Gallery > CMMI Progress

CMMI PROGRESS               



DataDrill has proven support for the Measurement and Analysis process requirements of the CMMI, including the complex quantitative needs of organizations at Levels 4 and 5.

Status by CMMI Key Practice Areas

cmmi

 

 

Automate the rigorous measurement and analysis demands of the CMMI or other standard adn clearly demonstrate progress and compliance.

CMMI Quilt Chart
cmmi

 

 

Measurement for Maturity and Process Improvement Using DataDrill Express

 

Introduction

The practice of measurement has been experiencing a well-deserved elevation in priority in recent years.  Demand is high across a wide range of industries for more complex software and software-intensive products that capitalize on rapidly advancing technologies.  Delivering these products requires new software project management control and monitoring to ensure that the end product is of top quality, delivered on time and within budget, and satisfies all customer requirements. The principle mechanism for providing managers with the information to achieve these goals is measurement. As such, a measurement process is a critical asset for executing business functions and delivering products.

This white paper provides an overview of the purpose, function and operation of an effective measurement process.  In addition, the use of measurement within the Capability Maturity Model Integration (CMMI) is presented, highlighting how measurement supports this popular process improvement framework. Guidance for implementing measurement is discussed, as well as requirements for using measurement within the CMMI, highlighting key aspects of the CMMI approach.

To help put the information here into action, the use of automated measurement tools is described using Distributive’s DataDrill Express product as an example. DataDrill Express provides a number of capabilities and features that speed the measurement process, especially for use in implementing CMMI and process improvement. DataDrill Express provides an off-the-shelf measurement solution that directly addresses the needs of the CMMI Measurement and Analysis process area.

DataDrill Express support for the CMMI falls into three essential areas:
•    contains a library of best practices measurement techniques that “jump start”
management and measurement deployment.
•    automates all key measurement tasks, including those identified in the CMMI Measurement & Analysis process area.
•    enables tailoring and configuration of every aspect of measurement including automatically configuring data collection based on IT project management needs.

The measurement guidance presented is appropriate for organizations embarking on process improvement, with or without a formal or informal/internal CMMI implementation assessment. This white paper is intended for executives, software managers, process group members and measurement leads who need an overview of the activities associated with the establishment of a measurement process.

 

The CMMI and Measurement

The Software Engineering Institute released the Capability Maturity Model for Software (SW-CMM) in 1990 as one of three separate Capability Maturity Models (CMM®).  After its initial release, the SW-CMM was adopted successfully by many organizations in many domains.  Released as a successor to the CMM, the CMM Integration (CMMI) integrates the three CMMs into a set of core process areas.  “The purpose of CMM Integration is to provide guidance for improving your organization’s processes and your ability to manage the development, acquisition, and maintenance of products or services. CMM Integration places proven approaches into a structure that helps your organization appraise its organizational maturity or process area capability, establish priorities for improvement, and implement these improvements.“ 1

 

CMMI Overview 

In the CMMI implementation models, process areas describe key aspects of such processes as configuration management, requirements management, product verification, systems integration, and many others. The most widely recognized depiction of the CMMI is shown below in Figure 1.



Figure 1. The Five Maturity Levels of the CMMI

An organization begins at Level 1, which is essentially an ad-hoc/non-repeatable set of processes, or simply an organization which has not been rated.  As process capability is demonstrated in specific areas, an organization is rated at one of the higher levels. The rating process is itself a formal activity involving assessors who are trained by the SEI in determining to what extent an organization actually has the capability which they claim. Informal assessments (i.e. those not performed by trained and certified assessors) can be useful at lower level organizations to identify gaps in the IT project management and engineering process.
_____________________
1 CMMISM for Systems Engineering, Software Engineering, Integrated Product and Process Development, and Supplier Sourcing (CMMI-SEISWIIPPDISS, V1.1) Continuous Representation, CMU/SEI-2002-TR-011, ESC-TR-2002-011,CMMI Product Team, March 2002 


Measurement and Analysis in the CMMI

In the CMMI implementation process, Measurement and Analysis is one of the areas that support the planning and management of other process areas.  Software managers use the measurement and analysis process to enter and establish plans, to review the progress of actual performance against the plan, and finally to take management action to identify and resolve issues of poor process or product performance.

