Thursday, January 26, 2012

Preemptive Quality Assurance

The quality revolution that swept through American industry in the last 30 years has been driven by one overriding principle—it's better to prevent mistakes than to correct them. The traditional method of quality control was inspecting products at the end of the assembly line. This was not only inadequate to ensure high quality, but it incurred a high cost in fixing or discarding products that failed to pass.

Unfortunately, many A/E firms haven't evolved much beyond the old inspection method. The common quality control process (and that's a generous term for what many firms have) still relies heavily on reviews near the end of the project. Even if the reviews are well done, which sometimes isn't the case, there's added cost involved in correcting the errors and omissions that are found.

Too Much Rework


How much cost? Some estimate that rework in our industry averages 10-20% of project budgets. Most companies in other industries couldn't survive such a high rate of rework, but we tend to accept it as normative. Granted, much rework is beyond our control, imposed by clients, regulators, or changed conditions. But there's still plenty A/E firms can do to substantially lower the cost, and improve quality.

The important thing is to step up your quality assurance measures in the early stages of the project. Prevention is key. You may be familiar with the 1-10-100 Rule, a general rule of thumb often cited by quality management professionals. The theorem goes that for every dollar spent preventing a mistake it will cost you $10 to catch and correct it internally. But the cost escalates to $100 once you experience an external failure.

The numbers may not be precise, but the 1-10-100 Rule illustrates a basic truth: The cost of an error increases exponentially the later in the project delivery process you are when it is discovered. Many of us can share stories of multimillion-dollar construction failures that resulted from preventable design errors. In fact, one of my former employers was pushed into bankruptcy by such an error—discovered several years after the design!

Key Causes of Unnecessary Rework

It's helpful to consider what leads to avoidable rework. As mentioned earlier, there will always be things beyond your control, but most rework can be averted by minimizing the following:

  • Failure to clearly understand the client's expectations
  • Not carefully planning the details of project execution
  • Failure to seek client and stakeholder endorsement of your project plan
  • Poor communication between the client, project team, and other stakeholders
  • Inadequately defining client deliverables at the start of the project
  • Failure to engage senior reviewers prior to the development of work products
  • Inadequate intermediate work product reviews
  • Poorly defined standard work processes
  • Failure to clearly define review roles
  • Poor communication with the client and contractor after the design is complete
A Basic Preemptive QA Process

The following briefly outlines some key steps in building a quality assurance process that seeks to prevent mistakes rather than merely catch them in the final review. This process establishes expectations and secures buy-in early. It promotes the type of planning and foresight that often gets shortcutted in our task-oriented profession.

Clarify client goals and expectations. Make sure you understand not only what the project is, but why it is happening and what it's supposed to accomplish—including how it will meet strategic and personal needs. Formally benchmark quality and service expectations with the client before proceeding.

Carefully plan project execution. A project management plan (PMP) is advised for all projects requiring substantial time and resources to complete. You should engage other project stakeholders in defining the best approach to meet client needs, deliver great service, and maximize efficiency and deliverable quality. Involve the client as much as possible up front.

Define deliverable content in detail. For design projects, this typically involves a design basis memorandum, concept design, or other similar preliminary work product. For reports and studies, a detailed report outline should be required that summarizes not only the structure of the document but its core content—observations, conclusions, recommendations. I call any of these products the "deliverable profile."

Gain endorsement of the PMP and deliverable profile. Secure the buy-in of the client, project team, and other key stakeholders (e.g., regulatory agencies). Endorsement involves not only getting approval but mutual commitment. Each party formally commits to do his or her part. It's important to get internal senior reviewers engaged at this point.

Actively collaborate with the project team. Superior quality and service is best achieved by project teams that work closely together. Conversely, "lone ranger" PMs who have inadequate involvement with their project teams are at the root of many project problems. Take steps to foster ongoing collaboration and communication, even measuring it to the extent you can.

Closely monitor and manage project progress. The PM is responsible for keeping the project on track in terms of meeting scope, schedule, budget, and client expectations. This must be done regularly so that potential problems can be identified and mitigated before they become serious. Develop a plan for managing project change at the outset.

Religiously stick to internal milestones. One of the most common quality and schedule problems is failing to adhere to internal deadlines. This results in task completion being constantly pushed toward the client deadline, meaning that the effort is rushed and inadequately reviewed as the deliverable heads out the door—a sure recipe for mistakes. The PM should strongly reinforce meeting internal milestones.

Clearly define internal review roles. Reviewers should have distinct review roles, avoiding unnecessary duplication or, worse still, gaps in the review process. Some degree of overlap is desirable, but reviews tend to be more thorough when reviewers are focused on specific things. Checklists are helpful in guiding the different reviews.

Diligently check all work products. Final reviews are still needed, even with a proactive approach to quality assurance. But hopefully you will be finding fewer errors and omissions because fewer are being made. All client deliverables (including some correspondence and emails) should be checked by a third party before they leave the office. More complex deliverables warrant a multiparty review.

Apart from these reviews, it's important to ask every project contributor to check his or her own work before it is passed on. Besides adding another level of review, you want to promote a culture of client focus, individual initiative, and ownership of the work product—even in its intermediate stages. Getting everyone to share this responsibility is one of the best ways to preempt quality problems down the line.

No comments: