-
Notifications
You must be signed in to change notification settings - Fork 0
Business Statistics
Kyle Coberly edited this page Jul 9, 2017
·
1 revision
Measure of how frequently something occurs.
- What our average turnaround time is
Type of histogram, ranked in order of most frequently occuring to least. Used for defect tracking.
- What are our most frequent errors?
- What caused a project to be late?
- What processes take the longest?
Chart that gives the mean over a specific number of values. Takes into account past performance, but weighted toward more recent performance. Used to generate estimates.
- How long will a project take?
- How much will a project cost?
- What is our throughput?
Measures the shape of a distribution.
- How consistent is a process?
- Compare same process over time
- Compare different processes
Chart that estimates the probability of rare events. Can be used to predict frequency of errors.
- How many resources do we need to dedicate to troubleshooting?
- How many resources do we need to dedicate to change requests?
Mean, and upper/lower control of +/- 3 sigmas. If something is outside of that, the process is "out of control." Other bad signs are 8+ consecutive above/beow, 10/11, 12/14, trend over 6+, 2/3 are in the outer 1/3, 4/5 are in the outer 2/3s, the centerline shifting from chart to chart.
Split of our work
- Cost centers, revenue centers, profit centers
- estimates vs. actuals
- How long do we wait / keep people waiting for info?
- New vs. Returning customers
- Close rate
- Benchmarking
- JIT
- Reengineering ("Blank Sheet of Paper")
- Map out every process, then Pareto, then Ishikawa