The measurement process is the mechanism that organizations use to provide the information that managers need to manage the way that the organization requires them to. Another way to look at it is this: An organization establishes management policy that managers are expected to follow. As part of that policy, managers are required to review status information and then perform specific management tasks based on the status information. The measurement process is the mechanism by which the organization delivers the status information to managers.

Measurement and Analysis is defined and managed as a distinct process area, separate from other engineering and management processes. Other IT project management practices, as well as engineering practices, integrate measurement into themselves. Figure 2, below, highlights the key tasks of the Measurement and Analysis process area.


Figure 2.  Measurement support of key management practices


While performing any of the activities within the software or systems lifecycle, the project or program manager uses the measurement process to enter and establish plans, then reviews the progress of actual performance against the plan and finally takes software project management action to identify and resolve issues of process or product performance.

Within the CMMI, the Measurement and Analysis process area also acts as a general status mechanism for other process areas, allowing managers to plan and review progress as needed. The high-level view of measurement within the CMMI is shown in Figure 3. Notice that the information needed to manage processes, shown as “Goals and strategy” and “Process measures”, and to track progress of goals and strategy become the “requirements” of the measurement process.

 

Figure 3. Overview of Measurement Process


Setting Up a “Capable” Measurement Process

This section describes the measurement process capabilities needed to adequately support software, systems and other types of technical IT project management, including satisfying guidance provided in the CMMI.

Following, is the list of capabilities that your measurement process should provide:
1.   Ensure business goals and processes are measured
2.   Support the tasks within the measurement process
3.   Establish a measurement approach and plan
4.   Collect and analyze data
5.   Deliver periodic measurement reports to managers and stakeholders
6.   Provide a repository and evidence of the use of measurement
7.   Provide compliance and assessment reports in parallel with status
8.   Track progress and effectiveness of process improvement

The scale of the measurement process will depend on how many information needs are identified and how many managers are to become “customers” of the process. Most organizations start with a small set of information needs, typically in one or two process areas, and then expand over time. The capabilities described here should be present in any measurement approach, tool or solution, allowing your implementation to scale with your business needs.


Ensure that Business Goals and Processes Are Measured

One of the primary purposes of measurement is to satisfy the information needs of managers.  The CMMI implementation specifically requires that measurement be aligned with the business goals and needs of the organization. Business goals can come from annual or periodic strategic planning, process improvement initiatives as well as short-term executive projects. Process measures should be extracted into information needs from the appropriate engineering and technical documents such that the measurement process is satisfying the information needs for managing those processes.

The measurement process should be able to store your organization’s information needs and allow you to assign them to new projects or programs.  The measurement process should allow you to consistently apply an information need across projects. For example, the “Software Quality” information need would be defined once and then applied consistently to several projects — ensuring that the managers and stakeholders share a common vocabulary and reporting format.

An information need definition should contain descriptive text, references, management guidance for planning, monitoring and control, rules for determining status as well as graphs and report formats.  When possible, you should avoid capturing measurement process definitions manually in cumbersome documents because the cost and time spent updating and maintaining your information needs will become unmanageable as needs evolve and the scope of your measurement process increases.


Support the Tasks within the Measurement Process

To aid implementing CMMI, the Measurement and Analysis process area describes a set of measurement and analysis practices, shown in Figure 4 below. These tasks specify the functions to be performed and the work products/artifacts to be created. However, precisely what to measure is not specified in the CMMI. Instead, each organization determines what measures are important for managing their organizational goals, strategy and software engineering processes.

 

Figure 4. Measurement Process Tasks

Your measurement process should be able to perform these tasks.

For more details on each practice in the Measurement and Analysis process, consult
Appendix A “Task Description for Measurement Process”.


