lean-ISD – My Masters Series Article for ISPI’s Performance Improvement Journal (2001)

This post republishes my 2001 article for ISPI, from back when I was the majority owner of CADDI Inc.

I’ve updated the graphics to replace CADDI with EPPIC.


by Guy W. Wallace, CADDI partner


John Swinney asked me to be a part of his Masters Series at the 2001 ISPI conference to tell the history of CAD. CAD is Curriculum Architecture DesignSM, one of three levels of lean ISD design in what my firm calls the PACTSM Processes for T&D. I have been doing CADs since 1982 and have done 70 to date. I have presented on CAD at more than ten ISPI conferences and at seven local chapter meetings in the past 20 years.

The PACT Processes for T&D are a set of instructional systems design/development (ISD) methodologies authored, practiced, and evolved over the past 21 years by me and my business partners, Peter R. Hybert and Kelly R. Smith. They are performance oriented and practitioner honed.

If one could design a set of collaborative processes involving ISD customers and suppliers with

  • A set of prescribed phases and tasks
  • A robust set of tools and templates to quickly and efficiently engage the right people at the right time to produce high-quality, performance-improving training & development (T&D)
  • An object/chunking design and development strategy of both shareable and unique modules for deployment in any mode—group-paced or self-paced or individually coached

then you would have achieved what we also set out to accomplish with the PACT Processes.


The concept of lean comes from the M.I.T. study in 1990 that looked at the worldwide automotive industry and practices and compared them all to Japan’s lean production system, in the book titled The Machine That Changed the World.

The lean approach is most prevalently applied to engineering and manufacturing processes, but it is not limited to those processes. The goals in these lean applications are to

  • Use the best of mass and craft production methods.
  • Reduce costs and cycle times.
  • Improve product and process quality and customer satisfaction.


The application of lean to the world of ISD should create a set of common, effective, and efficient processes for the entire ISD process that spans project planning and management, analysis, design, development, pilot-test deployment, and evaluation of T&D.

These lean-ISD processes would allow for

  • Dividing the ISD project efforts across multiple T&D organizations, locations, and personnel while ensuring that all of the T&D pieces will fit together into a seamless learning experience for the target audience
  • Planning and managing predictable projects with predictable schedules and resource consumption (peoples’ time and out-of-pocket costs)
  • Developing both shareable and unique T&D Modules (T&D product subassemblies) that are components of a systems view of the entire T&D product line
  • Reusing (with little or no customization required) the T&D products and subassemblies for various target audiences from across the organization
  • Involving and collaborating with both upstream suppliers and downstream customers


ISD is instructional systems design or instructional systems development, depending on your source. It’s been making the headlines lately. According to the press, the death knoll has been sounded. ISD is too slow and too ineffective for life at Internet speed.

Is ISD dead? I don’t think so. Is ISD dying? No. Is ISD hurting? Definitely.

ISD has been slowly evolving since its early days, when it was created for the U.S. military in the 1940s. Then the military and all of the industries it directed put it to use for mission-critical purposes, helping to ensure that tanks rolled, ships sailed, planes flew, and supplies reached troops all the way up to the front lines. Situationally, it was do or die. No kidding! Today? Today is different.

Too many things are just not thought to be that mission-critical in the enterprise operating at Internet speed today. And too often the wrong things seem and are deemed important.

Today too many enterprise initiatives are using Knowledge Management Systems to cram all sorts of “e” content into electronic warehouses wired to each and every employee. But they are doing it without regard to the total life-cycle costs of creating, administering, and then maintaining the thousands of knowledge module objects. Not good.

Readers should consider those life-cycle costs so that they don’t end up as a warehouse full of past-its-time thinking and models. In other words, only build it if they’ll log on and you’ll maintain it, or at least delete it when obsolete.

Will the Internet kill ISD? No. It will better enable it. It will help it get lean. But it’s not the only component in what CADDI has labeled (and service marked for our newsletter and my book) lean-ISD. And by lean-ISD we mean performance-based lean-ISD.

The ultimate goal of ISD processes is to create instruction that is effective and efficient. Yes? Good ISD processes are also themselves effective and efficient. Yes? If you’re with me so far, then you’ll probably go for the notion that the ultimate goal of T&D is improved performance by the learners as measured by enterprise metrics. Metrics such as cycle times, deadlines met, costs, returns, lost opportunities, safety, and a myriad of other stakeholder satisfaction items are much more important than training days delivered alone.

