CFOs’ work on data analytics reaps exciting results

When David Horan, CPA, CGMA, started motor loan company NextGear Capital with colleagues in 2005, they relied heavily on experience and gut feel to run the business.

But that approach has been transformed at the company based in Indiana in the US. Today, Horan, NextGear Capital's CFO, uses data analytics and visualisation through real-time dashboards to guide strategic decisions. This means it can make accurate predictions — for example, around loan underwriting — that were previously based on pure instinct.

The outcome has been a dramatic improvement in strategic thinking throughout the company, he said. It's an example of how data analytics are transforming the way finance leaders think about their role. But for Horan and others, it has also involved a deep journey of discovery.

Many CFOs are realising the potential of data analytics, and, though challenges remain, exciting results are emerging. Experts such as Chris Argent, managing director of Generation CFO, have declared that data analysis has become the most effective way for CFOs to add strategic value, and many are starting to agree.

Argent said, to achieve this, he sees CFOs upskilling teams with digital awareness and data analytics training. "We are seeing increased understanding of this new capability, and even a demand for program language training, which shows finance teams want to get the data faster and use it more impactfully," he said.

London-based Simon Bittlestone, FCMA, CGMA, is CEO at analytics platform Metapraxis and a CIMA Global Council member. He said finance functions' understanding of data analytics tools and techniques has increased significantly over the last five years, particularly the last 18 months during the pandemic.

Bittlestone said the key to analytics is combining financial information with nonfinancials, such as market, sales, operational, and people data. The latter are leading indicators of financial outcomes. So by connecting the two in a dynamic model, finance can help people across the business understand what drives performance and give them insights to make better decisions.

Once they have the data strategy, technology, and skills in place, finance leaders are finding they can use analytics to improve almost any aspect of the business from sales forecasts to risk management and supply chain efficiency.

Here are examples of how CFOs from around the globe have been doing this.

Big bottom-line savings

London-based Ilan Cohen describes himself as a digital CFO and has worked on data-led projects at several companies. He said these have brought significant bottom-line benefits in recent roles, including in credit risk modelling and underwriting for a loan company.

"Open banking [which requires banks to give third parties access to customer data] has allowed us to analyse and get a fairer, more accurate picture of a loan applicant's creditworthiness," he said. "A huge drain on companies who lend to small firms is the avalanche of applications with a limited team of underwriters forced to assess every application equally."

Cohen implemented a machine-learning data product that scores incoming applications on various credit-related factors. Knowing these upfront allowed the company to predict, to 88% accuracy, whether each loan would be funded. Ranking the hundreds of applications this way showed staff where to concentrate their efforts. Loans funded rose 300%.

A project at another company involved analysing customer churn. Many companies focus marketing budgets on new customer acquisition, Cohen said. But retaining existing customers is more efficient.

"We used machine-learning frameworks to profile customer behaviour, predict when they will drop off, and intervene," he said. "We segmented them according to churn probability and estimated customer lifetime value. Focusing marketing efforts and budget on those with high expected value and churn probability improved retention by 440% and revenue by 35%. Best of all, it was not a backbreaking amount of work. It just required the right focus and toolkit."

Delivering real-time benefits

Singapore-based Edwin Ang, senior financial controller APAC at food delivery company Foodpanda, said the company has implemented a dashboard to enable real-time, data-driven decisions.

"We use data to measure effectiveness of user incentives, manage rider supply and demand, reach new customers, identify fraud, monitor competition, predict orders, adjust pricing, and measure profitability of each brand," he said.

"We monitor closely each data point that affects our profit and loss. By deep-diving into live data we can analyse trends in real time rather than wait for the following month close. We can then advise decision-makers on any opportunities or risks impacting our profitability or growth.

"Waiting for the following month to make business decisions would be too late. For example, the challenges of the COVID pandemic required heavy reliance on data as countries went into lockdown overnight. We could see the impact of that and react swiftly."

Improving vision and cohesion

When NextGear became part of Cox Automotive in 2012, it joined one of the largest automotive companies in the US and brought access to ever-growing reams of data.

"There is too much raw data to absorb, so the value comes from filtering out the noise and structuring the data into an accessible, visually appealing representation that tells a story," Horan said. "This helps bring staff at all levels on the journey."

NextGear has found two main uses so far. First, it has developed an internal risk-scoring algorithm that helps determine pricing levels and predict business risks more accurately.

"Implementation of this tool over the last few years has completely changed the way we do business," Horan said.

Second, data helps drive use of NextGear's platform by improving understanding of opportunities in the wholesale automotive marketplace. This uses multiple data points from across the group to identify areas of opportunity with new and existing customers.

"Broad adoption of our data visualisation dashboards has created a culture where employees at every level can look at the same information and understand it the same way," Horan said. "This makes conversations between field and corporate teams, for example, more fluid and unbiased. Driving KPIs from these same dashboards improves everyone's focus and understanding of the business."

More accurate forecasts

New York-based Ben Goodband, ACMA, CGMA, has held senior finance positions at several top companies and is currently working as interim CFO for data and analytics at London Stock Exchange Group.

He said machine learning-led predictive analytics combined with cloud-based software has enabled significant efficiencies over the last few years.

"At a recent employer, we had a traditional sales forecast built on a manual process," Goodband said. "Each step involved another human who added their own subjectivity to the information, so it was not data-driven.

"This often resulted in poor accuracy and frustrated senior management. We started an algorithmic forecasting model that used 24 months of actual sales data to capture the behavioural differences in sales teams, products, and customer types. It then applied those differences to the current sales pipeline data to predict results, improving forecast accuracy by 90% and reducing time spent on forecasting by many hours across finance and sales teams."

Overcoming challenges

Goodband said his sales prediction project team faced many adoption challenges because people did not want to give up old processes, even though there was a huge productivity benefit to them. To work effectively with the new system, the team had to take a long-term, iterative approach, improving the algorithm constantly.

CFOs should not underestimate how important collaboration is to success. Bittlestone said: "Blending data analytics with a solid understanding of finance and the business itself probably remains the single biggest challenge and opportunity. Get this right and the CFO has an incredibly powerful capability that can bring competitive advantage. But it is not straightforward and rarely if ever sits all in the same head, so a big part of the solution is teamwork and communication."

Cohen said the challenge is sometimes that the finance function has a siloed mentality.

"They might hire data scientists and machine-learning engineers but not interact with them," he said. "They put them in the basement with some databases and hope the magic happens. It's a twisted approach because, in data science, the hypothesis is the hardest thing to formulate. Data scientists and experts are not close to the product or the customer, so they don't know what questions to ask. The CFO should be framing and contextualising the questions.

"Another problem is when finance leaders try to boil the ocean and solve hugely complex problems at once. You don't need a huge budget. In previous roles, I helped deliver a ton of value with just a handful of key hires. Start by solving problems you understand with open-source technologies and cheap, established frameworks. Your business probably has problems other businesses have already solved and released the solutions into the cloud, so use them."

Cohen agreed the biggest challenge is getting others to buy in and change their behaviour. "It's easy to change code or a website but harder to change people's way of working. No matter how sophisticated the system or intelligent the people, management and restructuring in the trenches is hard."

Horan said to address this, finance leaders must be persistent in transforming to a data-driven culture. "Some people will always trust their instincts more than data," he said. "While instincts are important, they need to be grounded. You have to work through the anecdotes and feelings that surround the business. But the more you validate the story told through the data, the more everyone will appreciate its power."

— 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 Neil.Amato@aicpa-cima.com.