You are here
5 Tips to Help You Forecast Sales
5 Tips to Help You Forecast Sales
It seems like the basic sales forecast is one of the hardest pieces of business planning for most people to deal with. I think if we dug deep enough w;d discover ther-s a fear of forecasting at the bottom of the problem. And -m hoping I can help with that by offering these five tips.
1. W're Not Really Supposed to Know the Future. W're Guessing.
Somewhere along the way I realized that a lot of people assume that somebody else knows better. As if having an MBA degree or CPA certification means somebody has a magic bag of tricks for forecasting, tricks that other people do't have. The truth, however, is more like good old fashioned educated guessing.
I do have an MBA degree, and I was also a vice president in a market research firm for several years. We produced a lot of forecasts. They were so much more educated guessing than anything else. It was't a matter of econometric modeling, surveys, or detailed research. Instead, w'd talk to product managers, compare opinions, do secondary research, and make educated guesses.
And quite often, the best forecasts were the educated guesses of people in the front lines of the business. Sure, we needed to research as much as we could, but that was to educate the guesses. They were still guesses.
2. You are't wrong for very long. Yo're Going to Revise
I's not like you have to forecast the future for that whole year in advance, although that's what a lot of people fear. Back in the real world, though, you forecast for a year but you follow the plan vs. actual results at the end of each month, and when things go different from the forecast - and they always do - you revise.
Look especially to where assumptions have changed. When assumptions change, your forecast changes.
You should never go longer than a month or two with a bad forecast. Revise it.
3. Break it Into Manageable Drivers
Don't just pull a forecast out of the air. Look at the components. Find the drivers. Almost any business, service or not, can be broken into sales equal to units times price. Units could be days, hours, jobs, trips ... you can still break sales into component parts.
For example, on all things web, look for drivers like page views, search terms and ad words. Then there are conversion rates, and unique visitors. Find a way to relate the numbers.
For a restaurant, figure out what's capacity - use the numbers of chairs and tables to determine how many meals, like breakfasts or lunches or dinners, you can sell in an average day. Then figure out how many days in a month you'll be at capacity.
For a limousine service, it's how many trips per day, week, or month. And what's the price of an average trip.
4. Find The Right Level of Granularity
Forecasting requires a sense of summarizing and aggregation. The bookstore forecasts by main lines of sales, not every book. The restaurant averages out the meals, the drinks, and so forth. If you sell hundreds of different things, find a way to summarize into main categories.
Don't forecast the whole business as one single row of sales. And don't break it into 50 rows of sales. Find the sweet spot, like 3, 5, 10 rows of sales. Find the level of granularity.
5. Don't Look for Answers. Generate Your Own.
One of the most common problems with forecasting is the idea that you're searching, during some hallowed research phase, for right answers. As if there is a standard growth rate, or standard forecast, which, if you find it, will make the whole thing make sense.
Sure, educate your guesses, but it's not a matter of searching online until you get the right study. It's a matter of developing a forecast you can live with, then tracking it, revising as assumptions change, and living with it over the long term.
*Not a government website.
About the Author: