Listen Now
Subscribe on iTunes
Check out the podcast on Google Play Music

The Software Process and Measurement Cast 433 features our interview with Jeff Dalton discussing holacracy. defines holacracy as, “a complete, packaged system for self-management in organizations. Holacracy replaces the traditional management hierarchy with a new peer-to-peer “operating system” that increases transparency, accountability, and organizational agility.” Jeff has implemented holacracy in his own firm and others and has a lot to share about this exciting form of management and leadership.

Jeff Dalton is President of Broadsword, a Process Innovation firm, and Chief Evangelist at, an Agile Leadership Research and Development center that develops models for high-performing agile teams.  Jeff is principle author of “A Guide to Scrum and CMMI,” published by the CMMI Institute, and is a SCAMPI Lead Appraiser and Certified Agile Leadership Consultant that specializes in software product development, self-organizing teams, and performance modeling.  His upcoming book, the “Agile Performance Holarchy: A New Model for Outrageously High Performance” will be released in September of 2017. (more…)

Outsourcing:  Metrics and Governance
Thomas M Cagley Jr.

Successful outsourcing arrangements require that all parties pay continuous attention to the goals of the arrangement.  The phrase ‘continuous attention’ is nearly an oxymoron given differing corporate cultures, short attention spans and competing channels of information.  Outsourcing contracts often contain one or more tools intended to focus attention (typically as a portion of the governance convents). The tools currently in vogue include metrics, service level agreements, balanced scorecards and the prescription of CMMI®[1] maturity levels.   In many cases these tools are used as weapons to interrogate and control the parties in the contract.

The goal of governance is to manage and reduce risk. A better practice is to construct governance covenants so that they build and strengthen the capabilities of both parties. CMMI, metrics and scorecards done correctly can fall in this category while at the same time mitigating risk. The choice of tool and the construct of the governance convents are as critical as the financial portions of the contract. All too often, the development of the governance processes is by contract personnel without a through understanding of process or metrics. The lack of process or metrics knowledge is exacerbated when the idea is championed by parties on one side of the table or other (Note: Someone has to champion the concept or most outsourcers will run, not walk, away from the idea.  Each side of the negotiation needs to have a metrics counsel.). These scenarios lead to agreements governed by convents that incent unbalanced behavior, that do not add capabilities or strengthen the relationships between parties to the contract.  Examples can be found across a range from contracts which pay penalties and bonuses on a single metrics (typically indexed productivity or time-to-market) to contracts with scores of embedded metrics which sap time and effort from delivering value. Soliciting input from and the involvement of your organizations process and metrics personnel (or consultants if you do not have these capabilities) is a best practice that does not find its way into many contract negotiations.

There are many techniques that can be used to facilitate developing a balanced approach to using metrics in the governance process. Goal driven metrics techniques are processes for developing useful metrics for the core of contract governance.  A few of the techniques that have been found useful include:

  1. Goal, Question, Metric (GQM)
  2. Practical Software Measurement (PSM)
  3. Goal-Question-Indicator-Measurement (GQIM)


A basic tenant of all of these techniques is to begin with an understanding of your business goals. A through and honest appraisal of the reasons why you are outsourcing will allow you to build a set of measures that have a chance at allowing you to know whether those goals have been met. The set of measures will also provide a filter to discriminate between what data will provide information and which will provide extraneous noise (this requires a careful and thoughtful understanding of data analysis).  Without the use of goal driven approach it is possible that the contract will incent behavior that does not accurately reflect the goals, targets and objectives of executive management (an old adage points out that you get what you measure).

As noted earlier the two worst-case scenarios typically are a focus on a single goal (cost reduction, increased productivity or better quality) or the development of scorecards with too many contributors (if you need an econometric model to understand the impact of metric you certainly have gone too far). Kaplan and Norton’s work on the Balanced Scorecard has certainly driven the point home that a single view of performance is at best not very useful and more than likely dangerous.  At the same time balanced and overly complex do not have to be synonymous concepts.

Dashboards and scorecards are excellent summarization techniques however they are not adequate to develop knowledge or to direct action. A granular level of detail will provide organizations with the facility to have event-centric and experience-based interaction with data in order to generate actionable insights rather than to rely purely on personnel interactions. Granular, query-able data will allow you to develop mechanisms to defeat the tyranny of average, which looses the ability to identify special cases (both good and bad).

Continuous attention does not have to be an oxymoron.  It does mean all parties in an outsourcing arrangement must decide on what is important when the contract is being framed (and have the guts to change when the business dictates). This process provides a basis from which to develop a well-grounded metrics program that will let you know if you are attaining your goals while increasing capability and value.

[1] Capability Maturity Model Integration® and CMMI® are registered trademarks of Carnegie Mellon University.