multi colored circles with earth icons

Perspectives

Building Trust in Machine-Powered FP&A

As published in CFO Journal for the Wall Street Journal

Leveraging a behavioral-backed approach can help alleviate reluctance around algorithmic technology by bridging communication gaps and aligning incentives.

When it comes to the adoption of algorithmic forecasting, finance departments face a conundrum. On the one hand, they’re increasingly looking to these tools to help them meet objectives nimbly and accurately. Yet team members often find it difficult to embrace the technologies and their outputs.

To address these human-centered challenges head on, finance leaders can consider a six-part framework for rolling out machine-enabled forecasting in financial planning and analysis (FP&A). While technical matters involving IT enablement and data management also play an important role, as discussed in a previous article, most of those vital areas focus on such people-oriented challenges as user experience and adoption issues.

Aligning the Operating Model

Equally as important as the immersion of the algorithmic solution in the forecasting process is the interaction between those who use it. While algorithmic forecasting is becoming part of the fabric of finance, there’s no silver bullet for creating an operating model that will foster adoption. Recent trends indicate a shift toward centralization of finance’s analytics capabilities, but the feasibility and effectiveness of centralization largely depends on the structure and scale of your organization.

But no matter how an organization is structured, there are specific skills and interactions that will allow people and machines to work smarter together. Removing silos that have formed, whether by design or organically, can enhance knowledge sharing across domains.

globe with greenery and plants
Bridging Gaps Between Data Scientists and Business Partners

The key knowledge domains at play sit in data science, FP&A, and the functions of the business units involved. The way they interact with both the machine and with each other is a critical indicator of the operating model effectiveness.

Data scientists are typically the ones doing the “heavy lifting” directly in the machine, especially at inception, when the capability is being established and drivers are being added and evaluated. They possess the skills needed to process high volumes of structured and unstructured data to create sophisticated and often complex forecasts. Meanwhile, finance business partners are responsible for driving and explaining the forecast to gain acceptance from their supported businesses and leadership. Mastering this task requires the agility to make and explain changes and can be a cause of communication breakdowns and poor model adoption.

Missing from the equation is often the translation of analytics to insights, and the traceability of insights back to the analytics.

Classically trained finance business partners may struggle with forecast outputs that at first glance either do not appear explainable or have an explanation that does not match a traditional forecast approach (e.g., more advanced algorithms vs. linear relationships). Data scientists in turn tend to focus less on the broader business implications, so they may struggle to explain the cause and effect of forecast inputs and outputs.

Missing from the equation is often the translation of analytics to insights, and the traceability of insights back to the analytics. Without this bi-directional collaboration, machine-powered forecasting loses its agility as outputs become less explainable and, as a result, less trustworthy. Even forecast decomposition, where a projection is broken down into its component parts or drivers, can be tricky to explain clearly. Therefore, the emerging “translator” skillset is becoming more prevalent where algorithmic solutions are deployed.

The role of translators is to serve as the interpreter at the intersection of data science and finance terminology. The role requires the breadth, but not necessarily the depth, of knowledge needed for each of the data science and finance roles. In practice, they can deploy data science to provide quick and meaningful analyses while collaborating with the finance business partner to explain the forecast, layer in business activities, impacts, and insights, and determine the appropriate course of action.

Enabling this translator skillset increases transparency between the complex underlying data science and the results for which the finance business partner is accountable. Such transparency is integral to building trust in the machine and driving adoption.

The translator does not need to be a standalone role in an organization, but the core skills are essential to any cross-functional operating model involving algorithmic forecasting. Positioning data scientists and translators as a centralized service to the business units, or a service delivery model, is one of the most effective ways to generate value. That’s because it allows the two roles to operate as functional and technical counterparts of the forecasting analytics. If data scientists are more localized within the organization, then deploying multiple strategic teams of translators and data scientists to service region-based business partners may fit an organizational model. Alternatively, a single center of excellence (CoE) will unify the data science and translator expertise across the enterprise to raise the impact across the business and functions.

A service delivery approach can enable the translator to transition from the intermediary to a trusted source powered by data science expertise. The skillset alone will not solve the adoption problem, so it is up to finance leadership to provide the vision and strategy for an effective operating and/or service delivery model that opens the pathways between the model’s creators and consumers. Effective interaction and collaboration across the capabilities within the organization is critical to success.

small tree going
Influence Decision-Making

A well-designed solution, process, and operating model alone will not drive decision-making. All are important considerations for adoption but in order to understand the real-time decision triggers at play we need to take a closer look at the underlying behavioral components. This can be considered the finance adaptation of the “last mile” problem, where adoption can be considered the final and most challenging leg of the algorithmic forecasting supply chain. Since this “last mile” is often embedded as a step or stage in a larger finance transformation journey, it can further add to organizational reluctance as the new operating model and the capabilities and tools to support it may all be evolving at the same time.

