Economic forecasting models can get very complicated and it is easy to lose touch with reality. Assumption on assumption gets added, new market segments come in, then your manager asks you to bump up that growth by 5% per year.
Your survival? Take the current year as a starting point, and break down your total market in actual physical drivers. Then check whether these same drivers still make sense 10 years later.
Let’s take the example of the coffee market. $ spend by customer segment that growth at a certain rate over time are abstract concepts. Cups, liters, people, price per cup are things you can touch and relate to. If you think there is a premium market segment that pays double the price per liter, then you need to back out the opposite segment that pays a lot less to get back to the average. Does that reverse engineered price make sense? If the coffee market doubles, but people pay the same price, and the population isn’t really growing, where do these liters come from?
Break down today’s total market into factors you can touch
Forecast these factors (not the total numbers)
Build up the total market from these factors
Sanity check and go back to 1
In most cases, you will discover that only a factors really make the difference. And if you keep yourself grounded to reality, you can pretty much include any breakdown, split, scenario, as long as the totals and averages stay the way they were.