These business metrics apply equally to both the suppliers and the customers of T&D. The customer uses these measures in those targeted operations, both pre- and post-T&D deployment. The supplier uses these same measures in their T&D development operations. But the supplier’s metrics only help us understand what it took to get the T&D there, available for deployment, and then to deploy it to its targets. That is the focus of lean-ISD: the supplier side. It’s about how to do performance-based T&D, and do it lean. But still the ultimate measure of T&D, lean or otherwise, is outside the T&D system box.

With T&D, it’s not really important that someone learned something, even to a very high proficiency, if what was learned had a miniscule impact on the enterprise given its costs.

All of the supplier and customer metrics lead to ultimate measures such as return on investment (ROI), economic value add (EVA), reduced future costs, and the potential returns on other reinvested savings. These are how T&D product quality are the best measures of T&D effectiveness and efficiency. Hey, a dollar saved is a dollar earned, a dollar generated with less than a dollar of expense expands the profitability of the enterprise, etc.

All T&D efforts incur a cost to produce a gain. Why spend $100,000 to gain back only $80,000? You wouldn’t if it were your money, so why would the corporation’s shareholders feel differently?


ISD is the label placed on efforts to plan, analyze, design, develop, implement, and evaluate T&D. There are various models in the literature that describe ISD, including the “big block diagram” (also known as “ADDIE”) from the work of Robert Gagne, Leslie Briggs, Robert Morgan, and Robert Branson.

There are many other ISD models—almost as many as there are ISD practitioners. Therein lies part of the problem for most organizations. In too many organizations, there are too many ISD models being followed. They are typically not common and not predictable in terms of the quality of the T&D outputs produced, or their costs and schedules, and they are not in control. They are often not visible to T&D management or to T&D customers.

This typically results in an ISD situation where

  • Content of the product line elements (courses, CBT, on-the-job programs, etc.) are both gapped and overlapping in terms of their content.
  • It is costly to produce in the first place, and more costly to maintain.
  • It is costly to deploy.
  • It is impossible to predict development schedules and costs and then predict ROI.
  • The look and feel of the T&D varies across the product line, and chunks of potentially shareable T&D aren’t designed with reuse in mind.

Many organizations have a significant opportunity in recovering and reducing resource expenditures for their ISD processes for producing T&D. They need to re-engineer their ISD processes.

The ultimate goal of the T&D is improved performance by the learners. That is how T&D product quality is best measured. The ISD process goals are to create this quality T&D in a reduced cycle time and at reduced costs.

The T&D products must have the desired effect in terms of the incurred learning in the learning environment (whether classroom, CBT, or on-the-job) and, most importantly, the ability to apply those learnings back on the job. The ISD processes must get this job done quickly and cost-effectively.


Over the past dozen years, the partners at CADDI have reduced to practice the prevailing ISD concepts, philosophies, methods, processes, and practices. Our efforts to model the ISD process are driven by the same need that has driven many businesses to first model and then re-engineer their core processes: to improve quality and reduce both cycle time and costs.

Many T&D organizations have undertaken efforts to re-engineer their ISD processes to make them common across the organization, predictable in their schedules and costs, and ensure that the T&D produced is effective. We began in the late 1980s and in 1989 coined the term “PACT Processes for T&D” which include

  • Analysis:      Performance Modeling and Knowledge/Skill Analysis
  • CAD:            Curriculum Architecture Design
  • MCD:           Modular Curriculum DevelopmentSM
  • IAD: Instructional Activity DevelopmentSM
  • PP&M:        Project Planning & Management

The three ISD processes of the PACT Processes are CAD, MCD, and IAD. Each operates at three distinct, different levels of ISD. Each is driven by the PACT Process analysis methodologies, including performance modeling and knowledge/skill analysis.


This methodology provides a structured, gated, in-control process for the fairly quick design of the overall curriculum architecture, or learning architecture. The design meeting may take two to four days and generate the macro-designs for all of the T&D Events in the curriculum architecture.

The CAD analysis process outputs of Performance Models and Knowledge/Skill Matrices are used to drive the CAD design and ensure it results in a performance-based orientation instead of a content/subject matter-based orientation.

Figure 1: CAD Project Phases

A CAD is built to support job performance. It creates an architecture of T&D Modules where shareable and unique modules of content are used to create performance-based T&D products such as courses; workshops; structured, on-the-job training; CBT programs; book reading assignments; project assignments; etc. These modules can be configured many ways, thus maximizing the shareability of content across various target audiences.

A CAD segments and organizes the content of training to ensure the greatest impact on an organization’s performance while minimizing life-cycle costs. It helps to prevent the allocation of resources to training that have little or no impact on job performance. Many T&D Modules and Events are never developed/acquired because there are no positive returns or economic value add. So why bother? You wouldn’t if it was your money.

