Bart Deckers Bart Deckers

Forecasting in the digital age: fostering value through driver-based approach

Accenture strategy forecastin business value based planning
Is your ‘crystal ball’ digital enough?

Every quarter companies are under pressure to deliver the most accurate forecast in order to identify growth opportunities, limit risks, and analyze root causes of profitability issues.
However, in a volatile, complex and competitive economic environment, there is an urgency to re-think the forecasting process in order to increase the agility of organizations to respond to trends that threaten its performance, or that provide specific opportunities.
In these times of digital technologies, companies are becoming better equipped to achieve this ambition by making use of big data, predictive analytics, scenario simulations, self-learning forecasting algorithms, etc. Yet all these methods are of little value unless they are grounded on a sound forecasting methodology that aligns forecasting process with strategy execution and value growth. By identifying the most material, actionable and predictive components of the business, breaking them down into their most fundamental building blocks, and utilizing those to instantly forecast the future – thus applying driver-based forecasting approach – companies are able to generate better and faster insights that drive better informed decisions and actions.

Although driver-based forecasting is already in place for some time, digitalization significantly boosts capabilities by reducing manual interventions and cycle time, integrating analytic findings into forecasting and reporting, and focusing on analysis and corrective action planning. Finance teams are now able to explore the whole range of capabilities that were unknown to many a few years ago, for example:

  • Big data that provides integral access to finance and non-finance data to effectively feed driver based models from S&OP, and operational data to strengthen the predictive capabilities on a shorter cyclical basis;
  • Self-learning forecasting algorithms that test and optimize forecasting methods at each forecasting cycle to adjust to changing trends;
  • Scenario analysis supported by powerful tooling that enables real-time assessment of alternative future scenario’s, allowing for a more valuable interaction between finance and business;
  • Monte Carlo analytical capabilities that help better quantify, assess and manage risk;
  • More powerful analytical capabilities that allow for extensive forecasting models that provide an integrated view over different analytical dimensions.
What is driver-based forecasting?

To exploit the great benefits from digital in forecasting, the basic driver based forecasting process needs to be in place. Today many companies still struggle to establish an effective driver based forecasting capability. Traditional forecasting processes are often characterized by long cycle times, resulting in late availability of forecasts, leaving few time for actions. Ineffective methodologies associated with massive use of spreadsheets and unjustified granularity of planned data, are leading to time- and effort-consuming process, without necessarily higher forecasting accuracy. Last but not the least, focus on accounting line items rather than on business drivers results in reduced added value of the forecasting process and less accountability.

The heart of driver-based forecasting technique relies on the notion of driver also called value driver or business driver. Drivers are measurable factors that influence strategy execution and shareholder value creation. They represent key “levers” to achieve strategy and can be financial or non-financial. Understanding and effective management of value drivers is especially important in the context of growth agenda that today’s CFOs need to navigate. For example, depending on company’s ambition to grow organically or inorganically, different drivers will be in the focus of management attention and steering.

Examples of value drivers are volume, price, headcounts, and macro-economic factors. However, those are not to be confused with key metrics (KPIs) which is a way to measure the success or failure to reach a strategic, tactical or operational objective after comparing actual results with forecasted data.

Driver-based forecasting methodology is grounded on the principle that a business can be accurately predicted by focusing on a selective set of key operational and financial data points. The ability to forecast by updating a limited number of value drivers results in increased process efficiency, allowing the organization to focus on meaningful root cause analysis and remediation action planning.

Driver-based approach is of use not only for the forecasting process, but also across the whole enterprise performance management cycle, including strategic planning and annual budgeting processes. It facilitates scenario simulation, fosters an integration of planning processes, allows to create faster outlook of future performance and adjust the action plans with agility and speed.

How does it work in practice? 

Implementation of driver-based forecasting starts with identification of value drivers. A combination of two approaches can be used to identify the companies value drivers: the first one consists in mapping drivers to strategy objectives whereas the second one consists in process decomposition in order to identify the drivers. Identification of value drivers is critical as the number of drivers should be limited even though different businesses may require different drivers. Moreover, each driver needs to be material, actionable, predictive of results and traceable in actuals – if there is no way to track the driver compared to the forecast, root-cause analysis becomes a guessing game.

Next, the cause and effect linkages need to be established between financial and non-financial drivers. Gradual decomposition of company performance measures in the form of value driver tree allows to analyze complex dependencies and identify the key sources of added value. Nonetheless correlation should not be mistaken for causality.

Drivers are then prioritized based on sensitivity (Which level of impact on economic performance will improvement to this driver result in?) and manageability (How likely is it that appropriate focus on this driver will help achieve our desired results?). This helps to maintain focus on few critical drivers.

Finally, the financial impact on profit and losses is assessed by establishing correlations between drivers and accounting line items. Development of an accurate driver-based model is a crucial step. The consideration should be given to the fact that not all data are driver-related: while some measures are driver-based (directly linked to the driver performance), some others are only driver influenced (indirect impact), the others are not driver-related (non-volatile and one-off items).

As the business evolves, drivers and their relationships need to be re-evaluated regularly to ensure an optimal driver-based model.

What’s the bigger picture?

To conclude, driver-based approach is a foundational concept that enables and blends together multiple elements of effective and efficient forecasting process:

  • It enables the organization to employ integrated business planning, ensuring consistency and alignment across the enterprise;
  • It is an important pre-requisite to exploit the potential of digital capabilities in forecasting process, e.g. sensitivity analysis and scenario-based planning;
  • With the use of the latest technologies that expand the limits of data analysis, it equips FP&A teams with the predictive capabilities at a new level;
  • – It represents the crucial component of linking the company’s performance to its strategy via the notion of value drivers and actively steering the strategy execution rather than looking for causes of not achieving it.

From workflow that updates with new information to advanced algorithms that analyze unstructured data and cognitive agents who identify opportunities to close gaps to target, digital forecasting will enable CFO to drive digital transformation agenda and to be truly positioned as a new digital apostle.

Feel free to contact me to schedule a private meeting to discuss how to make the New forecasting work for your business.


Bart Deckers

Managing Director, Accenture Strategy

Personal Page