Pareto analysis ranks recurring software failures by frequency, directing limited investigation time toward the largest contributors. It identifies what to investigate first, not why the failure occurs. section 8.2.3
Table 8.1 Frequency of errors observed
|
Error category
|
Count
|
%
|
Cumulative
|
|
Database timeout
|
512
|
60%
|
60%
|
|
Missing auth token
|
201
|
24%
|
84%
|
|
Invalid JSON payload
|
89
|
11%
|
95%
|
|
Network Failure
|
45
|
5%
|
100%
|
|
Total
|
847
|
|
|
The error counts show how a small number of categories can dominate the total: database timeouts account for 60% of observed failures. section 8.2.3
Prioritize, then investigate
In a week with 847 logged errors, database timeouts account for 512. Addressing them with a missing index eliminates more than half the error volume; use 5 Whys or fishbone analysis to find the underlying cause. section 8.2.3
Pareto analysis can work alongside 5 Whys, fishbone analysis, and Is/Is Not in root cause analysis: prioritize the problem, trace its cause, and check for missed failures. section 8.2.5
