner, Americas data science and artificial intelligence lead at PA Consulting, said the reason many companies struggle in this space is that they continue to approach mining and managing data as the goal. Solving business pain points is often an afterthought once all the data is acquired, structured, and cleaned.

It is vital, Vayner said, for CFOs to switch from a “bottom-up” view of managing data towards a “top-down” view of solving problems. This can help identify the critical data elements needed.

Vayner offered the example of a CFO facing discrepancies in sales reporting from different divisions and regions. “Instead of declaring a need to create a separate ecosystem, which will take lots of money and time, they should instead ask to fix one division in one region,” Vayner said. “From that, they can learn about discrepancies in data quality, consistency, and accuracy that can then be generalised.”

Critical misconceptions

Ireland-based John Pearson, ACMA, CGMA, has worked in IT risk and assurance and data management for almost 15 years. He has seen many of these issues playing out across projects in different organisations. He said that he has seen organisations expect that investing in technology alone will be the key component in solving problems and getting value from data. More times than not, this is a critical misconception.

“New systems do enable more rapid deployment and have better data functionality and interoperability,” he said. “But organisations must still design them carefully for a specific business use and incorporate a well-understood, end-to-end data model that is owned and managed on an ongoing basis.

“To get the best value from our people, processes, and technology, end-to-end data management is crucial and must be viewed as the common thread between these key value areas.”

Pearson said he has seen several pitfalls leading to companies not realising the benefits of data projects.

The first is poor planning around data models. Unless companies build out a clear map and lineage of their upstream data, it is unlikely they’ll fully leverage the functionality of any new data technologies and systems.

Some do not consider the impact on integrations with downstream systems and processes.

They may often lack understanding of the existing data model, leading to poor design choices. And they may not carry out planned data cleansing and enriching to match the required data inputs and reporting and process outputs.

Pearson said another pitfall he has seen is that companies lack a formalised data governance and ownership model across key processes and systems and do not have a centralised team to manage data.

Finally, some also have inadequate mastering and quality issue detection controls early in the process life cycle. This can lead to issues such as duplicate customers that can be time-consuming to resolve, Pearson said.

Making progress

Ireland-based Shane McArdle, FCMA, CGMA, is director of finance and Brexit readiness lead for retail company Boots UK and Ireland. He said that forward-thinking finance departments are making progress.

“Boots is good with the vast amount of data we gather,” he said. “That said, digital transformation is an ongoing project. I suspect much of the disappointment [in other companies] stems from not understanding the true goal when investing in technology to mine and manage data.

“Given the speed of … disruption, the danger is a reactive, knee-jerk focus on ‘doing something’ because the risk of doing nothing is larger. But without a well-thought-through data strategy, that investment could be squandered.”

McArdle said CFOs can address this issue by ensuring their data strategy is aligned collaboratively. It cannot be created in silos, nor should it involve senior leaders only, but also should include other people throughout the organisation who can be agents and champions of change.

McArdle said designing a holistic data strategy requires understanding of data capture, management, and interpretation across the business. For example, customer data can take many forms, from personal loyalty scheme information to satisfaction of experiences in store or online.

Also, ensuring the right intent and processes are in place for deriving insight from data is important. This includes a culture of identifying why and where learning can be gained, and either corrective action is required or good behaviours can be replicated in other areas.

“It means continuing to ask why something is so until you get to the true root cause of an insight, which may be many levels away from the original factor,” McArdle said.

Regardless of your strategy, a cost/benefit approach to data analytics is important, he said.

“CFOs can also instil a culture that not all data needs to be captured, and not all the data you do capture needs aeons spent on analysing it.”

Huge benefits

The rewards are great for CFOs who get data transformation right. Vayner said companies with advanced data capabilities can operate based on better insights, and numerous studies show that those companies outperform competitors. For example, they are twice as likely to be in the top quartile of financial performance in their industry if they have advanced data capabilities built in, he said.

Mueller said companies that have invested in their capabilities and are making progress on analytical development have seen significant benefits. But any company that does not invest in using their data assets to drive better business decisions will be at a huge disadvantage in the future and ultimately “out of the game”.

McArdle said that the pre-eminence of data availability, and hunger for it, will keep the disruption train rolling. Without the right strategy, organisations will feel overwhelmed and wallow in information overload.

“Recent trends have shown a seismic shift toward personalisation,” he said. “Marketing is now built on factors such as browsing history, prior purchases, frequency, geolocation, and satisfaction. Biometrics are being used to serve up suggestions of what you may like to buy, and Alexa is recording your preferences.

“More than ever, improved productivity and decision-making will rely on insight derived from the vast amount of data we capture.”

— Tim Cooper is a freelance writer based in the UK. To comment on this article or to suggest an idea for another article, contact Neil Amato, an FM magazine senior editor, at