COVID-19 has brought about one of the strangest business environments in memory, and more than ever leaders need data to determine how to survive it. It has pushed finance professionals, particularly those engaged in finance analytics, into the limelight. They are being asked to build forecasts and models for situations they have never encountered (and never anticipated). So, is finance ready to play a starring role during a crisis once again?
APQC conducted research into the current state and success drivers of finance analytics just before the true impact of the pandemic was evident. The study showed that for some companies finance analytics may play a prominent role in outlasting the accompanying recession, but for others, it will be a conspicuous Achilles heel.
Finance analytics is the process of searching for and gathering meaningful insights from financial data, often in combination with other business data, to inform decision-making.
We know that strong finance analytics can positively impact a range of business outcomes: risk mitigation, customer satisfaction, and bottom-line results among them. Moreover, analytics is the lever by which finance can distinguish itself as a valued business partner to the organization. When finance has data-based insights at the ready, senior leaders come knocking.
Overall, the APQC research showed that finance analytics is maturing at an accelerated pace. At this point, those organizations new to the game are in the minority: Half of the 200 finance executives surveyed have been conducting analytics in finance for more than 10 years, and 68% have been doing analytics since at least 2014. A majority of the organizations have also increased investment in finance analytics over the past three years.
Still, questions remain. Have finance analytics programs matured quickly enough to meet this moment? Are organizations’ data accurate and easily accessible to those who need it? Do organizations have talent capable of performing top-notch analyses? Have finance organizations moved beyond spreadsheets and embraced leading-edge tools? APQC dug into these questions as we examined the current state and success drivers of finance analytics.
Current State
Structure. Most finance analytics is delivered in either a centralized or hybrid model. The hybrid model leverages a centralized governance team or center of excellence (COE) combined with decentralized teams embedded in the business units. Only 18% of organizations have a fully decentralized structure (and a lonely few, just 4%, have no formal structure at all). That reflects the growing maturity of finance analytics.
Centralized governance provides strategic alignment, accountability, and consistent communication and implementation planning across the enterprise. One example of the benefits of centralization comes from Johnson & Johnson’s FP&A COE. J&J’s COE has successfully standardized finance processes, tools, and data warehousing. This setup allowed it to move faster in the integration of leading-edge technologies and approaches, such as predictive P&L analytics and cognitive services (a set of machine learning algorithms developed to solve problems in the field of artificial intelligence).
Most organizations (76%) engage in a low level of outsourcing for finance analytics, which APQC defines as outsourcing 20% of tasks or less. Today, CFOs are probably quite glad they kept outsourcing of analytics to a minimum — that’s one coronavirus-related disruption they avoided.
But there are also more long-term reasons why finance analytics teams don’t typically use much outsourcing. For one, the technology tools like robotic process automation (RPA) can handle the transactional tasks outsourcers used to. Two, keeping finance analytics skills and knowledge in-house can promote a culture of analytics and data-based decision making across the organization. J&J’s FP&A COE for example serves as a “talent incubator” that provides training and development opportunities for finance professionals across the company.
Practices. As Scott Wallace of eCapital Advisors told APQC, “In an ideal world, finance analytics should cut across all functions of the business because finance sits at the center of the data and the business.” The good news is, we’re getting there. APQC found that a vast majority (97%) of finance analytics programs incorporate non-financial data into their analyses, most commonly data from operations, sales, and supply chains. About one-third also leverage external data on industry, competitors, market trends, and benchmarks.
However, finance analytics programs have some room to grow when it comes to how they analyze all of this data. Only about one-third of survey respondents said they use the most advanced form of analytics, prescriptive analytics, for most major finance processes. (See “Progressing in Analytics,” below.) The most prominent application of advanced analytics occurs in the area of internal controls. A little more than 40% of respondents use prescriptive analysis to identify outliers and prevent fraud. Forecasting is another area suitable for advanced forms of analytics. APQC found that 73% of finance executives said their analytics programs applied predictive analytics.
Prescriptive analytics offers big benefits that many organizations are missing out on. For example, at a global tech company (which participated in a blinded case study with APQC and the Association for Financial Professionals), machine learning plays a central role in forecasting. FP&A professionals build algorithms directly into a machine learning engine that can be trained for a variety of purposes. It can do regression analysis for forecasting, budgeting, and workforce needs. This same tool can also use classification to look for patterns. For example, it helps the organization find new customers for a product based on their existing profiles. It can also perform risk management by mastering the characteristics of fraudulent transactions.
Tools. When it comes to technology, Excel is still very prevalent — 97% of those surveyed said it’s one of their primary tools for finance analytics. Most (73%) also leverage the finance/accounting modules of their enterprise resource planning systems. A substantial number (48%) have developed their own in-house tools for finance analytics, while 39% use off-the-shelf analytics software. Beyond this, the picture starts to get more complex. In addition to statistical packages like SAS and SPSS — used by 21% of those surveyed — there are a wide variety of visualization tools, business intelligence tools, Excel add-ons, integration tools, and programming languages being deployed.
