Screen assessments with minimal context

Activity UX content review of education software in process of being refactored. Mark up screens for future change consideration. Add to backlog for picking up later.

Challenges A very broad brief and no research budget. Context blind spots.

Resources Access to SMEs & domain knowledge, existing usability research (slightly old and unfocused)

Users/Audience School administrators and teachers

PROFILE

Assessing without the full context

When you start with a blank canvas, it's daunting, but there's many possibilities. When you start with a legacy of previous software and design, it's much harder to gain context and know what to change for the greatest benefit to users. Even simple changes have big repercussions.

Changing this pattern here means it needs to change everywhere.

Taking an approach to mitigate the challenges

Selecting an approach required me to first accept that I could not do much research or collaborate with designers/product teams to propose solutions. So I created my own set of rules:

Create consistency within the patterns. E.g. For windows of the same type, for forms, tabs, similar functions, etc.
Make ‘best guess’ plain language suggestions.
Go hard with the red pen. If I couldn’t be certain, at least be decisive.

Assessment activity

Identify content patterns (e.g. syntax action type)
Ensure key task is able to be met
Check that the simplest terms are used
Review language, such as 'records', 'range'
Suggest clearer language
Query the design of content elements
Annotate reasons for change

Making rules for patterns and calling out the big stuff

As I went, I catalogued the content patterns and gave them names, such as save records screens, pop-up confirmations, profile forms, one-off forms, option selections, etc. For each of these I tried to define rules, such as always use Save and Cancel, or ensure options are parallel.

While I put these forward as rules, I also pointed out where changing things in one place was going to impact many others. Such as Save and Cancel. This is how I communicated the risk of the change, so a future person could decide if the effort was worth it.

Pattern examples

Form navigation

Blank state

Button actions

Interactive calendar

Form field details

Outcomes

The client ended up with a well organised and tagged backlog of screen assessments. While I was unable to enshirine decisions in any kind of design system or style guide. I did create a list of enterprise level decisions captured in a single JIRA card, that their designers could pick up later. This prevented them having to go into each card to detect patterns.

The quick and dirty approach was not ideal. But the limitations eventually became more like clear boundaries and I could adjust what I was asking for to benefit the user, without creating new design work. Kind of like performing a makeover as opposed to a renovation.

“Design adds value faster than it adds costs”

Joel Spolsky