David Tilston has kindly shared with us his guide to cashflow forecasting, a subject hugely important at this point in time, where liquidity is the order of the day, and you can read it here. The article prompted me to think back to my days at Unilever and elsewhere, where I implemented forecasting systems, and I thought it worth raising two areas for further thought:
a. David mentions inconsistency of assumptions; this can mean, as per his example, different people making different assumptions e.g. around volumes, but the other very significant factor which explains why forecasts are often flawed from the start is people’s differing definitions and interpretations of what a forecast is. For example: is it what you expect to happen should you take no particular action at all? Or is it after certain actions are taken? If you are consolidating forecasts which are based on an entirely different philosophy or definition, then the outcome won’t make any sense. So the issuing of guidelines and definitions to all involved in the process is hugely important
b. range forecasting is a very helpful approach, whereby you ask colleagues ‘how good could it be?’ vs ‘how bad could it get?’ You can then work out a most likely outcome (it shouldn’t usually be in the middle of the range, which demonstrates why you will get a different result from if you start with a single point forecast and ask for risks and opportunities around it)), but also have the richness of the qualitative discussion to hand, so you already have a clear picture of risks and opportunities and their real magnitude. You can add up all the different parts of the business on the basis of ranges, and then, using probability assumptions, focus in on the statistically most likely outcomes. This ultimate range for your outcome is something you can work with more effectively – not least because you have a chance of being right!
