COSYSMO Adoption Process

Chris Miller
Systems & Software Consortium, Inc.
Herndon, VA / Ricardo Valerdi
Massachusetts Institute of Technology
Cambridge, MA

ABSTRACT

As the COSYSMO model transitions from the development phase into the adoption phase, industry stakeholders are beginning to embrace the model and integrate it into their existing measurement processes. To date, much of the guidance provided by the COSYSMO development team has been focused on the usability of the model. In the adoption phase, users need guidance on the adoption process that can help them convince management to invest resources, scope the effort needed for successful adoption, and implement the adoption of the model.

This paper provides such a process which provides guidance on the piloting and institutionalization of the model described in eleven steps. These recommended steps have been developed as a result of interactions with seven organizations that have participated in the industry calibration of the model and have subsequently begun to integrate the model into their internal processes. The diversity in the approach that each organization has taken has inspired different aspects of the process which we plan to refine further as our experience evolves.

Overview of the framework

The ten steps in the framework are provided in the figure below. Certain opportunities exist throughout for verification & validation of the model and its scope relative to the organization’s current measurement system and needs.

The Process

It is recommended that organizations follow these eleven steps in the order provided to maximize the probability of success in the adoption of COSYSMO.

  1. Call for participation

The first step involves the solicitation of support and participation in the adoption of COSYSMO. Most large organizations often have people working on similar issues at different sites. Connecting with these individuals can help strengthen the case for investing time in COSYSMO and associated activities.

  1. Check relevance/informal mapping

In order to determine the relevance of the model it is recommended that each organization do an initial feasibility assessment of adopting COSYSMO in their organization. Much of this can be done by reviewing briefing materials and interacting with other users in similar organizations.

  1. Train champion

If relevance and applicability have been confirmed, an individual with the appropriate level of influence and visibility should be assigned as the COSYSMO champion. This individual should agree to take on this initiative until it the end, act as the internal advocate, and attend COSYSMO Working Group meetings. From this point forward, the champion will be the central conduit of questions, data, and experience for an organization. The champion will receive training from an experienced COSYSMO user on issues related to model inputs, data collection, and calibration.

  1. Understand inputs and identify pilot programs

The champion must clearly understand the data requirements for the model inputs and outputs. Most of this information is available through documentation but its use is best learned through tacit knowledge provided by experienced users. The identification of pilot programs is also important since they represent they way the organization’s business practices and organizational roles.

  1. Informal mapping at the WBS level

A review of the internal company’s WBS and time collection structure will inform data collection needs as well as any possible customization of the model scope as initially defined.

  1. Test run industry calibrated model

Using data from completed projects, the industry calibrated model should be tested to determine whether the COSYSMO model provides an estimate that is close to the actual systems engineering effort used.

  1. Tailor COSYSMO to organization

If any customization is needed, the model definitions, scope, and activities should be adjusted to represent the organizations way of doing business.

  1. Historical data collection

The tailored model should guide the data collection instrument used to gather historical information on completed programs. This activity involves the individuals that have access to the quantitative aspects of the project (i.e., size drivers) as well as the stakeholders involved in the project that can assess the qualitative aspects of the project (i.e., cost drivers).

  1. Training for users

As users are exposed to the model inputs, they should receive training on how to use the model to estimate systems engineering effort. They should also be exposed to the main assumptions of the model and the local tailoring that has been performed as a result of the organization’s dominant processes.

  1. Local calibration

Once a critical mass (i.e., at least 5 data points) of historical data has been obtained it can be used to derive a new calibration constant for the model. This is an essential step for improving accuracy and relevance of the model.

  1. Large-scale rollout to other projects

Armed with a locally calibrated model, the COSYSMO champion can field a large scale adoption of the model that reaches across a larger breadth of the organization.