Putting rolling reforecasting into practice

In the first piece in this series, I suggested that despite it being widely recognized that rolling-reforecasts help improve both the accuracy of the forecasts and corporate agility, only a third of companies actually use them and discussed the reasons why that might be. In the second installment, I suggested that adopting a driver-based approach to planning is an essential element in rolling reforecasting, and listed some of the important capabilities needed to deliver this in an enterprise planning and budgeting solution. In this final piece, I’ll assume you have got the right software and executive sponsorship to make it happen, and will share some of the tips on implementing rolling reforecasting that I have picked up from observing companies put it into practice.

Coordinate rolling reforecasts with other planning cycles

Clearly month-end reporting needs to be complete before a round of reforecasting can begin. However, it also pays to synchronize the enterprise-level reforecasting cycle with other repetitive sales and operational planning cycles to ensure that the most recent data is available.

Help contributors focus on the immediate future

It is also important that contributors focus most of their time and effort on analyzing recent variances, understanding why they occurred (something that is easy in a driver-based model) and forecasting what will happen in the near future. This means that the majority of their time will be spent on updating periods they have previously forecast, rather than on the new periods added at the end forecast, as these are far more uncertain and much more susceptible to forecasting error.

Using different shading for the cells in the last six columns of an 18-month rolling reforecast helps, although you might also consider auto-populating them using predictive algorithms that project the trends in the forecast for the previous 12 months.

Design for speed

Free from having to rework their own planning spreadsheets, contributors will soon find it takes very little time to update a rolling forecasting model that leverages business drivers. However, you need to design the process for speed. If you are implementing quarterly rolling reforecasts, that means aiming for a cycle time of a week. But if you are implementing monthly rolling reforecasts, then you need to reduce that to just two or three days. Clearly there is little point in contributors participating until the driver data they work with has been undated by those upstream from them. So a first step to achieving such short turnarounds is to identify the exact order in which individual contributors need to access the model, and creating a workflow and alerts to expedite that process.

However, once they receive an alert to invite them to review and update their own line items, the clock should start ticking to limit the time they have to complete their input. If the goal is to complete the entire reforecasting cycle within 48 hours, then individuals can really only be allowed 3 or 4 hours at most. Today it is easy for contributors to review and update data on their mobile device. But if they still fail to submit a revised forecast within the allotted time window, there are two choices. You can either alert their manager to become involved or you can recalculate their part of the model. Both of these options can be built into the workflow, so that the process is expedited without the need for constant monitoring.

Design for light-touch

When a contributor has only a few hours available to review and update a forecast, it helps if their attention is directed to the handful of things that really matter. For the most part, these will be the business drivers and input costs that impact the main variable line-item expenses and other large elements of expenditure, such as the marketing budget. Regardless of where they appear in the standard chart of accounts, it helps focus the attention if these key elements, together with recent actuals, prior forecasts, and the results of any rule-based calculations, are displayed in a separate pane. This will help contributors be more productive as they can do most of their input and “what-if” modeling in a single view.

Providing contributors with an easy-to-use selection of predictive tools will also improve their productivity, as will automating the spreading of higher order values in a dimension to its children using an algorithm based on trailing averages.

Make it a consistent and repetitive process

Any business process that occurs infrequently usually takes longer to complete and is fraught with errors. So once you have identified and configured a workflow to drive your monthly reforecasting process, automate it so it happens like clockwork. That way, contributors will soon accommodate it into their own cycle of monthly tasks and become increasing quicker at reviewing and updating their forecasts.

Implementing rolling reforecasting will help an organization improve its agility and keep its operational capacity aligned with fluctuating demand. It will also give senior executives better insight into their future financial performance. This will help them provide the market with more reliable guidance and reduce the risk of missing quarterly earnings targets. As market analysts put more value on companies that deliver predictable results, the ultimate outcome should be a mark-up in the stock price resulting in a tangible benefit for investors. Yes, with rolling reforecasting, everybody wins.