When negotiating anything that includes a formula, make sure that you thoroughly understand the formula and what makes up each element of the formula to understand the formula’s impact. Do an example using real numbers to see if the formula works and does exactly what you want it to do. When using percentages, understand and manage the two components of that formula, the numerator and the denominator.
In a negotiation I had a Supplier propose the following formula to calculate what an epidemic defect was:
“For purposes of this section, Epidemic Defects shall mean the occurrence of multiple failures of the same component, for the same root cause, to the extent that the world-wide annualized failure rate (AFR) for the identified defect(s) for a period of 3 (three) consecutive months the annualized failure rate exceeds five percent (5%) or zero point four, one seven percent (0.417%) per month.“
Aside from the wording problems, when I read this the first time my head was spinning. For anyone that is not familiar with epidemic defects, its usually a clause that Buyer’s want to include to protect against cost they incur when there are an excessive amount of defects.
In negotiation anything that is based upon a percentage, it needs to be clear how both the numerator and denominator will be calculated to calculate the percentage. For example, if you were to base the rate on defects in a thirty-day period, you would want the numerator and denominator to be based on that same thirty day period. If you were going to measure it against the Products under warranty that were defective (which would be your numerator), you would want the denominator to be the total products under warranty. If you measure against different period, you need to take that into account in the rate you agree upon. For example if you have a short measurement period for defects (the Numerator) being measured against a much larger volume such as annual purchases (the denominator) would require a much smaller percentage to have the same impact.
In the proposed language there was a mix of measurements that I think were meant to confuse. The reference to zero point four one seven percent (0.417%) per month appeared to be there to confuse the issue and make it look attractive. If you measure on a per month basis and you are at 0.417%, of that month’s purchases, your annual average should be no more than 0.417%. To get to the five percent (5%) number would require the measurement on the annual volume (0.417 times 12 = 5%). So what they were really proposing was the defects in a month had to exceed 0.417% of the total annual volume. The proposed language also required 3 consecutive months of failures exceeding the percentage. If they simply said if the average percentage failure rate exceeds 5% for three consecutive months it would be clear but the calculation would then be against a lower monthly number and that could make it easier to trigger their liability. If they said if the average percentage rate of defects with the 12-month period exceeded five percent, that would have been clear. You would take the total defects in the 12-month period and divide it by the annual volume.
From a worst- case perspective the combination of the two measurements had a huge potential impact. Here's an example
| Month | Defect Percent | Quantity of defects |
| 1 | 100% | 100,000 |
| 2 | 100% | 100,000 |
| 3 | 0.417% | 5,004 |
| 4 | 100% | 100,000 |
| 5 | 100% | 100,000 |
| 6 | 0.417% | 5,004 |
| 7 | 100% | 100,000 |
| 8 | 100% | 100,000 |
| 9 | 0.417% | 5,004 |
| 10 | 100% | 100,000 |
| 11 | 100% | 100,00 |
| 12 | 0.417% | 5,004 |
|
| Average failure rate 66.8% | Total failures 820,016 |
In the above example, there was never a three consecutive month period that had failures in each month during that period exceed the 0.417%. So it was never an epidemic defect even though the average failure rate for the period could have been well above the 5%.
Every time you use a percentage measurement that ties to a remedy Suppliers will always be concerned with it triggering at too low a real number. You can always deal with that by including a fixed amount requiring both be met for the trigger to occur. For example, if you agreed upon 1% as the rate and on average purchased 10,000 items a month, you could have the trigger be that it exceeds 1% and a minimum of 100 units so both conditions must be met. Conversely, if you were concerned that a percentage only measurement would be problematic with very high volumes you could use a fixed number tied to an “or” so either the percentage of the number would serve as a trigger such as “it exceeds 1% or 100 units” a month so either would trigger the remedy.
If what a Supplier proposes will either limit the numerator or expand the denominator that's used to calculate the percentage, you either need to say no or you need to adjust the percentage accordingly.
It’s also very important to carefully structure pricing formulas. I remember coming across a situation
in conjunction with a divestiture there was an on-going purchase commitment to purchase product from the company that bought business. They included a formula where future pricing would be set based on the average of 3 bids, one of which was the Supplier. Sounds simple enough, but with that type of formula combined with the fact that they had to buy from that one Supplier, what type of results do you think they got? The Supplier knew they would be awarded the Business no matter how high their bid. They had no incentive to include a competitive bid. In fact the higher they made it, since their bid was included it the total that would be divided by three to get the average, for every dollar they increased their price they would get to keep 33 cents. Someone obviously didn't think through the formal and the behavior it would drive. A smarter formula would have been to either not have the Supplier bid or throw out the high bid so the Supplier couldn’t manipulate the average. Never agree to a formula where the Supplier can unilaterally manipulate the price or anything else in their favor.
