MISTAKE #69: Your volume forecast is over-inflated
Back when I was making my living in construction, building out neighborhoods with single-family residential homes, I thought I made conservative assumptions to properly manage the risks that came with carrying millions of dollars of debt.
As the weakness in the house marketing (and the economy overall) was becoming apparent in 2007, I remember telling my wife:
“Don’t worry, I’m pretty sure the worst case scenario would be just selling one home a month and we’re still in great shape if that happens.”
Of course, I’m an idiot and was facing the challenge of staying solvent 9 months later after we hadn’t sold a single house and my world was imploding.
While forecasting sales of $300,000 homes in a single neighborhood is very different than forecasting sales of a $3.00 item in 3,000 stores, here are six principles I believe can still apply to both:
Don’t let irrational exuberance get the best of you: As Alan Greenspan observed when he first popularized this term, the excessive rise and fall of stock prices reflect the human nature to get overly excited about an opportunity that can feed other people’s over-excitement. Don’t let the optimization that your new product is a big idea lead you down a path of unrealistic expectations.
Make conservative, well-supported assumptions: Volume forecasts are full of assumptions…particularly during year one when there is no hard sales history data. Before you evaluate the numbers, focus on the validity of the assumptions made to support them. Consider the range of possible scenarios for each, not just the most likely scenario or the most optimistic scenario.
Realize how several small errors can compound into big errors: For simplicities sake, let’s use the components of the volume formula as the volume forecast for your product and you’ve projected each component to equal 1. So penetration, frequency, and transaction size each equal “1” (I know this doesn’t make practical sense). This would give you the world’s most rudimentary volume forecast of 1 (1*1*1=1). Now, let’s say each of these proved to be 5% too high, and the actual values were 0.95. With just three variables each off by 5%, your results fall 14% below your projections (0.95*0.95*0.95 = 0.86). Being 5% off on one component doesn’t sound like a big deal, but having sales 14% below estimates is going to give someone a really bad day.
The components you ignore can have a bigger impact than the ones you include: Complicated volume forecasts aren’t always any better or any more accurate than simpler forecasts due to all the assumptions. However, forecasts that are missing critical components can be almost worthless. Most important to consider here is how competitive response is accounted for. This is perhaps one of the most difficult to predict in both the timing and size of the impact. However, a lack of any realistic assumptions at all in these areas can make a new product almost helpless. In the case of competitive response, the results not only drag down the topline sales, but those lost sales could deplete the little funding you had set aside for your own counterattack.
Let your plans drive your volume forecast. Don’t let your volume forecast drive your plans: While I’m a big fan of top-down planning (set an objective and then figure out what needs to happen to accomplish it) in other areas of business, I believe volume forecasts need to be bottom-up (look at how well different components can be delivered and sum up what they’ll produce to arrive at your objective…with a little extra to encourage stretching for more). Top-down forecasting encourages the fabrication of unrealistic assumptions or components in the name of supporting an equally unrealistic overall number.
Use simulators to assess downside risk, not create the perception of greater upside potential: Hopefully, your volume forecast already exists in Excel and is all formula-based. It might be in a single tab and printable on a single piece of 8 ½ by 11 paper. Or it might contain a dozen tabs and over 100 inputs. Regardless of the complexity, you should spend time running through the impact different scenarios will have on the overall results. Doing this, of course, will reveal which components have the biggest impact. This can create the temptation to tweak numbers to significantly grow projections (like increasing the overall forecast by 25% simply by increasing the anticipated trial from 4% to 5% but offering no concrete tactic to deliver that big bump in awareness). Instead, use the simulator to help identify the critical components that need to be delivered or the activations that absolutely need to be funded even when sales start coming in slower than expected.
While I’ve helped clients build volume forecasting models, this is not an area I specialize in. However, it is yet another area where major mistakes can be made by fudging a few numbers only to have the error cascade into compromising other areas that jeopardize the overall viability of the product.
Make sure someone with significant volume forecasting experience and conservative views is responsible for this part of your business. Do yourself a favor and also give them the authority, without repercussions, to play the 10th man and respectfully challenge the optimistic assumptions of others.
It is best to build your plans off an accurate volume forecast. Lacking the ability to guarantee the accuracy, it is better to build your plans based on a conservative forecast than to find yourself with unrealistic, under-funded plans that, in hindsight, were fantasies from the start.