11 ways to stabilize your infrastructure

What do you do if you’re struggling with infrastructure and stability? Maybe you’re a startup with no time for infrastructure. Maybe you’re a business that is reluctant to invest, with an IT gap that is hard to close. The stress, the worry and the escalation calls can keep you up at night – and then beat you down the next day.

To regain your balance, first take a deep breath. Then consider these 11 steps to help you better align the people, processes and technology around your IT infrastructure:

1. Reduce IT complexity. In a recent survey of 800 CIOs, more than three out of four say increased IT complexity could make it impossible to manage digital performance. Indeed, a single web transaction may now cross an average of 35 different technology systems, up from 22 five years ago. Is that sustainable? If you’re unable to handle your current level of complexity, review your operations, peel back the layers and make sure you’re building from a solid foundation.
2. Delegate. Assign members of your team to key activities. Make sure they are enabled and understand they have authority to solve problems. It sounds simple, but many entry-level IT professionals worry that they can only follow the SOP and are not encouraged to be creative. If you hire smart people, give them room to maneuver. Creative problem solving will make you and your team more effective.
3. Simplify and streamline. Where does your time go? Do you even know? With too much on your agenda, you could end up spending a large part of your day simply switching from one complex task to another. Can an engineer really be effective working on eight projects? (Does that mean a different project every hour of the workday?) If you’re a manager and leader, try to reduce that number and the time your team spends between tasks.
4. Prioritize. Focus your resources to solve the problems that give you the most ROI. First, score projects on key criteria that are important to your overall roadmap goals. Then pick the top projects and map them against your resources. Let the numbers talk, but be flexible: politics or customer status could change the ranking. If you lack the capacity and can’t hire new personnel, don’t be afraid to put something on backlog. If necessary, renegotiate deadlines and always set proper expectations.
5. Trust. Lack of trust shows up as micromanaging and turns into delay. On the other hand, trusting those you hire to solve complex problems can be like putting wind in their sails. Sometimes you just have too much to do, and trust won’t solve that. But if you doubt and second-guess your team all the time, you’re slowing things down. Find ways to build trust and see your capabilities expand.
6. Tolerate imperfection. Don’t let perfection be the enemy of progress. Not every bug needs to be fixed before a release. And solutions for bugs that are showstoppers need not be elegant. Don’t do sloppy work, but waiting for something to be perfect will cause delays and frustration every time. Hold back on engineering for an A++ and settle for the A.
7. Share knowledge. Make sure that your team has a way to share knowledge. From Jira to Slack, to Microsoft Team, there are plenty of secure, collaboration tools. What a shame to work on solving a problem only to lose out because the solution wasn’t documented and shared. Documenting upfront takes extra time, but saves in the long run.
8. Have that tough talk. Delaying difficult conversations is like carrying around baggage that becomes heavier each day. More importantly, in a large project, not talking about those things that nobody wants to hear about will just mean that eventually you’ll be asked, “Why are you telling me this now?” As my boss says, “Bad news isn’t like a fine wine. It doesn’t get better with time.” If you know something isn’t going to work, outline what you have done and your concerns. Overcome your fear, talk things over, adjust and move forward.
9. Ask for help. We know you’re competent. But there comes a time when you still need to ask your management and other teams for help. Don’t worry, this all comes around. You’ll reciprocate at some point. Vendors can also help. Of course, they will charge you, but sometimes contracting with an external expert to solve a problem is the most effective way to take something off your plate.
10. Anticipate. Get into the mode of anticipating problems. Think through the solutions and, where possible, have a plan of action in place before the crisis occurs. This is key, because frameworks, training, knowledge, great technology – these can all fail. Indeed, according to one classic theory, complex and tightly coupled systems are more prone to accidents. If you haven’t prepared, accidents and failure are more likely to cascade into catastrophe.
11. Automate. With current-generation tools, there is no excuse for not automating a significant amount of your work. Do you still have an army of people handling alerts? Or skilled labor rebooting machines for known issues? If so, think again. Many companies (including the one I work for, NTT) have sophisticated platforms that allow companies to automate their operations. The amount of money and time saved with these platforms that you’ll see when the ROI analysis comes back is amazing.

Gain perspective and time

Each of these steps has value of its own. But linked together, they add up to something more. Preparing for a crisis or disaster (step 10), for example, may get you thinking about the complexity of a key system (step 1), which may lead you to revisit the documentation (step 7) of a related project. In other words, this list is not necessarily sequential. Pick one, take action and see if that leads you into a virtuous cycle.

Then, once you’ve stabilized your infrastructure and stood up again, you’ll gain something really valuable: perspective and time to think. That combination will help you become more strategic and reach whatever your next goal may be.