Direct collaborators: SophieAnne Peijnenburg
PostNL // Interaction design, Visual design, UX research, Tester
Project duration: 2 months at PostNL (4.5 months in total)
As a Communication and Multimedia Design student you have to choose your project in the 2nd year of your school career. I chose the interaction project, this project was in collaboration with PostNL who wanted a user-friendly interface for their delivery, pick-up and mail vending machine. After the 10 week project PostNL would pick two students who would continue this project for another 10 weeks to finalize the designs.
This was a school assigned project at first but after the project ended I was one of the chosen students to keep working on this project. First I thought we could continue where we ended but we had to start over and reassess our research.
The main problem that was plaguing the screens were the unclarity in the flows. It wasn't clear when or how you finished a flow. The system didn't give a lot of feedback about your actions and the labeling that was used also didn't help.
Within the project there were 5 flows that were assigned to students at random:
- A young adult that has ordered a package and wants to pick a place to pick it up through 3rd party applications.
- A young adult that wants to send a package.
- An young adult that wants to pick up a package.
- An older person that wants to pick up a package.
- A worker from PostNL filling up and emptying the machine.
Because PostNL valued my idea and how I visualised / tackled the problem they gave me the chance to continue this project at their office. Next to me Sophie was also chosen to finish this project. Here we got the chance to see all the work other students had done, and how we would use their answers for our own solution.
The first thing we had to do was look at the the solutions others had offered and how we could use that. So we visualised that into a poster. After that visualisation we looked at the research that supported the idea's, and discovered that some of the research wasn't strong enough to support the idea. The conclusions that were assumed didn't always have the research to support it.
The first thing we did was try and find correlating findings in the research we had received. Once we looked through all the research, insights and answers people came up with we decided to make a list of 10 insights and ideas we found useful (in consultation with PostNL), and we decided to use those ideas as a roadmap we could follow. With the insights chosen and the road clear we researched and interviewed various stakeholders with multiple techniques: Fly on the wall, Survey, Expert interviews, user tests. Everything we could use to make the insights more tangible.
For example, the older people had issues with the screens font because it was too small, so they'd rely on the illustrations but the illustrations didn't clearly show the process or what has happening for the users.
The error states were lacking, they were present but once you saw them they wouldn't tell you enough about the screen that you were looking at and what the exact problem was. Which in turn frustrates the user as they feel powerless, and have no idea how to fix the problem at hand (if it even is fixable at hand).
And finally the screens didn't communicate about the state of the lockers. We've done multiple user tests and before we changed the screens not one user understood which locker was open. To make matters worse, the users didn't even know that a locker was currently open. Which was quite a problem.
Once we'd reach the point where we could prove every small detail with data that supports it we started making new designs for the screens. We rigorously tested the screens with eyetracking and user tests. We used heatmaps to spot the area's we had to target and tweaked the designs accordingly. With every iteration the tests gave better, more positive results and the screens that we made eventually were understood by every tester (who hadn't seen the screen previously).
After testing like a madman we made the changes needed to give clear instructions or help to the user when needed. Check the final prototype out here (it's in dutch though).
The buttons were big enough for all users (but not too big), the illustrations said more about the actions and what was happening (you would actually read and see what locker was open). There was a breadcrumb menu that the users had so they could easily see what next step would follow, or what their previous step was.
Every step had atleast one error-state and every error-state would explain what was wrong and what could be done (or not done) about the problem. Every locker had a number and would be highlighted so the user would know which one was open. If the user made the mistake of going to the next screen to quickly, we added a forgiving element in which you could re-open your locker before you would end the flow.
This was the first project in which I worked for quite a large company. It was a fun experience as we didn't have to worry about any work outside of the task we had. Except I really like challenging myself and growing within a position, and I feel that a company that big doesn't always facilitate that. Nonetheless the experience was important for me to even form an opinion about this.
I really liked working hard on this project and testing to remove all the errors. It was very rewarding to see the previous wrongs made right by Sophie and me. I also enjoyed going out and interviewing people about their use of the system, I think that a lot of companies miss that nowadays. They mostly use surveys (that work very well and help a lot don't get me wrong) but once you step out and actually talk to the people you create it for, it starts to make more sense.
I'd definitely do a project like this again!