The good news is that for those looking to go beyond Excel, there are many new tools and technologies that are more accessible and affordable than ever. Vendors are increasingly integrating sophisticated analytics capabilities into their most popular products. The bad news? Some finance analytics teams may be hard-pressed to choose the right tool for the job. In addition, organizations need to get both their people and their data ready to make the most of these leading-edge tools.
From Good to Great
The research showed that finance analytics programs are growing stronger, but many are struggling to make the leap from “good” to “great.” Overall, 70% of the finance executives surveyed rated their finance analytics approach “effective” or “average,” but only 24% rated it “very effective.” APQC identified key practices in three areas — around data, talent, and technology — that survey participants indicated drive the success of very effective finance analytics programs. If your organization does not already have these practices in place, consider adding them to put your organization on the path to great.
Data. Ensure your data is clean and clear. Dirty data is worse than useless; it’s dangerous. IBM estimated that poor-quality data costs the U.S. economy $3.1 trillion yearly. Bad data also slows employee productivity and encourages executives to rely on intuition and instinct.
APQC’s research showed that very effective finance teams are highly focused on data standardization, quality, and accessibility.
Standardization is one of the keys to successful analytics — without it, an organization will never really know if it’s comparing apples to apples. Consider this anecdote shared by Armeta Analytics’ managing director Jim Rushton: The owner of several gas stations wanted to see sales-per-pump across different locations. But what, exactly, constituted a pump? Was each of the fuel nozzles a pump? Was each fueling station a pump? Was each side of a fueling station a pump? If an organization does not clarify terms, it cannot properly gather and compare data.
Cleaning up an organization’s data also means tidying up and maintaining the integrity of key tools, such as the chart of accounts (COA). Messy COAs cause all kinds of problems, from increasing process costs to causing difficulties in reconciliations and reporting. They can be a major roadblock in the adoption of new systems and add-on technology tools.
Most organizations are great at adding new accounts to the COA, but they need to be more proactive in removing inactive accounts on a periodic (at least annual) basis. Accounts with little or no balance are often a sign that reduction and simplification are in order.
Talent. Invest in your people. APQC’s research showed that investing in technical analytics expertise pays off. We identified a statistically significant relationship between statistics skills and statistical software package knowledge and overall effectiveness.
However, recruiting talent is not just finding people with the most impressive resumes. Growing your people is also important. Analytics specialists need to be provided with opportunities to continuously refine and expand their knowledge. They want to learn not only new statistical methods, but also statistical software packages like SPSS and R, visualization tools like Tableau and Power BI, and programming languages like Python and Visual Basic. Many also need to be presented with stretch assignments that further their understanding of the business.
“There’s often a gap between having a data science degree and knowing how to apply it in the real world,” said Beth Lahaie, program director of Divergence Academy, a data science-focused career college.
Leading finance analytics programs partner with human resources to develop competency models, formal in-house trainings, and structured stretch assignments. Because they want their teams to receive the best education, they’re also not shy to invest in external training.
APQC identified a statistically significant relationship between reimbursement for external or university coursework and the overall effectiveness of efforts. Many institutions of higher education, including community and career colleges, offer analytics programs and courses that can accommodate the schedules and learning needs of working professionals.
Technology. Explore new tools. Excel is great — and spreadsheet wizards know how to push it to its limits — but there’s a lot of emerging tech out there that some analytics teams are using to pull ahead of the pack. As Jay Giannantonio, ERM advisory principal and senior project manager of Column5 Consulting, said, “Let’s face it: Excel is the tool of finance and accounting. No tool is more widely used … but you need tools beyond Excel to go into your ERP systems and combine that data with what’s in your organization’s customer relationship management and manufacturing requirements planning [systems] and tie it all together.”
In particular, APQC found that RPA and interactive self-service reporting tools are associated with superior effectiveness. RPA is a huge factor across finance. A majority of finance functions already use it to streamline transactional processes such as auditing expense reports and processing vendor payments. But RPA doesn’t just save time — it also enables analysis of massive amounts of business data. It can help finance analytics teams integrate large datasets from other functions (e.g., operations, sales, supply chain) and from external sources into financial planning, budgeting, and forecasting.
Interactive self-service and querying capabilities allow users across the organization to access data when and where they need it for decision-making. Leaders can explore financial results, visualize data in formats they’re comfortable consuming, apply restraints relevant to their goals, and dig into the drivers of results and variances. Self-service tools also ensure everyone’s looking at the same single source of truth. Thus, user experience and governance must be taken into consideration when adopting these tools.
While finance may have been late to the game in leveraging and taking ownership of analytics, more and more organizations are making progress. They are adding the talent and tools necessary to generate analytical insights that drive decision-making in and well beyond the function.
As principal research lead, Rachele Collins, Ph.D., is responsible for APQC’s best practices research in financial management. As an analyst on APQC’s research team, Mercy Harper, Ph.D., is responsible for writing best practices and benchmarking reports.