Improving Sales Forecast Accuracy with a PERTy Formula

By Scott Allen on 15 March 2011 (Updated 25 March 2011) 0 comments
Photo: Maica

If there is just one aspect of strategic sales planning that all companies would love to improve on, it would probably be more accurate sales forecasts. For some, sales forecasting is a little guesswork mixed in with some black magic and a lot of luck. However, for others, it’s an involved process predicated on accuracy and continuous improvement.

Companies that excel in strategic sales planning do so by constantly raising the bar on their forecast accuracy. They understand that accurate forecasting means accurate inventory counts and accurate cash flow planning. After all, sales drives inventory. Without an accurate sales forecast, firms are left flying blind. And without an accurate cash flow forecast, potential cash shortfalls can be both costly and disruptive.

What is it that these businesses do that others don’t? More importantly, what can the rest of us learn from the best when it comes to improving sales forecast accuracy?

Sales Targets are NOT Sales Forecasts

The biggest mistake many companies make is trying to reverse-engineer their sales forecasting numbers. They start with how much money they want to make and set sales targets from there, often without any input from the salespeople. They’re often trying to get to the numbers they think someone else – usually investors – want to see. Or, they may ask the salespeople for their forecasts, but they only ask them for one number.

The problem with both of these approaches is that entrepreneurs and salespeople tend to be overly optimistic. That’s not necessarily a bad thing – it comes with the job. It’s what allows them to handle repeated rejection or failure and still go to work every day with a positive attitude.

Accountants, on the other hand, tend to be pessimists. They’re the ones who have to make sure that payroll gets met. They’re the ones who have to walk the line between making sure that sufficient inventory is available to meet customer demand, but that costs are controlled by not having dead or stagnant inventory.

A more refined approach solicits the input of the sales staff and then balances it with a healthy dose of realism.

Sales Forecasts are Based on Sales Totals to Individual Customers

Every sales forecast starts with the sales professional reviewing business at individual accounts. The sum of these individual accounts then becomes the forecast for the entire territory. But what is the analyst looking for When it comes to sales forecasts, every sales professional essentially asks three questions:

  • What’s the best possible outcome (outcome "A")
  • What’s the most likely outcome (outcome "B")
  • What’s the worst possible outcome (outcome "C")

These three variables (A, B, C) form the basis of an analytic approach that was initially adopted by the United States Navy in its design of the Polaris submarine. Faced with multiple variables, and a number of critical paths while managing the project, the Navy came up with Project Evaluation & Review Technique (PERT). The principle behind PERT is to use these three aforementioned variables and assign values to each. Next, a calculation is then used to come up with the most accurate estimate possible.

PERT analysis calculation = {1(A) + 4(B) +1(C)} / 6

Now, you might be asking how PERT can be used to help increase the accuracy of sales forecasts. Simply put, the salesperson assigns a value to each variable. This value is based on the number of units the salesperson believes she can sell to each particular account. Let’s assume the salesperson feels the best possible outcome would be to sell this customer 20 units. The most likely outcome would be to sell 15 units, and the worst possible outcome would be to sell 10 units. These values would then be placed inside the PERT calculation.

  • What’s the best possible outcome = A = 20 units
  • What’s the most likely outcome = B = 15 units
  • What’s the worst possible outcome = C = 10 units

PERT analysis calculation = {1(A) + 4(B) +1(C)} / 6

PERT analysis calculation = {1(20) + 4(15) + 1(10)} / 6

PERT analysis calculation = {90}/6

PERT analysis calculation = 15

In this example, the most likely total to be sold to this customer is 15 units. The sales professional could forecast 15 units with some confidence. However, even with this approach there's no guarantee. The PERT calculation may help to narrow down the figure, but it’s still incumbent upon the sales staff to close those sales.

Most sales forecasting software programs adopt some variant of this approach. They require the salesperson fill in some values for each customer account and then the program comes up with the most likely total to be sold across an entire territory. However, for those businesses that simply can’t afford to spend lavishly on forecasting software, PERT works extremely well. The calculation itself is very straightforward and can easily be implemented in any spreadsheet application.

You can still set ambitious sales targets if you choose, but setting unattainable goals that people consistently fall short of doesn’t really serve anyone well. Now, instead of setting those targets based entirely on the desired outcome, or on overly optimistic projections, you can incorporate truly the best possible estimate into the rest of your projections.

0
No votes yet
Your rating: None
ShareThis

comments

0 discussions

Add New Comment

CAPTCHA
This test helps prevent automated spam submissions.