Test IT Solutions Before You Invest

Written by Jaren Green | 18 January 2010

The Marketing Department says, "We asked IT for a bike, but they delivered a car." How do these things happen?

In my corporate job, my role was part Business and part IT. From the middle, I saw many disconnects. Today, I want to talk about one that causes frustration among the Business people. Occasionally, they will ask for a simple IT solution, but it gets out of control and grows into a big monster project.

You are probably t‌hinking that the Business request was not really all that simple -- the business person just thought it was simple. I know that happens, but we can save that discussion for another time. A realistic example of the situation referred to above is a test marketing project.

At some point, your company may need to show the world a website and a specific offering to see how people will respond. The results of this test might justify the creation of a more complete solution. On the other hand, the test might just as easlily justify shutting down the project.

Best practices for online market tests include:

• Agree to testing ideas before investing in them
• Understand what is being tested and measured
• Agree on the plan if the test fails
• Agree on the plan if the test is inconclusive
• Agree on the plan if the test is successful
• Make it a temporary process on temporary systems
• Provide for a fast start and a fast termination
• Manage to the lowest cost and effort for the test
• Manage to the lowest possible distraction for the group
• Let others do what they can without IT labor

We must protect the company image, assets, and security, but that should not

preclude us from setting up a temporary website for testing, especially when all agree on what will happen next. There is a pain you feel deep inside for those IT Groups that spend months on a full-blown solution that will ultimately fail. Test before you‌ invest.

Keep the scope down on your test system. Both Business and IT will be tempted to make it too good. It is better to test on a low cost, temporary system that requires manual processes.

Do not over-deliver if it increases time or cost. If the test represents a significant investment to the Business people, they are more likely to hesitate when it comes time to shut the thing down. Allow the test system to have flaws and limitations. Cut some corners and save money. Don't invest too much until after the test.

Make sure you have an exit strategy and a date on the calendar. If the test works, but the funding is delayed, hold to your exit strategy. Never allow the temporary system to become the permanent solution.