A CAD builds a design for a training curriculum with individual parts that add up to a logical whole within the context of a given job or category of position. It ensures that all training works together to produce the desired results by providing employees with all the knowledge/skills needed to perform. A CAD’s modular design includes both shareable and unique modules creating the capability for infusing the enterprise with a more common language, viewpoint/perspective, culture, and the local unique needs. It contains generic content chunks and specific content chunks. Then the existing T&D can be assessed for fit, and gaps in the curriculum can be identified.

A CAD project engages the training customer in the prioritization of all training development efforts targeted to fill gaps in the overall architecture of T&D. All of the priority training content really required becomes visibly apparent to the training customer. The customer’s knowledge regarding the affect of training on specific areas of performance allows them to prioritize gap training development efforts that will help them meet their business needs. The collaboration creates many win-wins.

The CAD’s architectural design will help reduce the overall life-cycle costs of the entire T&D product line. Initial, “first costs” will be reduced by eliminating and minimizing redundant content development. “Life-cycle costs” will also be reduced because there will be no redundant content to maintain. This systematic approach to the modularization of training content will reduce maintenance and administrative costs.

The CAD’s macrolevel analysis and design outputs become guiding MCD inputs to the midlevel analysis and design efforts, and they are further leveraged in the IAD’s microlevel analysis and design activities.

Key Outputs of a Curriculum Architecture Design Project

CAD projects typically span a two- or three-month cycle, but small CAD projects can be conducted in as little as five days without formal documentation.

We have used this methodology to help a pharmaceutical company develop 90 days of technology transfer training for their proprietary processes after two days in analysis and two days in design.

Although the analysis and design data was very macrolevel, it guided the development efforts. The end results were very close to all initial estimates for both development time and deployment time. This project was also a hybrid effort of both the CAD and MCD processes. CADs almost always lead to multiple MCD projects, where there is a clear, key business priority. Many potential T&D products identified during the CAD process are never built because the ROI and EVA figures or the strategic value to the enterprise do not warrant the efforts and expenditures.

Just because T&D professionals are skilled at uncovering T&D requirements does not in and of itself warrant meeting those needs.

The PACT Processes can save the organization from low-value T&D and steer the resources to T&D with strategic, business-critical, high-payoff implications. And it can do it without overly complex ROI algorithms.

A number of methodologies are used throughout the CAD project; however, the most critical from both a quality and cycle time standpoint is the use of teams throughout all phases.

The use of appropriate company personnel on the designated project teams will ensure higher quality of both the project inputs and outputs. In addition, it will provide for a level of participation in the project activities that will create increased ownership of the results and more support for eventual implementation.

The project’s overall structure for key roles and the teams is as follows:

  • Project Steering Team
  • Customer-side project manager
  • Supplier-side project manager
  • Analysis Team
  • Design Team
  • ISD Team

CAD Phases

CAD Phase 1: Project Planning & Kick-off. In this phase, the project priorities, direction, and resources are defined. Potential issues and/or stakeholder requirements should be uncovered and planned for during this phase to ensure the success of remaining phases.

CAD Phase 2: Analysis. The purpose of this phase is to establish a common view of the positions, personnel, performance requirements, and knowledge and skill requirements. In addition, demographic information about the target population and information about existing training will be gathered. This common view will form the basis for the CAD and all priority-setting activities later in the project.

CAD Phase 3: Design. The purpose of this phase is to produce a CAD to address the performance tasks and knowledge/skills derived in the Analysis Phase. In this phase, tradeoffs may need to be made in order to maximize the return on investment for the overall corporation.

The intent is to create a CAD that is robust to future variation in job assignments; individual trainee experience, background, career goals; delivery facilities; and maintenance requirements. It also needs to be designed for content “updatability” and future adaptability to potential changes in the business (e.g., organization structure, competition, technology, etc.).

CAD Phase 4: Implementation Planning. In this phase, the priorities will be established by the Project Steering Team for all of the T&D Events (and T&D Modules) and will be translated into a CAD implementation “development/acquisition plan.” The plan could include deployment planning and other T&D systems and infrastructure requirements, depending on the situation within the T&D organization and/or the enterprise.

CAD Benefits

Quality, performance-based T&D exists exclusively to improve human performance, and that human performance exists within the context of business or organizational processes. Any other goal for T&D has almost zero ROI.

The CAD’s architectural design will help reduce the life-cycle costs of the entire T&D product line.

