Finance discipline for optimized performance

July 22, 2021

A blog post by David Cutbill, principal, Deloitte & Touche LLP

In part I, we offered insights into the need for these new finance disciplines to optimize performance. Part II highlighted the methodologies for developing resilient scenarios—the first step to optimizing finance performance. In part III, the next step to creating more finance discipline, we discussed creating capital agility now and more capital resilience for the future, with a better understanding of capital and cash needs across the business as well as how to continually optimize the allocation of this capital. The final step in the journey to these new integrated finance disciplines introduces analytics capabilities for each scenario to inform new data and scenario-driven statistical forecast models. Agile planning, short-term capital allocation, and predictive forecast models collectively foster a more resilient business that can improve business and finance performance now while setting up future success in an ever-changing environment.

What is enhanced predictive forecasting, and how can it improve decision-making?

Predictive forecasting is the act of forecasting and assessing a number of potential scenarios. However, this process needs to be more rapid and flexible to achieve capital optimization in these uncertain environments. Organizations can start by hotwiring traditional planning and forecasting processes and leveraging more predictive analytics and data technology for data consolidation, planning, budgeting, and scenario assessment.

Creating agile, scalable, and expedient predictive forecasting capabilities with new technology
Predictive analytics and data technology can be optimized around five capabilities in a new forecast model that leverages driver-based analysis to provide greater transparency into decision-making and its impact.

Robotic process automation: The automation of large quantities of data from markets and regions using robotic process automation rapidly consolidates data and builds the forecast baseline.

Predictive analytics: Predictive forecasting using statistical models that incorporate trends with less manual effort increases accuracy while narrowing focus and reducing human bias.

Visualization dashboards: Plans, budgets, and decisions are shared via self-service dashboards and collaboration tools that enable faster organizational alignment and more time for more impactful conversations.

Outcome sensing: Central and regional teams track real-time data to gain performance insights, opportunity, risk sensing, and sentiment analysis, which arm finance with “impact” insights and more forward-looking analysis.

Agile forecasts and decision updates: Agility is enabled through capabilities, but is predicated on finance, bringing a point of view that mitigates risk, captures more opportunities, and is supported by updated processes in real time.

Using new forecast model capabilities to predict the impact of different scenarios

In our previous blog, we shared insights into developing scenarios for enhanced finance disciplines using an example of four possible scenarios that may result from the pandemic and the impact of each possibility. We apply each scenario identified to the new forecast model to develop predictions and potential implications for the business.

The new forecast model

• Identify key potential drivers and significant data sources for each possible scenario, and evaluate these drivers for quality and relevance.
• Within a scenario, understand the impact on some of the key external drivers. To use our previous example of the pandemic, there are more indirect impacts, such as infection rates or travel patterns, that directly affect product demand or raw material availability and costs.
• Develop a model to link these external drivers to their impact on the business and financial performance.
• Use the forecast model to identify actionable and future-proof strategies for each alternate scenario.
• Continuously refresh driver inputs and data sources using AI-based tools to monitor these external drivers, allowing for ad hoc analysis and strategy adjustments.

Applying predictive forecasting to different types of decisions

Predictive forecasting capabilities using this model can work for multiple purposes and organization levels for different decisions. Some examples are discussed below:

Top-of-the-house model
The corporate-level model uses strategic scenario models, target setting, and validation of business-unit forecasts. Organizations looking to turn around forecasts frequently, create guardrails to compare bottom-up numbers, and generate high-level forecasts and simulations may want to consider this top-level forecast model.

Baseline model
The business-unit or product-level predictive model creates a baseline plan and forecast at the granular business unit level that cascades to the lowest level. This forecast model works for organizations looking to statistically generate first-pass baselines. A user focuses on adjustments and exceptions to the statistical forecast and produces bottom-up planning and forecast levels of detail.

Specialized model
Specialized predictive models may be used for targeted and strategic forecasting, demand planning for sales forecasting, and risk forecasting for R&D expense. Organizations looking to focus on specific, highly complex P&L line items (e.g., new product and SKU revenue forecasting) are prime candidates for more specialized forecast models.

The value of predictive forecasting around scenarios for decision-making

A more agile approach to planning with this forecast model is particularly valuable in the current market, rife with uncertainty, allowing for clarity against future-focused implications and potential opportunities, better planning and confidence in decision-making, and improved operational resilience that can increase the chances of success across diverse businesses in a time of transformation.

• Better risk mitigation with scenario analysis capabilities can identify the degree of influence of inputs and outputs and lessen the impact of management biases on decisions and execution.
• Industry-leading accuracy, improvement over the current forecast practice, and on-demand forecast generation that reduces cycle time (versus latency experienced today) can allow for more time spent on strategic solutions.
• Incorporating exponential data sources and information, from macroeconomic indicators to industry-specific drivers, delivers more valuable insights and visibility into business drivers and relationships to P&L, balance sheet, and cash flow.

In our series on finance discipline for optimized performance, we offered insights into the need for these new finance disciplines. From there, we went into methodologies for developing resilient scenarios and creating capital agility now and more capital resilience for the future. Now that we have detailed the methodology for scenario planning and predictive forecasting, in the final part of our series, we will wrap up with further insights and actions you can take now.