In just a few minutes, using some basic statistical analysis, you can predict the maximum number of agreements any given location, district, region, or company is capable of producing.

"For a store to improve, it must be out of control"

When a colleague asks, "How's your month?", most of us respond based on last months numbers. "We gained 6 agreements", or "We're up 70 BOR from the same month last year". Looking at your results in those narrow terms is not only inaccurate, but also dangerous. Having such a narrow view of your results can give a false sense of security.

Being "up from last month, or last year" may give you a 'warm and fuzzy feeling', but it is absolutely meaningless as a gauge of long term business health. Though a bit geeky, the proper response would be..."I'm out of control on the high side!"

Every system is perfectly designed to achieve the results it achieves
If, when your agreement count drops, your first instinct is to 'have a store meeting to discuss performance'...you're not paying attention. Your store is producing X number of agreements because that is what the system is capable of. A robust system will thrive even if an individuals performance slips.

If your store has been open for more than a few years, you have noticed some leveling off in gain. Using statistical analysis, it is possible to, not only accurately predict revenue, but also predict with certainty the number of agreements your store is capable of producing.

For the purposes of this exercise, we'll use 3 years worth of fictitious agreements  Our objective will be to predict the maximum and minimum number of agreements we can expect from this location. You will need at least 3 years of monthly agreement numbers to accurately predict your process capability limits.

First, lets get familiar with some basic terms.

Control Limits
Upper and Lower control limits are calculated to let you know if your store is 'in control (stable), or out of control (trending up or down). Variation within these control limits is normal. Being 'Out of control' on the high side is a good thing... more on this later.

Process Capability Limits
The upper and lower limits that your store (system) is capable of producing. If done correctly you can predict, with up to 99% level of confidence, the number of agreements any location, district, region, or company is capable of producing.

X (Pronounced "X Bar)
X Bar is the average agreement count for three months
example: (380+412+392) / 3 = X (these X will represent the individual data points on our graph)

_
X  (pronounced "X Double Bar") also called 'mean'
X Double Bar is the average of the averages
example: (Q1+Q2+Q3+Q4+...etc) / (number of quarters being analyzed) = X Double Bar
 

R (Range)
The difference between the highest and lowest number of agreements in a 3 month period (X)
example: In Q1, (Feb)412, minus (Jan)380 = R (in this case R = 32)

R (pronounced "R Bar")
R Bar is the average of all ranges (R)

A₂ (Pronounced "A Two")
Used to calculate Control Limits to a high confidence level. It is dependant upon the sample size. When your sample size is 3, ie: each point on the graph represents the avg of 3 months, then A₂ = 1.023 Note: The bigger 'n' is, the more accurate the analysis.

d₂ (Pronounced "D Two")
Used to calculate Process Capability Limits to a high confidence level. It is dependant upon the sample size.

The first table shows the raw agreement count by month for 3 years, the X bar, R, R, Upper and Lower Control Limits, and Upper and Lower Process Capability Limits.  The graph below shows a statistical analysis of this particular 36 month period.

Calculations
The calculations are included HERE in case you want to 'try this at home'.

Rules of thumb
A system is said to be out of control if...

• Any 1 point is lying beyond the control limits
• 3 or more consecutive points lying on one side of X 
• 5 or more consecutive points going in the same direction (indicates a trend)

According to our chart, we can tell many things...

• The store is in statistical control and will range between 379 and 426. That means that our system is designed to produce a 16% swing in any given quarter. (Difference between Upper Control Limit and Lower Control Limit)
• Something statistically significant occurred in year 2 (All 4 points are 'above the mean (X Double Bar) What was it? Can we reproduce it?
• Most importantly, Unless the system changes, this store will never, ever produce more than 444 agreements or less than 362 agreements (Upper process capability limit and Lower process capability limit) because that is what it is designed to produce.

If you ever hope to increase above 444, a true system change must occur. System changes are not easy. They require change on every level of operations.

Do one thing at a time, do it extremely well, then move on to something else
If you're satisfied with never exceeding 444 agreements, and potentially dropping to 362...then go fishing. If you want to narrow that range or grow...

• Analyze your system
• Make one or two significant changes
• Track the results for a minimum 3 months
• Chart the results. Do you see a statistically significant change? (3 points above the mean (X)
• Implement the next system change and chart the results, etc, etc, etc.

Statistical analysis, while not sexy, is the only accurate method of determining whether or not your system is improving, or is even capable of improvement.