In order to persuade people to choose to use algorithmic forecasting capabilities, they need to understand the reasons for their use. In this case, making sense of algorithmic forecasting is less about inherent model logic and more about understanding the adoption barriers clouding users’ choices. Once the source of the noise is uncovered, companies can influence behavior through “nudges” that remove barriers and reinforce the machine capabilities. That requires broadening the lens beyond design and process by considering the prevailing culture and incentives.

For example, meet Alan, a business partner whose performance and compensation is driven by his adherence to corporate standards and targets. With the help of his data scientist counterpart, Alan runs his company’s sophisticated new predictive forecast across various scenarios, but none of them enable him to meet the corporate revenue target for his business unit. Instead, he chooses to use his preferred manual method to input his target and back into his forecast. Alan’s forecast misses the mark, but he doesn’t mind because he met his performance objective.

At the same time, however, there are many real-world scenarios where parties accountable in the forecasting process do not have the same objectives. Consider business partners who pressure finance to sandbag their plans. If forecast accuracy is at odds with effective business partnering, then the benefits of predictive capabilities are instantly viewed as diminished, because they did not match a preconceived expectation or business outcome that diverges from creating an accurate forecast of business activity.

The Importance Of Shifting Incentives

This is where incentive shifts matter. In Alan’s case, if his performance is measured by the level of variation between his forecast and actuals, he might turn to his sophisticated algorithmic forecast to inform his baseline. Then, he might push back on corporate targets, because the various scenarios he ran indicate the business unit’s revenue target is not likely attainable. The upshot: Both corporate and Alan’s business unit have better visibility into their financial outlook and Alan has achieved his own objective, powered by the algorithmic forecast. Additional upside: now there’s an opportunity to talk about potential investment to help change the business outcomes.

Inserting steps to monitor and analyze performance can be a powerful tool for continuously improving performance while reinforcing its use. This can be done by comparing the machine outputs to status quo forecasts and tracking for accuracy. Asking, “Where were my assumptions more accurate than the algorithmic forecasts?” can help inform machine improvement opportunities. Asking, “Where was the model more accurate than my assumptions?” can demonstrate where users might partner with the machine to create a more accurate forecast. The constant review of results and learning journey needs to be part of the ongoing conversation.

Additionally, incorporating metrics or measures that reflect some of the gut feel in forecasting today—such as a reference to compound annual growth rate (CAGR) or a metric to measure the variable relationship between revenue and trade—can help provide guideposts and sanity checks to encourage comfort with the algorithmic forecasts. In the end, understanding the limitations of both the model and one’s own judgment will paint a clearer picture for users, enabling the behavioral nudge toward trust and acceptance.

Across the marketplace, company functions are adopting AI and advanced technologies and realizing the intended value and advantage. But finance is still catching up to its counterparts. A user-focused framework can help companies navigate a successful rollout, however, and enable more time spent analyzing the impact of the difficult decisions that finance supports as a business partner. With the accelerated insights and the power of algorithmic forecasting, organizations can stay competitive and responsive to change in a rapidly evolving marketplace.

Read the full report, “Building Trust in a Machine-Powered Forecast

by Eric Merrill, managing director, Finance & Enterprise Performance practice; Paul Thomson, senior manager, Finance & Enterprise Performance practice; Nick Shkreli, senior consultant, Finance & Enterprise Performance practice; Alison Levine, former consultant, Finance & Enterprise Performance practice; and Alan Kryszewski, manager, Emerging ERP practice; all with Deloitte Consulting LLP

Disclaimer and Copyright

This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional advisor. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication. About Deloitte. Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. In the United States, Deloitte refers to one or more of the US member firms of DTTL, their related entities that operate using the “Deloitte” name in the United States and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public accounting. Please see www.deloitte.com/about to learn more about our global network of member firms.

Copyright © 2022 Deloitte Development LLC. All rights reserved.

About Deloitte

Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. In the United States, Deloitte refers to one or more of the US member firms of DTTL, their related entities that operate using the “Deloitte” name in the United States and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public accounting. Please see www.deloitte.com/about to learn more about our global network of member firms.

Copyright © 2022 Deloitte Development LLC. All rights reserved.

Did you find this useful?