Bloggo back to the blog
Development Intelligence within an Application Lifecycle Management (ALM) solution.-->
Business and technology studies confirm the commonsense adage that organizations can only manage what they can measure. Research has shown time and again the power of goal-setting and measurement in influencing organizational behaviour.
Indeed, Research shows that only 15% of Fortune 500 firms have complete software development measures in place; and fewer than half have quality measures. And we would expect these numbers to only go downhill for smaller companies with fewer resources to spare.
Why are we so bad at measuring, even when we know the good that can come from it?
What stops most of us in our tracks is the difficulty of capturing consistent and meaningful measures. In order to truly succeed at development intelligence, your environment needs to be able to do two
things. First, it needs to capture metrics automatically and behind the scenes, minimizing
administrative busywork. And next, it needs to be able to surface those metrics in real time, at the
right level of detail for the task at hand.
Without development intelligence, managers spend most of their time asking for information from their teams. With it, they can become a manager hero – a manager who is able to positively influence the success and direction of his or her projects.
According to Capers Jones, projects with strong measurement practices have much better success rates than those that do not – Capers Jones, “Measurement, Metrics and Industry Leadership,” 2009, and Software Engineering Best Practices, McGraw Hill, 2010.
ALM provides for measurement; studies strongly correlate measurement to project success. For example, the three measurements listed below are practiced by less than a 50% of all organizations in the Capers Jones study:
Quality measures: 45%
Productivity measures: 30%
Complete measures: 15%
Here are some suggested do’s and don’ts regarding measurement practice.
DON’T Ignore performance measures.
Define performance metrics that are appropriate for your organization.Simple metrics such as Build Duration, Build Pass/Fail rate are simple place to start if you haven’t already.
DON’T Take a ‘big bang’ approach to instituting measures and metrics.
Identify a weak spot. Choose a practice to implement improvement. Determine how you will measure the improvement. Choose a tool that collects and reports on the team’s activity using the data from
DON’T Expect to get it right the first time.
Conduct retrospectives and identify the next set of improvements.
DON’T Try to manually collect data by hounding the team for status reports.
Use live dashboards that provide transparency of information and dashboard reports based on data
coming from the team’s activity.