Quality and quickness
The emphasis on speedy design and delivery that characterizes rapid development projects can work well with traditional project management skills and produce quality deliverables. Quickness is effective when supported by constant customer contact and communication so that new and evolving requirements are understood by all and integrated smoothly.
Some suggest that a rapid approach leaves a project or product more susceptible to deficiencies in quality. “Often, usability suffers, performance suffers, maintainability suffers and testing is not adequately done,” according to the article “Rapid Application Development: Project Management Issues to Consider,” which appeared in the American Society for Quality's Software Quality Professional journal.
While utilizing rapid methods is most pertinent in software development, the demand to produce quickly or within a short timeframe is relevant to any industry—and the same risks apply. To manage and counteract these risks, project management techniques must be part and parcel of a rapid development approach. “To perform rapid application development, organizations need to employ project management fundamentals, which include the tasks of estimation, planning, tracking and control (measurement practices and risk management practices) and people management,” the article states. Four executives share the best practices they're using to ensure both quality and speed.
Vice President, Federal Sales,
CDW-G, Herndon, Va., USA
CDW-G works with IT companies to deliver more than 100,000 different computing solutions and products. For a recent U.S. Air Force deployment, CDW-G created more than 35,000 systems—some with as many as 16 different custom software images and hardware configurations—that had to be shipped to 30 locations around the world. The day after our contract was awarded, we were delivering projects, and while this was a single contract, we had to adapt our system to execute almost 1,200 different priority-handling orders.
Competency in program management helped us go above and beyond. In our PMO, we make sales personnel an integral part of the project team to ensure what we communicate and commit to at the time of sale is what we meet or exceed at the time of delivery. By formalizing that process of sales and program management working hand-in-hand with the customer, we're able to handle a lot of large-scale deployments.
Because the scale and magnitude of this deployment were unprecedented, we set up auxiliary supports to make sure we were on track. We held weekly status meetings with both organizations’ program management teams while we were getting configurations. The meetings often included a broader set of constituents, such as representatives from the Air Force Commodity Council, CDW-G contacts in sales management and logistics, and representatives from our various manufacturing partners. Essentially, we had a mini supply-chain integration, from end-user to acquisition, sales, logistics and manufacturing.
These weekly calls were followed up with hard-number reports that served as the basis of our next meeting, which helped us deal with problems—both real and perceived—much more quickly. From here, we formalized metrics based on the weekly status and worked closely to coordinate our efforts with all the applicable manufacturers: The Air Force wanted the best technology, and we were constantly working to put together the best of what the industry had to offer.
In rapid deployments, you have to recognize that you're always dealing with a new set of circumstances. It's also important to have estimated requirements, volumes and timeframes provided to you by your stakeholders. In our case, guidance regarding the required volume actually was quite lower than demand turned out to be, so when the upfront planning didn't quite hit the mark, we had to immediately assess the downstream implications. When the plan of record is altered by outside events, you have to think about the long-term effects immediately and figure out what you need to do to change.
MICHAEL NORMAN, Ph.D.
CEO, Scapatech Technologies Ltd.,
Edinburgh, Scotland, U.K.
Our company does a lot of work with the open-source organization Eclipse where we use a continuous development model that puts a fixed iteration structure of six to seven weeks around a cycle. In this time, you're designing, building and testing, and each week overlaps the next phase. There are two iterations in a quarter and at the end of every quarter you get a new point release—the idea is that if things aren't going well, you get that visibility within three weeks and it forces you to take this situation into account. Seeing things that may go wrong very early is the key to reducing risk in a rapid process.
Eclipse has a very interesting property because individuals and companies are collaborating to build something on a volunteer basis. In trying to manage the work, you're attempting to control resources you don't really own, so the sanctions you have as a project manager are just not there. But we have structures in place that sort this out. We elect committers—the only individuals allowed to change items in the computer code— who are responsible for making sure an entire project component is completed. Committers determine system architecture, control what gets dropped into code base and vote on ultimate plan approval. This is how you get real quality from open-source projects.
If you're trying to control everything through a single project manager, that's a gating process. The bottom-up process works very well because there's an aspirational/motivational aspect that promotes good working levels on a tech-to-tech level, largely eliminating the need for top-down monitoring.
