We are a specialty consulting firm exclusively focused on measuring and improving the financial return from marketing investments. With experience in dozens of industries, we use a broad toolkit of unique approaches to find just the right way to break through the political, cultural, and structural obstacles to help you crack your toughest measurement challenges.

View our two new webcast series, designed to help you meet the challenges of marketing measurement and resource allocation.

How Much Will We Sell Next Year?

The marketing director at Beverage Company "D" was pulled out of a meeting by his vice president one day. In the hallway, the vice president told him that she had a meeting with the CFO in an hour and needed as accurate a sales forecast for the next two years as he could provide for the company's best-selling liquor product. Worse yet, she needed it by month. After a brief discussion, they agreed that for simplicity he could assume there would be no material changes in the competitive marketplace over that 2-year forecast time.

The marketing director went back to his computer and pulled down the last four years of sales data by month for the product in question. Everyone knew that sales were seasonal, with the majority taking place over the holiday months and a second (much smaller) spike in the summer. Then, he stared at his computer wondering if he would be able to get an analyst from finance freed up in time to help.

Well-armed only with a $100 Microsoft Excel plug-in, we solved the problem in less than 15 minutes.

First, we recreated the chart of monthly sales data for each of the past four years.

Next, we ran the simple forecast macro included in XLForecast by Analycorp and requested a 2-year projection of the present trends (by month), taking into account seasonality.

But to be as accurate as possible, we needed to test the resulting forecast model. So we decided to see how the model would have worked in Year 0 had we attempted to predict sales for the year just past. We had to tune our forecast model a bit to match the most recent sales using Excel's built-in Solver routine. Here's the forecast chart:

With this forecast so closely tracking against Year 0 actual sales, we felt pretty confident in our approach to forecasting Years +1 and +2. As you can see from the chart below, our forecast model was better than 98% accurate when compared to actual data.

So using our new forecast model, we were able to get the answer to the vice president with plenty of time for a discussion of the risks of assuming that marketplace conditions wouldn't change over the next two years.