Skip to main content

A colleague recently shared this LinkedIn article by Jack and and Suzy Welch with me. I receive plenty of article recommendations daily, but the fact that this one included Jack Welch’s name and the catchy title “This is the Most Mind-Numbing Ritual in Business (But Doesn’t Have to Be)” convinced me to take the time to click on it. The article argues that companies should ground their annual budgeting exercise in reality and link it to true operational what-if planning that can get corporate and field team energized behind common goals. For those who have read, Jack: Straight from the Gut, this is not new territory. In the book, Jack provides a lot of detail on how he conducted GE’s annual budgeting exercise. On that note, it’s a great read for anyone wanting to transform their budgeting process.

However, the article set me thinking on why anyone would restrict this grounded-in-reality, market-based approach to an annual exercise. With the rate of change most businesses deal with today, it is certainly time well spent to have this same discussion. Can we at least agree to omit sections of the budgeting process when there are no significant changes in the business environment?

The key to conducting operational planning more frequently is adopting processes that allow stakeholders to express their viewpoint on important business metrics with accompanying analysis, assumptions, and plans. The teams then get together to review these inputs to arrive at a consensus plan.

On the demand side, key inputs I proactively engage with my clients on include:

  • Statistical Forecasts – Leverages 1 or more statistical models that have been validated upfront, typically managed by Supply Chain teams.
  • Customer Forecasts – Gathered from a Collaborative Planning, Forecasting and Replenishment (CPFR) process, typically managed by Supply Chain teams.
  • Account Management Forecasts – Gathered in meeting with Key Account teams (typically managed by Sales Operations teams.
  • Market-based Forecasts – Generated using syndicated data that represents the size of the overall market, typically managed by Category teams.

Given each of these are done at different levels (e.g., Market-based inputs at Subcategory by Month level whereas an account team provided their forecast at Key Account by Month x Key SKUs level), there needs to be a standard way to compare and contrast these forecasts at a uniform level of granularity to then drive a consensus. Teams that are accustomed to looking at absolute numbers need to get comfortable thinking about ranges. For example, if each of these forecasts are in a range of +/- 5% from each other, you’ve got a winner; choose a number and move onto the discussion of assumptions, upside scenarios, and execution plans.

If the ranges are further apart, as they are apt to be given the wide range of inputs, a robust discussion of assumptions and Forecast Value Add (read: “Who recently did the best job of forecasting?”) become critical. Based on these inputs, a consensus number can be reached with an accompanying storyline on assumptions that need to pan out to make that outcome possible. However, remember that the consensus number is typically at an aggregate level and that operationalizing the forecast for execution purposes will require it to be disaggregated down to the lower levels (e.g., Category->SKU, Month->Week, Account->Store).

If this sounds like a monumental challenge—and I have had clients tell me it does—that might explain why Jack recommends an annual instead of a monthly consensus process. But, shouldn’t we be able to leverage modern technology to meet this challenge? And if so, what would be the core set of capabilities it must fulfill. I have posited an initial below for your consideration:

  • Allow each department to create a forecast based on their view of the world. Moreover, restrict sharing of that information until the appropriate time.
  • Collaborate with internal and external partners
  • Serve as repository for syndicated Market Data and leverage that to develop forecasting algorithms
  • Aggregation/Disaggregation
  • Contextual assumption creation and sharing
  • Versioning forecasts
  • Accuracy reporting such as Forecast Accuracy, Forecast Value Add
  • What-if planning – What’s the sensitivity of the forecast to a change in a certain variable?
  • Scenario planning – How will changes in business drivers impact input variables, which in turn will impact the output forecast?

If I missed something, please let me know in the comments.

At o9, we used these requirements to architect our decision management platform, mPower. The result is that large-scale enterprises looking to adopt a more rapid budgeting process can now do so a frequency aligned with the pace of this world, all without increasing the burden on the workforce.

Lastly, for you finance folks, if “mark to market” is your rallying cry, we hear you. Come with us and we’ll take you on a journey that moves your company toward an ideal, reality based budget.

Suneet Upadrasta