Establish a Measurement Approach and Plan

Before a measurement process can be started, the organization must allocate resources to establish and sustain it.  These resources should be documented in the measurement plan, and would typically include:
•    Managers and their responsibilities for oversight
•    Budget, hours, and personnel commitments
•    Program and project resources
•    Support personnel for IT, integration and deployment

Each time resources are assigned and commitments are made to measurement, the scope of measurement should be captured in terms of the organizational units (e.g. departments, programs, projects) and specific information needs or policy to be covered by the measurement process.

Once the set of organizational information needs have been established, it is time to create a measurement plan that contains supporting details such as the specific data to collect, collection procedures, analysis procedures and similar rules.  The measurement plan may be a single measurement approach to be used across the organization, or it may contain tailoring such that different business units, types of products projects, or development models have a unique measurement approach.

Again, you should minimize the reliance on printed or written documentation to the greatest extent possible. When your measurement plan exists solely in documentation, it becomes increasingly difficult to keep the documentation, process execution, automation and management behavior consistent. As you expand your measurement process in depth (by adding information needs) or breadth (by adding organizational units), you increase the burden of maintaining a document, and at the same time, modifying your process.


Collect and Analyze Data

Your measurement process must collect the data that is specified in your measurement plan.  Ideally, the data elements specified in the measurement plan would be collected as part of automated measurement collection. As obvious as this sounds, many
organizations don’t build their measurement process this way. Instead, they might collect
what is easy, or use what they already have, and then modify the measurement reports accordingly. The recommended approach is to collect the data you need and no more, using automated techniques when possible.

Once data is collected, it must be analyzed before being sent to managers.  Analysis includes calculating base measures from the collected data, combining base measure data to create derived measures and assessing status by comparing measures against plan, organizational thresholds, or historic data.  Analysis may result in a status or indicator value which represents the “goodness” or “badness” of the current data with respect to plans.  Analysis may also involve manual review of data by a subject matter expert. Analysis results are stored in a measurement repository for future review by managers and other stakeholders.

Measurement implementers should take a look at the state and technology of their IT departments before completing the measurement planning process.  The measurement process is usually the first time that an organization has attempted to build a consolidated picture of progress using information from disparate/disconnected sources. There are often difficulties in obtaining data, establishing the contents and then determining how best to use it.

As part of examining your IT infrastructure, you should uncover the sources of your measurement data and also establish the mechanism, or mechanisms, that can be used to extract needed data.  Such mechanisms could include web services, a programming language API, a report extract or a direct database connection. Many commercial and in-house tools are built in SQL databases —  having a measurement process which can support this saves a great deal of time. In addition, data may also be provided in common file formats such as an Access database file, a comma separated value file, or an XML file. Your measurement solution should be able to accept these file formats without requiring integration or scripting.

In designing measurement, you should recognize that your software processes and software project management needs will change over time.  To accommodate this, the mechanism for collecting data should be practically extensible. “Practically extensible”, means that you should implement a solution that provides universal integration with your data sources. You should avoid a solution which takes a point-to-point approach to integration. What you want to find is a solution that provides all required tools and technology for maintaining and extending the collection of data.

Without the automated collection of data, a measurement process will either not have the data needed to satisfy the stated information needs, or will rely on time-consuming manual methods.


Deliver Periodic Measurement Reports to Managers and Stakeholders

An essential element of measurement is the actual delivery of reports to managers who need them.  Once the information needs are established for each manager, the measurement process delivers information products (satisfying the information needs)
so that managers can monitor their progress and take action when needed.  The delivery
of measurement reports allows managers to have as complete a picture of their project as practical, and to replace subjective status assessments (“I think we are 50% done!”) with a quantitative and objective picture of performance.

Your measurement solution should deliver reports in a method that requires little effort, and impacts stakeholders and managers as little as practical.  When there is a burden in gathering data, updating status, or generating reports, managers will tend to stop using the solution. To be effective, the measurement process must efficiently deliver reports — ideally requiring less and less time on the part of managers as the measurement
process matures.

