Stewardship of fiscal performance has always been a numbers game; now, big data and automation are allowing financial leaders to take their key performance indicators to a new level.
While the data deluge has created ample opportunity to improve financial KPIs, managing that data and transforming it into actionable insights is proving to be a challenge — an issue explored in a panel discussion at the 19th annual MIT Sloan CFO Summit.
The challenges are especially acute in companies with a data estate that is spread across different systems and is punctuated by silos, data gaps, and inconsistencies in the type and quality of data stored, chief financial officers agreed.
To fully capitalize on data-driven performance, financial organizations need to start back at step one: Getting enterprise data in order to ensure the right data is captured and that data initiatives are aligned with core fiscal strategy and business goals. Key to that effort is a renewed focus on data governance and working through the question of who owns the data model.
“Before we can even think about getting insights and value out of the data, we [need] a good way of getting that data into our systems and a good way of governing it so that it can be usable for us,” said panel moderator Peter Irwin, a partner at KPMG Lighthouse, which specializes in data analytics, automation, and artificial intelligence.
Here are key takeaways from the discussion:
Finance needs an ownership stake in the data model
The data explosion is a doubled-edged sword. There’s so much potential in leveraging data and advanced analytics to boost fiscal performance, yet without a single source of truth, organizations can be stuck in a cycle of never-ending reconciliations and questionable data integrity that diminishes data’s value to the business.
Historically, the information technology department has had sole responsibility for the data model, but without complete understanding of what’s required for fiscal KPIs, that can lead to a lot of redundant and unproductive work. Aligning goals and responsibilities is central to data governance; as part of that process, finance should have some ownership stake in the data model, along with IT.
“Finance has a deep understanding of the calculations, the sources, and the definitions and mapping” of financial data, said Kae Arima, vice president of finance at Workday, which has an ownership stake in the data model at the provider of cloud-based finance and human resources software.
“We’re able to reduce the time spent on our reconciliation … and there’s a lot more data quality and integrity.” The ability to effectively leverage financial data has enabled Workday to scale as an organization during a period of explosive growth with relatively low general and administrative costs, she added.
The right data — internal and external — makes all the difference
In order to generate useful insights and execute on data-driven metrics, organizations have to identify the right data from both internal and external sources. At ChaosSearch, a venture-backed, cloud-native data lake platform, the emphasis is on top-of-the-funnel KPIs to evaluate the velocity of go-to-market plans, marketing programs, and qualified sales leads, explained Melinda Smith, the startup’s CFO.
“From a KPI perspective, it’s all about driving as much monthly recurring revenue as we can,” she explained. “Therefore, it’s important to identify the point in the sales funnel where you flip from still selling a deal to where you’re confident it’s going to close.”
As a startup building IT infrastructure from scratch, ChaosSearch has no legacy platform or data silos to consider — all of the company’s relevant data resides in a variety of cloud systems, making integration between them key, Smith said.
Workday is also combining data from different functions to deliver new insights and KPIs. For example, by marrying data from HR and finance, the company has created a new metric called “performance to plan,” which is used to help project managers optimize execution plans for different initiatives.
The company is also tapping into 15 external data sources on the finance side — databases like Bloomberg and Dun & Bradstreet — to help automate and scale processes, Arima said.
For example, automated workflows tap into transactional data from E*Trade, the firm’s stock administration platform, to generate accounting entries and to create a virtual subledger with high assurances of data integrity. “We have peace of mind that we know the source of the data, and it’s saved a ton of hours for us,” Smith explained.
TraceLink, a software-as-a-service provider that helps pharmaceutical companies track and trace products throughout their supply chains, combines its internal data with data from external partners to offer a new service to its customer base, according to Michael Mozzer, TraceLink CFO.
Data generated from a network of 200,000 entities tracking pharmaceutical product serial numbers has value to customers for a variety of use cases, from automating the recall process to optimizing production runs, he said.
“We can help these pharma companies understand whether it’s time to spin up another production run or if they have enough product already sitting there in the supply chain,” Mozzer explained. “We have anonymous rights to all this data so that we can monetize it and sell it back to our customer base.”
Automation and AI help optimize financial processes
While most companies are still in the experimental stage, automation is starting to come into its own in finance processes such as leveraging OCR technology to scan invoices and automating other accounts payable processes. Workday is marrying data from a weekly employee engagement survey with attrition data, then employing AI and predictive analytics to facilitate adaptive planning financial forecasts.
For its part, ChaosSearch has leveraged various cloud applications to automate approval flows around invoicing, reducing the number of on-staff employees, which is critical to a startup with a limited budget. “We used to have a dedicated payroll person, but I now have one person that does both accounts payable and payroll, for example,” Smith said. “That’s because of the efficiencies we’re gaining from these tools.”
Progress is incremental, and that’s okay
The process is a journey, not a sprint, and financial organizations shouldn’t let perfection get in the way of good as they moved toward achieving some automation of KPIs, the panelists agreed.
Understanding how to define a requirement for a minimum viable metric can still deliver some benefits of automation. That still requires that data capture processes be well-tuned, because ensuring data is of the highest quality leads to accurate KPIs.
“If it’s garbage in, it’s garbage out. So we really do spend a lot of time making sure we have the right validations in place at the front of the process to make sure everything downstream is good quality data,” said Workday’s Arima.
Prioritization of projects is key. “Finance people have such a big appetite to measure everything,” Arima said. “We’re trying very hard to drive automation and forecast accuracy that goes to sales data and people data. It’s about figuring out what you need to measure first.”