This is a guest post by John Hawkey of freeforumofideas.com.
I came across this quote the other day:
“Plan financially for it to take you twice as long to sell half as much as in your original business plan.” (Toby Reid, director of BioCity Nottingham.)
It caused me to reflect on one of my obsessions from my earlier consulting days. This was the inability of startups to produce an accurate sales forecast. This is, after all, the most important forecast there is and every startup I was involved with got it wrong!
The importance of sales forecasting
Sales are the lifeblood of a business. Nothing is more important. Without sales there is no business.
A business’s operations are planned around the volume of sales it will generate. Sales volume influences cash (or capital) requirements, office or factory capacity, staffing levels, transport requirements, operational activities, in fact everything that happens in a business. Sales forecasting is, therefore, the vital ingredient of the business plan.
To accurately forecast sales in an established business is difficult for many reasons. These include the variations in the economic cycle, unusual events such as a pandemic, and more mundane issues such as increased competition, disruptions to supply and/or manufacture, strikes, staff problems, etc., etc.
However, with an established business one has historical data to work with. Actual sales from previous years set the benchmark from which estimates of future sales are made. Sensible sales assumptions based on presumed events within the economy and the business, as well as probable activity by competitors should lead to a forecast that is accurate within, perhaps, a ten percent variance.
However, with a new business (or startup) you have none of these benchmarks. To accurately forecast sales here is of a different dimension all together. There is, by definition, no history in a startup: there are no past sales on which to base future sales. Most things are a guess.
Startups in established industries
To help overcome the difficulty of having no history, potential startups in established industries undertake market research. This research attempts to establish such things as the size of the potential market, the activity of competitors and the likely share of the market that the startup is hoping to capture.
Information can also be gathered by:
Talking to vendors or wholesalers
Studying trade magazines
Considering how similar products preform
Looking at how rival businesses preform
Deciding whether your product is seasonal or not
Established industry forecasts
Using the methods above, startups in established industries can hope to produce sales forecasts that have some credibility. However, they are still usually inaccurate. Without exception they are inaccurate on the upside, plagued by mistakes such as assuming that sales will increase by a set percentage every year and happily feeding this information into a spread sheet.
The result is a forecast that is not only wildly optimistic about quantum, but also about timing. This is where Toby Reid’s comments above are so relevant.
I used to say to my startup clients: “Estimate the numbers, divide them by two, and then divide them by two again and then add a year.” Pessimistic? Maybe but, in the event, usually closer in outcome than the original numbers spewed out by the computer.
New business in new industry
Occasionally a startup is so innovative it is attempting to create a whole new industry. Although this is rare, it does happen. Where the new business is in this situation (that is, going into an industry which will be new), the process of forecasting its sales becomes doubly difficult or impossible.
The business cannot undertake the research mentioned above because there is no industry information, or competitor information or existing products to find out about. Where there are no industry sales statistics (because the industry does not yet exist), it is, of course, impossible to forecast what sales might be using tried and tested (albeit unreliable) methods.
Forecasting for a new industry
So, what is to be done about sales forecasting in these circumstances? Is there any point in attempting any forecast at all? Or put another way, can one place any credence on these forecasts, or are they just wild guesses?
The first answer is probably yes, you should produce sales forecasts. The second answer is no, you cannot place much credence on them. In these circumstances the startup owners must go on gut feel as much as anything else. The forecasts are probably just wild guesses and, therefore, mostly useless, and the business will have to be flexible and nimble and adjust to the sales actually achieved.
The inaccuracy of forecasting is only one element of a startup’s uncertainty. It is likely that the business will have to undergo numerous changes in its early life, such as adapting or changing its products, its location, its management and its method of sales to survive. This ability to “pivot” (to use Silicon Valley’s jargon) is vital to the continued existence and growth of many startups.
In the struggle for survival and emergence as a viable business the accuracy or otherwise of a startup’s initial sales forecast will have long since been forgotten. A bit like the relevance your GCSE results when you are seeking promotion to CEO.