While the measurement process provides updates to these reports on a periodic basis, managers should be able to use these reports anytime. For example, managers might rely on measurement reports during team meetings, reviews, program analysis, re- planning discussions and a range of other reasons. One of the beneficial aspects of measurement is that organizational information is made available to those who could benefit from it. The measurement process makes the information stored in various organizational silos available to managers for decision-making.


Provide a Repository and Evidence of the Use of Measurement

Implementation of any process should lead to having evidence that the process is actually used wherever it is deployed across the organization. In performing a CMMI assessment as well as when determining internally whether your standards and policies are being followed, you need to have evidence that the process is being used. With respect to measurement, this means that there should be evidence that a measurement process is deployed, information needs have been defined, and that measurement reports are being used by stakeholders.


Provide Compliance and Assessment Reports in Parallel with Status

In managing a capability assessment initiative or process improvement, you should be careful to make sure your organization’s processes are the focus, not the assessment or improvement goal. For example, processes should only be modified to address CMMI if gaps are identified and the change is beneficial to the organization.  Then managers and engineers should be trained on how to follow the process and process improvement specialists should update the process documents. The fact that the organization is focused on a CMMI maturity level, an ISO standard or any other compliance or assessment is not useful to the rest of the organization.

With respect to measurement, this means that the measurement process may be updated by changing existing information needs, or the addition of new ones. You would not want to have a set of information needs for the “standard processes” plus another
set for “process improvement”.  The measurement process delivers an integrated set of information needs which allows managers to control the processes for which they are responsible. If some if these processes are improved, this could show up in the form of modified information needs.

In parallel with on-going management, it is reasonable for process improvement specialists to need to understand the status and progress of just the process improvement activities, separate from the manager’s integrated set. Your measurement process should be designed to allow you to re-use one or more of the process measures to create a parallel process improvement or capability measure.  Most desirably, the measurement process would allow you to organize these measures according to the framework or improvement initiative.

For example, if you were to attempt a staged CMMI for Software Level 2 initiative, then you might want a way to display your information needs in terms of the process areas required for Level 2.  A measurement process should allow you to map your information needs to the CMMI, policy, model, standard or other guidance.


Track Status of Process Improvement

There is a growing recognition that organizations should treat process improvement like a project —initiatives are committed, planned, performed and managed like any other type of product or program. Essentially, process improvement is a type of product to be managed. In this light, the measurement process, since it can measure any process, should be used to monitor the status and progress of process improvement.  Managers involved in process improvement should be provided with resource commitments, status and progress related to each effort. One key aspect of measuring progress is to have a small set (i.e. between 1 and 3) of indicators that compare current, improved performance, to a previous baseline.  By having such a running comparison, you can develop an understanding as to how effective your process change is, and take action if needed.

One of the key things you should determine after the improvement is how much better you are, or how much more capacity has resulted. The measurement process should quantify the following information needs for executives and process improvement leads:
•    Process Improvement Readiness
•    Process Improvement Capabilities
•    Process Improvement Progress

As an organization repeats the process improvement cycle, additional information needs could be added to focus on success factors and best practices that have been shown to lead to effective improvement.


Using DataDrill for CMMI Measurement and Analysis

DataDrill enables an organization to implement a measurement process by identifying and defining information needs and makes possible the delivery of that information to knowledge workers - senior executives to technical managers to developers and staff - in a way that supports decision-making and improves overall program management.

DataDrill assists organizations of any size in implementing a measurement program that manages systems, software and IT developments as well as supports process evaluations such as CMMI.  DataDrill is a scalable, flexible, easy-to-use solution that helps an organization:
•    Bring vital information together in digital dashboards for empowered decision- making
•    Satisfy and sustain process maturity requirements with accurate, objective compliance information
•    Facilitate defining and sharing of best practices for IT project management and measurement
•    Support management information needs through measurement process implementation

Specifically, DataDrill provides solutions to the primary challenges encountered when satisfying the Measurement and Analysis (M&A) requirements of the CMMI by:
1)  Supporting all activities identified in the measurement and analysis process area
2)  Providing pre-built best-practice information needs which jump start CMMI Level
2 implementation
3)  Automating all aspects of a measurement plan, from requirements, security, analysis, collection and reporting, eliminating time spent writing documentation
4)  Allowing tailoring for organization-specific measurement or project management practices

The remaining subsections describe how the DataDrill Express solution can simplify and speed a measurement process implementation.


