How to Design Systems that Actually Solve Problems

Have you ever noticed that systems designed to provide solutions sometimes end up creating more problems? This unfortunate side effect is often caused by a lack of understanding of what users need and expect. Let’s take a look at how you can avoid creating more problems and instead, design a system that delivers the solutions your users need. 

1. Identify needs and users

This is where you make hypotheses based on what you know about your customers and their needs so far. Lead with your customers to make sure your perceived business needs and goals are in alignment with what your customers really want.

2. Conduct research with your customers

Gather analytics, conduct user tests, talk to focus groups, create personas and journey maps -- whatever you need to do to test your hypotheses. For instance, the Lokion team frequently leads client teams through the creation of personas (“summary” users who represent key audience segments) and journey maps to walk those personas through their interactions with the client. Our team has also created dozens of unmoderated online tests allowing users around the world to navigate a proposed solution and give feedback.

3. Design features according to customer needs

Populate your backlog with features and success criteria, solicit user feedback early and often, evaluate features and success criteria based on feedback, update the backlog accordingly, and repeat.

4. Build feature slices

Feature slices are small, testable, freestanding pieces of functionality within a larger project -- using the same kind of feedback loop. The feature slices don’t have to be released publicly; they can be tested by small groups first.

5. Prepare for launch

When you release a set of features for user acceptance testing, keep evaluating features and either revise right away or add revisions to the backlog depending on priority and schedule.

6. Launch

After you launch, keep measuring actual performance against your success criteria and follow up with customers. Track the analytics, track customer feedback, and keep an eye on what’s working and what needs improvement. Seed your backlog with potential improvements and formulate new hypotheses for research. Then return to step 1!