I’ve made every forecasting mistake possible, but I still firmly believe in the
As part of a continuing series on forecasting, I’ve created 5 steps to help you build your own system! This is no theoretical list of steps… My steps have been developed from 30+ years of building (and using) forecasting models for clients and companies I’ve run. I’ve made every forecasting mistake possible, but I still firmly believe in the process. Let me share what I’ve learned:
Step 1: Assess your data. You will need almost all the financial data your organization generates. You may find you need to have more reports and data produced by your financial team, as the forecasting process requires you to use all your knowledge of how your business generates revenue and incurs expenses. I’ve often found that this process will generate new ideas about products or services or developing new competitive advantages.
Step 2: Organize your historical data. Since forecasts are frequently shared with others such as bankers and investors, it’s traditional (and I think best) to format your historical data as a common income statement or P&L. Most businesses develop their own P&L format, which is great. The important thing is: your forecast should be formatted exactly like your historical reports because you and your team already know how to read and consume that data.
Step 3: List Key Variables. Every organization has external factors that impact operations. The cost of raw materials, the price of gas, weather, interest rates, competitors and customers–all of these are your variables. They make business difficult to predict. Start listing these variables. It’s okay to begin with 3 to 5 just to get started. You can get more sophisticated and have as many variables as you choose after you have the hang of the process.
Step 4: Assume and calculate. You will need to make assumptions about each of your variables to begin your forecast. Using your historical data, begin calculating future amounts using the variables and your assumptions. For instance, if customer orders have been increasing 1% per month for the last 20 months, a reasonable assumption is that trend will continue. Continue this exercise for each line item in your P&L for at least 6 months into the future.
Step 5: Review and Analyze. Review the forecast the same way you would your monthly P&L. How does it look? Does it make sense? The evaluation of the process is completion of the process. Without the evaluation, there is no learning. Learning is the ultimate goal. Compare and determine why your forecast might be different from the actual results during the forecasted period and ask why. The why tells you how your assumptions were inaccurate and how you can make better assumptions next time.
Now, Go back to Step 3 and repeat, repeat, repeat.
You can learn more about the forecasting process from this article by McKinsey & Company. For more reading recommendations, or to discuss other entrepreneurial tips and tools, contact Harvard Grace Corporation at firstname.lastname@example.org.