Supports All Activities in the CMMI M&A Process Area

Express was specifically designed to replace the manual tasks associated with measurement. As listed in Table 1 below, there are eight activities and four primary work products identified in the CMMI M&A process area.  The table describes briefly how DataDrill Express provides each required element.

 

Table 1. Measurement and Analysis Process Elements

 

Table 1. Measurement and Analysis Process Elements, continued


Provide a Set of Pre-Built Best Practices

In DataDrill Express, the time between needing a best practice and applying it is short. By simply selecting from a Library of best practices, gathered from leading software organizations, managers can deploy proven measurements/metrics in their project. Extensive automation in DataDrill Express allows managers to decide for themselves which techniques to use, without needing process improvement experts or consultants.

 

 

The pre-built Library contains information needs and management practices that have been extracted from a variety of standards, policy and guidebooks, from leading organizations such as ISO, IEEE, SEI and others.

The information needs in the Library have been pre-assigned to CMMI level 2 process areas.  This assignment provides a starting point for establishing your organizations management and measurement processes. The assignment of information needs to process areas is contained in Appendix B.


Allow Tailoring for Organization Specific Management Practices

As an organization matures, DataDrill Express can capture best practices and share them across an organization quickly and effectively.  Management techniques that save time, money or resources become competitive advantages.  Methods for reducing requirements creep or raising product quality can be captured so that the organization can address other critical strategic needs, and not re-learn the lessons of past performance.

Using the built-in Library, your organization has a method to define, review, share and quickly deploy the management and measurement techniques, metrics and reports that lead to better performance and less risk.


Eliminate Time Consuming Documentation and Integration Tasks

In a measurement process without DataDrill Express, the measurement process is described in a management or measurement policy document, and then integration scripts and reports are developed based on the policy.  The problem with this approach is that the policy can easily become disconnected from the integration scripts and the reports. Another problem with this approach is that updating the measurement process is time-consuming, as all three elements - the policy document, integration scripts and reporting - must be updated manually.

Contrast that manual approach with the automated approach of DataDrill Express, where the policy document is a by-product of defining and selecting information needs within the Express application, and integration and reporting are automatically configured based on the selected information needs. The measurement policy is always up-to-date always accurate.

DataDrill Express provides automation for measurement policy, information needs, compliance, analysis, collection methods, reporting, users and groups, scheduling, and security and access policy. All key measurement elements are automated with Express such that the manual and labor intensive process of generating documents is streamlined.

 

Present Project Status Using Management by Exception

DataDrill Express presents managers with key color-coded status data, allowing them to quickly spot trouble areas in the projects and programs for which they are responsible. The measurement process is typically used for multiple projects or programs simultaneously, delivering a set of graphs to each project manager on a periodic basis. For example, your organization may have 50 projects being measured, each with 5 information needs and 25 graphs, for a total of 1250 graphs. To aid in finding and spotting problems quickly, DataDrill Express provides a number of management-by- exception features that allow managers to efficiently find trouble areas.

To simplify reviewing the performance of a single project, Express provides an array of status views, each with color-coded status linked to specific project performance. Managers can also enter their own business rules for color-coding.  Express automatically evaluates collected/analyzed data against all rules and targets.  With color- coded status and alert emails available for requirements, software size, quality, configuration management and scheduling, managers quickly spot troubled areas, assess and correct issues during the project lifecycle.