The T&D Modules can be configured many ways, but if they follow the “rules of modularity,” they will maximize the shareability of T&D content across various potential target audiences. They will create and/or reinforce common language across more target audiences, while also reducing the T&D suppliers’ costs by reusing content chunks over and over again (but only as appropriate!). Elsewhere in business, especially in design engineering, this is known as configuration control or platform design.


This methodology provides another structured, gated, in-control process for the fairly quick design, development, pilot testing, and revision/release of the T&D Modules and T&D Events of the CAD.

CADDI uses a proprietary process that is designed to incorporate representatives from stakeholder groups into the overall project’s activities and tasks.

An MCD project is conducted in six phases using a team process. Figure 4 shows an overview of the CADDI PACT Process for MCD.

Figure 4 MCD Phases

The MCD process allows the various concerns of management, job incumbents, and staff support groups to influence the design decisions. The project will be controlled by a Project Steering Team that will make the final decisions. Teams of top performers will be used to identify both the performance requirements and the associated knowledge/skills required. Additional teams will be used in the MCD process to ensure that all decisions reflect the needs/issues of the company.

The six-phase structure in Figure 4 provides the framework for the project activities, deliverables, and team structure.

MCD projects typically span a four- or six-month cycle, but small MCD projects can be conducted in much less time. We built and pilot tested, in a 90-day cycle, a four-day “labor relations” course with more than 50 percent of class time spent in intense simulation exercises/role-plays. Three ISDers were involved.

The CAD outputs of the performance modeling and knowledge/skill analysis process and the CAD design specifications are used within the MCD process to drive the design to ensure it results in shareable T&D Modules and Events.

Key outputs from CADDI’s PACT Process for MCD are shown in Figure 5.

Figure 5 Key MCD Outputs

The project’s overall structure for key roles and the teams is as follows:

  • Project Steering Team
  • Customer-side and supplier-side project manager
  • Development Team
  • Pilot-test participants
  • Pilot-test instructors/administrators
  • ISD/T&D Team

MCD Phases

Phase 1: Project Planning & Kick-off. Project priorities, direction, and resources are defined; potential issues and/or stakeholder requirements should be uncovered and planned for during this phase to ensure the success of remaining phases.

MCD Phase 2: Analysis. A common view of the personnel, performance requirements, knowledge and skill requirements, and appropriateness and completeness of any existing T&D is established; this common view will form the basis for the training design.

MCD Phase 3: Design. In this phase, the Design Team is facilitated through a systematic design process; some details are completed after the design meeting.

Note: The intent of the team approach to design is not “to design by committee” but to influence “the designers by committee” during the actual design activities.

MCD Phase 4: Development/Acquisition. In this phase, the training is developed and/or acquired/modified per the Design Document (produced in Phase 3).

MCD Phase 5: Pilot Test. In this phase, the training is delivered (pilot tested), and extensive evaluations are conducted.

MCD Phase 6: Revision & Release. In this phase, all materials are updated (per the “revision specifications” from Phase 5) and are released into the training system.

MCD Summary

The PACT Process for MCD is a powerful process, if populated with the right people to do the right things at the right times. The gates ensure that our customers and key stakeholders for our T&D product line are systematically engaged for our collective success—collaborative win-win.

MCD uses the multiteam approach to plan and conduct a predictable project to develop and test performance-based T&D. Whether preceded by a CAD or not, MCD takes a proactive approach, with tools and templates to accelerate and ensure the quality of both the analysis and design efforts.

The MCD methodology engages the right stakeholders at the right time for getting the right inputs and right decisions at the right time. It shortens the project time cycle and reduces costs for T&D projects. It increases the quality of the T&D product/service by focusing on desired performance as the terminal learning objectives. It structures T&D content into more shareable chunks, thereby reducing future costs.

The MCD methodology provides a gated process for working with all project participants in an accelerated manner to produce performance-based T&D. The T&D professionals retain control of T&D decisions, and the stakeholders in your marketplace gain and exercise control of all the business decisions inherent in T&D projects and resourcing.


This methodology follows the same six-phase approach as an MCD effort (refer to Figure 4). It provides another structured, gated, in-control process for the quick design and development of various types of instructional components, including

  • Knowledge Tests
  • Performance Tests
  • Simulation Exercises
  • Performance/Job Aids
  • Electronic (or Paper) Desk Procedures
  • Instructional Content: awareness or knowledge level

This methodology produces outputs that could be part of a T&D solution, or could exist on their own. Our intent is to be able to build T&D components and wrap other T&D components around them later, as needs dictate/allow.


