YES WE CAN Align our Volume Forecasts with our Financial Forecasts and Work with a Single Number for a Long Term Time Horizon

Mark Covas

Mark Covas

Dr. Jain (Chief Editor of the IBF’s Journal of Business Forecasting), and I have recently been collaborating on an article that is a wake up call for Planning organizations to have their house in order so they can recognize the signals for an economic recovery. I am a proponent for completeness and integrity of the forecast across the full planning horizon. In fact, right now I would bet most Planners are only focused on the next 3 months. Of course, not all industries are alike.  So forecast horizons and forecast buckets differ (weekly vs. monthly), but in most cases, the further out in time you go the less attention the organization gives to the quality of the forecast. You can often see year over year comparisons where volumes may actually show a decline only because the Planners have not included future promotions or new product launch activity. How can we give Senior Management confidence in the sufficiency of our volume forecast to meet our growth goals if we are not giving the same focus of attention to the 18, 24, or 36 month out forecast that we give the 3, 6, and 9 month out forecast? The discussion I would like to generate is how ‘Yes We Can’ align our volume forecast with our financial forecast in such a way that we can work with a single number forecast for at least an 18-24 month time horizon. Plus, how the S&OP process can close those gaps and provide better forward visibility.

As Anish Jain (Managing Director of the IBF) has planned these presentation paths at the IBF’s Orlando Conference, you can see that one discussion easily builds on another so we end up with a chain of conversations over the course of 3 days.  This helps to fully vet the possible options for us take different best practices back into our work environments.  As a member of the IBF for over 10 years and a huge fan of the Journal of Business Forecasting, I have to say that attending the Best Practices conference in Orlando is always my favorite event. To be able to speak this year is a great honor, but the highlight for me is meeting and sharing best practices with Planning experts from around the world and from every industry imaginable.

Your comments based on what works and does not work in your process to build forecast quality and integrity are welcome.  Hopefully, we can have a meaningful dialogue where we all benefit.

Mark Covas
Innovation Diamond Management
Procter & Gamble

See MARK COVAS SPEAK at The IBF’S:

$695 (USD) for 3 Full Days!

October 12-14, 2009
Orlando Florida USA

5 Responses to YES WE CAN Align our Volume Forecasts with our Financial Forecasts and Work with a Single Number for a Long Term Time Horizon

  1. I would first define what each time horizon is meant to do and who owns it. Example:
    1. Months 0-6 (operational) belong to operations to load and ship product.
    2. Months 7-12 (tactical) belong to sales to focus on quota
    3. Months 13-36 (strategic) belong to marketing to set promotions and set management objectives
    4. Months 37-60 (long view) belongs to capacity planning to look for space in Singapore or Budapest…. Read More
    5. Quarter and annual numbers belong to finance to see general trends and to report to management performance

    Each of these may/will have a forecast created by the different organizations. They should not overlap and they should closely meet. In this model the motto should “Mind the Gap… Read More”. (Think of the London Underground)

    There can be one forecast but it came from more than one organization. Be careful that you always finish the sentence with “Forecast of What?” What time-frame and what organization MUST accompany the title of the forecast. Top Management can facilitate this model by giving authority to the organizations and limiting their time horizon.

    How can we give Senior Management confidence in the sufficiency of our volume forecast to meet our growth goals if we are not giving the same focus of attention to the 18, 24, or 36 month out forecast that we give the 3, 6, and 9 month out forecast?

    You will not give the same focus to every segment of the forecast. You will give the most focus to… Read More the time-frame and organization that needs it now. The focus will change next month depending on what management needs.

    Remember the forecast only exists to allow better decisions to be made. Management makes those decisions. Senior Management confidence will be improved when they KNOW you are working for them.

    Do not confuse “who has the focus” with “who needs to do a forecast”; every organization must do their best and do their own. If there is one group (or better yet and application) with a specific charter to glue all of the separate pieces together, then everybody will be winners. That one group could facilitate best tools or best practices and help in consistent training across the company.

    I have done this several times inside of Hewlett Packard supplies. It is not easy and it takes a few years to get there, but it can be done.

    Emmet Jones

  2. Emmet – thanks for the great build on this topic. I hope we can get some other perspestives as well. Also hope you will be able to attend the Orlando conference. I would love to spend the hour I have generating this type of a discussion and just use the slide presentation to kick us off! Great insights!

  3. think the periodicity that Emmet mentions are important. The second criteria that is critical is being able to forecast the product mix. In our experience, if you can forecast the volume correctly (within acceptable error rates) but the mix forecast is off, you might as well have not forecasted.
    We had a scenario when we forecasted for an Ag equipment manufacturer. They have a spring cyclical pattern for their products. We were into March and the orders had not started coming in. But when the demand finally arrived, because the mix was correct, the order fulfillment was very high and lead time was fantastic. That is because the material/parts were for the right mix of machines, with detail of engine, transmission, tires, lights….etc. All the stuff that matters!!
    Read the paper at:
    http://www.emcien.com/index.php?option=com_content&view=article&id=129&Itemid=101

    I would like to see IBF put some focus on this. In these economic times, this level of mix forecasting is critical.

  4. In my opinion that you are completely wrong tryin to focus on forecasting 18-36 months out.

    Three reasons:

    1. The confidence interval of the forecast decreases as the time horizon increases

    2. Inability to predict future consumer behavior in a brutal economic climate where for the first time in a decade peopl are saving 5-6% of their income. Could you please tell me with 80% confidence when the unemployment rate will drop to 4-5% and consumers will start spending money like it’s 1999? Please share, I’d like to invest in the stock market 12 months prior to your estimated month-year.

    3. The pace of innovation, capital mobility and other factors could pulverize any competitive advantage that your firm might have and your competition could soon gain a huge chunk of your market share.

    Attempting to build a viable regression model that will incorporate all these variables would be an extremely arrogant exercise in futility.

    Let the big boys worry about cost avoidance, product innovation, sourcing opprtunities etc. You worry about what you can control: service and inventory levels, so that your company can maximize profits.

    The only thing is you can hope for is that executives have some foresight and will not run your company into the ground.

  5. John makes a very valid point that forecasting mix accurately is also key. To Emmet’s point, there are different needs of the forecast over time, and short term, accurate mix is essential because of the short amount of time to make an inventory correction. In the longer term horizon, and in trying to address Dan’s concern, the company may make a strategic decision to place greater focus on a category, say Baby Care. It takes several years to build a manufacturing facility and to build it in the right part of the world. With China transitioning out of a 1 child per family policy you could envision a greater need for Baby Care items in the long term. With innovation geared towards that economy we can see the need for a long term forecast. A concept I will talk about in Orlando that is used at P&G is the use of ‘building blocks’. Instead of building complex regression models we do use statistical forecasting models but layer on building blocks of activity that we believe will create incremental volume. Dan is correct that forecast error is likely to be greater farther out but that assumes you are forecasting at the same granular (SKU) level across all time horizons. For Baby Care, a 36 month out forecast for diapers (Product family)is fine for capacity planning. 18 months out we may need to build lines to support specific models of diapers (new born, toddler), and at 3-6 months out we will want to get the SKU by location right. What is interesting so far for me is the varied response based on the industry being represented. I look forward to seeing other people build on this conversation!

Leave a Reply

Your email address will not be published. Required fields are marked *

WordPress Anti Spam by WP-SpamShield