Choose your language:

Hong Kong
New Zealand
United Kingdom
United States
mountaineer arms raised at peak

Adventures in managing large IT projects

February 27, 2014

Much like epic adventures, major IT projects offer great rewards but are fraught with peril. Technological transformations pose risks not just for organizations, but also the people running them, i.e., you. (Just ask the folks behind In fact, a McKinsey study found that 17 percent of major IT initiatives go so poorly that they jeopardize the existence of their business.

In the interest of full disclosure, I’ve never run a big IT initiative (or slayed a dragon). But I’ve had many conversations with some of the best IT project leaders in the business—the IT superheroes who lead major transformations for the biggest companies—about what works. Several common themes emerged; hopefully they will help you conquer your next epic IT undertaking.

Build it right the first time. Clean is better. Thoroughly understand any new software you’re taking on, and try your damndest to avoid workarounds. Using out-of-the-box technologies will help you keep up with software changes and avoid the rabbit hole of custom patches. If you need to, hire the best consultants—the ones who understand your software and hardware better than their makers do—and you’ll save money and headaches.

Document everything. Don’t waste time reinventing the wheel for common, repeatable tasks. For instance, you probably have a number of administrative hurdles for new hires to clear before joining your team. Onboarding may not seem like a big deal until you truly consider the time it takes someone to acquire system access and equipment, go through training and jump through HR hoops. Gather those requirements once and write them down—it will help immensely in getting new people up to speed quickly. And that can really pay off when a key player leaves during your project or you need supplemental help. The same goes for software and hardware architecture maps and uniform coding standards.

Track and understand your process. When planning projects, you need a realistic, data-based assessment of how long tasks will take. While that information might not be available before your project begins, you should begin collecting it as soon possible. Use that data to update your timelines and plan for personnel capacity.

Invest in automation. Automate anything you can. For instance, larger organizations should set up role-based system access for all technologies to prevent playing musical passwords. Unlike many back-end necessities, people in your organization will notice and appreciate this investment. Likewise, automating QA functions as much as possible will benefit your project because you can test everything (instead of a sample), and they’ll save your team a ton of time.

Choose intuition over training. Don’t lose sight of the cost of bringing end users up to speed. If you come across an area where you have to decide on spending more for an intuitive interface or training users to learn new skills—choose the interface. Many technological innovations fail not because they don’t work well, but because the people they’re supposed to help don’t understand them. Also, training gets expensive, especially when you consider the cost of bringing new users on board down the road.

Hire great people. At TEKsystems, we believe strongly in adhering to developing and adhering to the best IT project management processes. But we also know that a project—no matter how well-developed the process supporting it—can still fail without the right people. You need a team of people with strong attention to detail, who care how clean their code is, who stay up late studying data architecture for fun, and who check and double-check their work. And find workers with flexible intelligence, because problems will inevitably arise and you need people who can think fast and creatively to handle them.

Blog Archive