Customers respond well when they have a lot of visibility, get a firm release date and see a credible, detailed roadmap. Communication skills are key: You've got to resolve things quickly. It's also key that everything we do is public—you need to have a public audit trail of every meeting, every set of minutes, every phone call. You've got to educate your staff on how to work in public, because normally, these people would not have that level of exposure. From a company perspective, the committers have a distributed sense of responsibility that helps propel the project. I can't quantify, but I know these methods have decreased the time it takes us to add more features.
Chief Operating Officer, US Technology
Resources LLC, Aliso Viejo, Calif., USA
US Technology utilizes a hybrid development model where we have a strong contingent of United States-based project managers, architects, designers and analysts who interface with our customers, and then we have a world-class team in India that builds source code and does our testing.
Our business model is predicated on very strong project management skills to ensure things get done on time. Our company has a 98 percent success rate on fixed-price projects because we recognize the balance between rigor and time-to-market. We have crystallized best practices to react to customers’ needs rapidly. First, you must have good, solid requirements upfront; if you don't, you surely will fail at the end. Second, we utilize round-the-clock development and make sure our work is being executed in parallel streams, taking advantage of the 24-hour cycles between us and India. The customer work and refinement is done during the day and our India-based facilities do coding and testing during the night.
Third, we maintain a very expansive knowledge management database with lessons learned from all of our projects. When we start a software project, we're not starting from ground zero: We are looking at risks, strategies, technological issues, organizational tact from previous projects. We do a tremendous amount of reuse, so every time we do a project of a specific nature (upgrade, conversion, etc.), that discipline is incorporated into the current project. We apply a reuse process to every project and measure the amount of reuse in each project, and track this infrastructure on a weekly and monthly basis. Tracking reuse is a part of our natural project review, and we require all project managers to take a look at what else has been done before they begin a project. We also have a reuse committee that takes artifacts from these projects to produce case studies.
We improve our speed by educating the customer on process and ensuring our customer's staff is aware of our best practices. We're very cognizant of having the customer tied in to what we're doing. Our customers also have complete visibility into project at all times, and they can say, “Hey, you guys might be off,” rather than waiting a year when it becomes more costly and the stakes are higher. I haven't heard a lot of people talking about full transparency but that's our culture. We let our customers see everything: They can have any information at any time.
President, Engineering & Construction Division,
and Director of the Board, Larsen
& Toubro Ltd., Mumbai, India
The five divisions of our company represent $3 billion in business, and 85 percent of this comes from engineering and construction sector. As a company, 80 to 85 percent of our projects are fast-tracked based on the lump-sum, turnkey (LSTK) scheme where things can move faster in execution. To help facilitate this, we look to our growing lessons-learned database. We are trying to promote more widely utilized guidelines and standard operating procedures, and tying this into an online knowledge movement. We also are always looking for ways to continue to integrate our ERP system with our own firm-based tracking systems.
One of our major goals is to achieve 20 to 25 percent volume growth each year, and we are trying to secure high-quality talent. A lot of our India-based talent is being sent to or used by other places around the world, especially the Middle East and United States, so we believe very strongly that a system built around solid project management will help us stay competitive. In this way, doing things with speed is a true necessity: Completing our projects faster is one of the ways we work around this talent deficiency.
Fast-tracking our projects has given us real results. We've increased the number of projects we prequalify in the Middle East and Africa, and this is all based on our three elements of good project management. First, you must start with good customers who are decisive with their approvals. Next, you always must utilize the most qualified project managers and engineers, because they drive projects. Last, you must have the systems to manage resources throughout your supply chain and always remain flexible. I always say a very good project manager and good project management can save a mediocre project, but bad project management will surely fail the best project. We still are growing and learning in our company—we're nowhere near where other companies might be or where we want to be—but we know that our best practices help us do our LSTK projects more quickly. We also recognize the importance of project management in strengthening the whole infrastructure of India's workplace. We're glad we've had the opportunity to spread this movement.
PM NETWORK | APRIL 2005 | WWW.PMI.ORG