Support for High Maturity Organizations

High maturity organizations, those at Levels 3 or higher in the CMMI, rely on more advanced measurement process capabilities, as compared to organizations at Levels 1 and 2.  As an organization becomes more mature, their managers rely on a measurement process to provide information at a greater breadth (e.g. more process areas and work products) as well as greater depth (e.g. more detailed analysis of resources).

DataDrill has a long history of support for high maturity organizations, which enables DataDrill Express to support the unique and more demanding needs of these organizations. High maturity features in DataDrill Express include:
•    Tailoring for business, program, customer or other needs
•    Flexible analysis through extensive equations
•    User-configurable best practices
•    Stability and statistical process control
•    Analysis and reuse of historical data
•    Attribute tagging of projects by lifecycle, domain and others
•    Centralized administration and security controls

Your organization can start process improvement at Levels 1 and 2, and continue to rely on DataDrill Express as you raise your maturity to Level 5.


Data Repository

DataDrill Express captures and stores all data required for generating information needs and management reports.  Express provides a data mart which captures the raw data
needed to construct measurement values, as well as the historical plans and progress data contained in reports.

This data repository can be used to establish organizational capabilities by examining organizational performance in completed projects.  For example, by analyzing the historical data for requirements growth in completed projects, an estimator or project manager would have a very good indication of future performance.  In DataDrill Express, technology is built-in to allow this type of analysis to be performed quickly and easily so that past performance data (stored in the data repository) is accessible across the organization.

The data repository can also be used to compare the current performance with previous performance. One example is a measurement report which shows current progress along with a three month trend. Yet another example is a measurement report which compares current progress to the same period in a previous year (or quarter).

The data repository in DataDrill Express is based on a SQL database.  Because it uses common SQL technology, you may use a variety of third-party tools for reporting, specialized analysis, security, administration, backup/restore and other purpose.


Quickly Initiate a New Software Development Project

Software project startup can cripple an organization embarking on process improvement. In the first weeks or months of a new project, managers must get authorizations, establish measurement plans and policy, and setup a myriad of new tools and environments.  When time runs short, the measurement process is often left undone, leaving managers to deliver critical technology without the tools they need to monitor and control their project to success.

DataDrill Express saves managers’ valuable time during project startup, automating the time-consuming and repetitive tasks of establishing an effective metrics capability. Managers can select desired information needs from a library of best practices, and see those techniques in use within minutes.

The burden lightens for IT and process specialists as managers are less reliant on them to apply best practice measurement, metric and analysis techniques.  Express provides the technology needed to take the ideas of world-class measurement and implement them without the need for supporting personnel to become involved.  From integration to analysis and from graphing to management by exception, DataDrill Express quickly establishes the metrics that a project manager needs to start planning and controlling their project.


Track Compliance with Standards, Models and Policy Documents

DataDrill Express addresses compliance by integrating the measurement process with compliance tracking.  In Express managers describe a compliance policy, regulation or standard with which to comply, and then assess a project’s performance against it.  With Express, software projects are easily assessed against one or more corporate policies, standards or maturity models such as the SEI’s CMMI.  Express allows tailoring of key process areas, sub-processes and other sub-compliance aspects to suit a variety of compliance uses.

Using DataDrill Express, managers stay informed about the compliance of their projects by reviewing automatically generated alerts or periodically checking a set of compliance indicators available within each project.  Project managers review and verify their project compliance along with other project performance data -- never being the last to see project compliance.  Express makes compliance a benefit of effective project management, not a burden.


Measurement and Process Improvement Resources

This section presents resources for the measurement process, as well as process improvement.

Measurement Guidance

Software Engineering Institute     www.sei.cmu.edu
Distributive Management    www.distributive.com/resources_articles.html
Practical Software Measurement    www.psmsc.com
Crosstalk Magazine from STSC    www.stsc.hill.af.mil/crosstalk
Scott Ambler’s Web Site    www.ambysoft.com
Karl Weiger’s Web Site     www.processimpact.com


Summary

A measurement process provides valuable guidance for improving your organization’s capability to plan, monitor and control programs and projects. Software managers use the measurement and analysis process to enter and establish plans, to review the progress of actual performance against the plan, and finally to take IT project management action to identify and resolve issues of poor process or product performance. The measurement process is the mechanism that the organization uses to provide information that managers need to manage the way that the organization requires them to. The CMMI contains a Measurement and Analysis process area that outlines key elements of a successful measurement process.

DataDrill Express provides an off-the-shelf measurement solution that directly addresses the needs of the CMMI Measurement and Analysis process area. DataDrill Express supports the CMMI in these essential areas:
•    provides a substantial library of information needs representing industry best practices
•    automates the pratices in the CMMI Measurement & Analysis process area
•    enables tailoring and configuration of measurement
•    provides a repository and evidence of measurement use
•    supports tracking process improvement


Appendix A —  Measurement Task Descriptions

This appendix describes each of the tasks recommended by the CMMI for a measurement process.

Establish Measurement: The organization determines the resources, people, facilities and techniques to be allocated to the measurement process.  Organizational management oversight is identified, either a single executive or a team who is responsible for deploying, supervising and verifying measurement process implementation. The organization establishes the criteria for which business areas and business functions must use measurement as part of their management functions. This criteria includes identifying specific programs, projects and other business area, generally called “management units”.  The resources allocated to measurement and the requirements for using measurement across the organization are documented in a measurement plan.

Specify Measures: In this task, the specific measures to be used for each management unit (identified in the previous task) are identified.  Measures are documented using information needs, where each information need contains one or more measures. The complete set of information needs is captured in the measurement plan, which contains high level organizational objectives for measurement. Each information need contains guidance, reference information, measures (including plan data), stakeholders and reporting requirements which aid the organization in achieving its goals and objectives. Additionally, each information need will contain a description of how status, thresholds, targets and analysis are created.

Specify Data Collection:  With a measurement plan in place and information needs defined, the organization is able to plan the physical collection of measurement data. This step identifies the sources of data for each information need used within each managed unit.  Note that while the source of data may vary from managed unit to managed unit, the information need definition does not. Specific data includes the source of data, the method for collecting it, and the frequency of collection are documented in the measurement plan. Team members and other stakeholders who work with the data to be measured are trained or informed of procedures for ensuring that correct measurement data is obtained.

Specify Analysis Procedures: The organization establishes “rules” against which the measurement data will be evaluated for gaining greater insight and spotting trouble early.  Approaches include: statistical analysis, mathematical manipulation of the data (summing, averaging, etc.), alarms, comparison of actual against plan, and tracking policy and regulatory compliance.

Collect Data: Actual and plan data is collected on a periodic basis for the specified measures.  Data may be entered manually or collected through an automated process and is generally gathered monthly, weekly or even daily, depending on the needs of the organization.

Analyze Data:  The collected data is evaluated according to the analysis procedures specified previously, after which information users review the results to assess the impact of recent changes and also to determine if trends in data and status are significant enough to warrant further investigation. Digging deeper into the chain of data to follow problems to their root cause is a key to evaluating alternative action plans.

Store Data and Results: The collected data and resulting measures and analysis is stored in a secure, accessible place.  Herein lays the ability to share the data and results across the organization and with customers, and also to use historical data as a valuable contributor to lessons learned and to future planning and estimating efforts.  Stored data may also serve as documented proof of process and policy compliance.

Communicate Results: The results of having collected and analyzed measurement data (graphs, dashboards, alarm indicators, management notes, etc.) are shared within teams, across departments and up the corporate management chain so that each measurement stakeholder uses the information to improve decision-making and control performance toward organizational goals and objectives.  It is critical that the results are up-to-date, accessible and easy to interpret.


Appendix B —Information Needs for CMMI Level 2

This appendix contains a matrix which maps information needs from the Express Library to Level 2 process areas in the CMMI.  Since your actual management and measurement processes are based on your engineering processes, this matrix should be tailored to add and remove information needs, and to modify the mapping so that all processes can be managed.