Early in my career, I was introduced to the work of Rummler, Gilbert, Mager, and then Harless through ISPI (then NSPI). One of the first books I read on the subject was Bob Mager’s book Analyzing Performance Problems, and it really made sense!

I started climbing the learning curve of ISD and HPT and began applying a Rummler-like performance analysis approach as a front-end to the design and development of self-paced (paper-based and video-based) T&D for contractor sales, and then inside sales, and then inventory management.

In 1981 I started working at Motorola’s Training & Education Center (MTEC), the predecessor organization to Motorola University. Our leader, Bill Wiggenhorn, had us attempt a project to create a Geary Rummler Design Process. That process would help us produce performance-improving T&D as well as help the organization deal with those issues in the work processes that would not be affected by our T&D. The project became entangled in the 13 perspectives on “how I like to do it,” and it didn’t happen while I was there. Later, Jeff Oberlin, the design manager, did develop a methodology, but the Rummler-esque version I had craved didn’t materialize during my time there. Since then, it has always been my goal to do just that.

Our effort at MTEC failed. Too many cooks and not enough decisions imposed, and the team drifted and fought. But it was a great idea. My time at MTEC was very good for me, especially working for Bill Wiggenhorn. His vision inspired all who worked with him.

When I heard Deming say, “80 percent of all quality problems are in the control of management,” I smiled. We at ISPI knew that for a fact. We know it is not always the performer. We know that it’s not always the performer’s need for training that is the root of any performance issue. I’ve heard Rummler say it many times, “Put a good performer in a bad environment, and the environment wins every time.”

In 1982 I joined Ray Svenson at R.A. Svenson & Associates, and I began to look at T&D as more than a partial family of curricula for a job or job family. I began to look at it more holistically. We studied jobs and created end-to-end paths of curricula to create performance competence. It was here that I began formalizing what became PACT.

My first Curriculum Architecture Design project was done here, in 1982. It was a combination of the approach Ray Svenson had for curriculum architecture from The Bell System for Technical Education (BSTE of AT&T) and a Rummler-esque Performance Model and Knowledge/Skill Matrix to drive the performance orientation into the design of the modules of content.

During my 15 years at what evolved into SWI (Svenson & Wallace, Inc.) and then into CADDI (Curriculum Architecture Design & Development, Institute, Inc.), I evolved the CAD methodology into a four-phase process, with tools, templates, and standards. Next I added a T&D Module development methodology, which is now MCD. MCD picks up where the system engineering of the CAD approach leaves off. The high-priority “gaps” of the CAD are addressed in subsequent MCD projects using an ADDIE-like six phases. Again, task procedures, tools, and templates guide and streamline the process without shortcutting the real meat of ISD.

My most recent CAD project, number 70, was done here at CADDI, in 2000. In early 2000 I published my book: lean-ISD. I was happy with the reactions by people I really respected.

lean-ISD embodies a collection of many best practices from many business disciplines. The current, leading ISD concepts, models, methods, and tools included are intended to create an engineering practice of ISD.

Why? Because after all my years in the T&D business, my view of ISD was still a muddled mess. All the books and articles I read, and the workshops and presentations that I attended at ISPI, ASTD, Lakewood, and others, did not add clarity. They did not prescribe, which is exactly what the neophyte thirsts for. At least I did. Maybe today it is different. I think not.


The ultimate goal of the T&D is improved performance by the learners. That is how T&D product quality is best measured. The ISD process goals are to create this quality T&D in a reduced cycle time and at reduced costs.

We believe in performance-based T&D, but only when the ROI and EVA are sufficient to warrant the investment. Otherwise, don’t bother. Our team and management approach facilitates the “for the sake of the business” only, no-nonsense nature of PACT projects.

Over the past dozen years, CADDI has attempted to reduce to practice many of the prevailing ISD concepts, philosophies, methods, processes, and practices. Our efforts have resulted in what we call the PACT Processes for T&D. Much of our ISD insight came from our affiliation with ISPI. For that we are very thankful. We hope you are thankful for our attempts to share over the years as well. Thank you for your interest.

This article was intended to overview what the PACT Processes are as an example of what lean-ISD is. This article is also intended to thank all those responsible for our opportunities to learn—and provide the history John requested.

lean-ISD, PACT Processes, Curriculum Architecture Design, Modular Curriculum Development, and Instructional Activity Development are service marks of CADDI, Inc.

*****   *****   *****

For a PDF of the article please go – here.

Guy W. Wallace

Shut down CADDI in 2002 to go solo with



The PACT Process are Guy’s intellectual property, acquired from
Svenson & Wallace when that firm dissolved in 1997.


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.