Sprinting through design

We're working in design sprints for the usask.ca homepage upgrade. And after only one week we already have our first prototype.

${vImageAlt}
A typical web design process operates in a big loop with a lot of time spent building and launching. We want to shorten that loop and focus more on learning from users and improving. Image adapted from this one by Google Ventures in their article on design sprints.

Every week we’re doing a design sprint in the usask.ca homepage upgrade. We’re condensing what is traditionally a multi-month web design process into one-week loops. That means each week we’re doing the following:

  • Reflecting on user research
  • Designing a prototype informed by that research
  • Testing that prototype with users

These sprints help ensure we work fast, but they also ensure we’re not wasting time on things that don’t matter. If our testing one week tells us users don’t need a particular feature, we no longer have to spend time building that feature. We can focus instead on improving the things users do need.

This week we kicked off our first sprint. The design team had collected a few dozen user stories by interviewing real people. We condensed those stories into six representative versions to narrow our focus:

  • As a prospective student, I want to find out what programs the University offers so that I can get my applications in on time.
  • As an alumna I want to find news stories so that I can keep up to date with the U of S.
  • As a staff member I want to find the contact information for another staff member so that we can coordinate our work.
  • As a parent I want to find schedules and costs for swim lessons for my kids so that I can keep them active and busy over the summer break.
  • As part of a newly engaged couple I want to find out what the U of S has for catering and event space on campus so that I can know if the campus is a good option for us.
  • As a current student I want to find out what classes I can take so that I can finish my degree.

Because we’re working based on assumptions rather than requirements, we’re making an assumption that the homepage should help each of these users satisfy their story, even if that’s just through the right link on the page. These five stories do not represent the full spectrum of stories out there but they’re a place to start.

We’ve also got a bunch of data from our current homepage usage that can help inform our assumptions. For example, from that data we can assume users want a link to PAWS and the ability to search from our homepage.

So we started sketching.

${vImageAlt}
Team members sketch out their first ideas for the usask.ca homepage

The team divided into five pairs with each pair sketching a homepage design. Some pairs sketched with the paper in portrait orientation, some in landscape.

We put all five designs up on the wall and gave everyone some stickers. Each person placed stickers on the individual pieces in designs other than their own that they thought were the best ideas. Some areas of some designs got lots of votes, some got a few, some got none.

${vImageAlt}
Silent voting reveals which pieces from all our designs resonate the most
${vImageAlt}
${vImageAlt}

Popular ideas were:

  • Big search box
  • Prominent PAWS link
  • “Campus services” link

Unpopular ideas were:

  • Navigation bar
  • “Cool facts” button
  • Instagram gallery

We also noticed that most designs had much less going on than our current homepage.

We used those results to do another round of sketching. This time we had two groups of five people. Each group made a homepage sketch based on our voting results in addition to our user research.

We placed those two designs on the wall and had a simple vote to see which design was preferred. 

${vImageAlt}
The design on the right wins, but just barely. It will become our first prototype.

This all took about an hour and it gave us our first prototype design. We now need to test that design with users so that next week we can learn from the feedback, make the necessary improvements, and then test it again.

After that second sprint we’ll be ready to open it up for anyone to test. And then we’ll keep on going.