Building Effective Measurement Solutions Workshop
Measurement Challenges
- Goals for measurement
- Tell a story of the value of testing
- Estimate number of test cases for test planning
- Enterprise level measurement
- Standards across multiple teams
Reasons for Measurement
- Drive improvement
- Planning of testing
- Demonstrate value of testing
- Provide testing status and control
- Improve testing efficiency
How to Establish Enterprise Metrics
- Different areas have different goals
- Different areas have different challenges
- Overarching goal - make more money
- Investment in automation
- Reduction in cost (through offshoring)
- Reduction in time to market
- Create a pallet of metrics
Group 1: Demonstrating the Value of Testing
Name of Metric: Test ROI
Description: Value of doing testing
Formula: (Benefit - Cost) / Cost
Measures: Test Efficiency
Time to Test
Time to Fix
Source: Defect counts
Time to Test = (Test Time) x (% Test Complete)
Test Efficiency = (# of defects in test) / (total defects)
Benefit=Quality product
Business efficiency
Meet requirements
Legislative compliance
Costs= Cost to test
Cost to fix
Cost to re-test
Back to top
Group 2: Providing Testing Status and Control
Name of Metric: % test cases executed
Description: Test execution progress (point in time)
Formula: (# test cases run) / (total # test cases planned)
Measures: # test cases
# test cases executed
Time to Fix
Source: Test Management Tool
When gathered: During execution cycle
When reported: Monitored daily
Reported weekly
Analysis: Redirect resources
Dashboard: % test cases executed - where we are at now
1st run fail rate - how much more testing will be needed
% test cases blocked - how much more to do
Back to top
Group 3: Improve Testing Efficiency
Name of Metric: Improvement (Efficiency)
Description: Established Baseline - Monitored change => improvement or not
How do we do earlier defect discovery in smoke tests?
Formula: (defects) x (effort) / size = defect hours per unit of work
Measures: # of defects in 1st run
When defect discovered (which phase)
Size (e.g. dev. hrs. for change, lines of code, function points)
Effort (# of testers)
Source:
Defects <= Quality Center
Phase <= Quality Center
Size <= # of requirements
Effort <= E-Time (time reporting system)
Analysis:
- Establish Baseline
- Frequency of collection - every time we run smoke test or on demand
- Frequency of reporting - monthly
- Who / how to collect?
- Tools to use / record
- Collect 3-4 months and establish baseline (mean)
- Measure / Review Improvement
- Threshold? + or - 3 standard deviations
- ^ threshold = go look at change
Reference: W. Ed. Deming "Out of the Crisis"
Back to top
Post-Conference: Agile Panel Discussion | Better Test Planning Solutions Workshop | Risk Management Workshop
Building Effective Measurement Solutions Workshop | Challenges of Test Automation Solutions Workshop