KAELAN sIMPSON

Prop Builder Post Mortem

The Problems I Wanted To Solve:

Make it easier to find props and assets

Our team repeatedly brought up that it was difficult to find assets in the project when building out or decorating spaces.  Even though our content browser was organized, we had so many props that it was difficult to keep track of all the options.  I wanted to solve this problem and give artists/designers an intuitive and visual way to see all the props at their disposal. 

Make it easier to populate spaces 

When I tested our art and design pipelines I discovered that placing small props took the bulk of total time spent populating a given area.  I wanted to see if I could create a tool that made placing small objects such as coffee cups and keyboards more efficient in Unreal..

Major Iterations:

Adding filters

At first I displayed all available props in bulk, this was an improvement over digging through the content browser but still took unnecessary time.  I add a filter section to the tool which allowed certain props to be filtered out.  For example, an artist could choose to only display props were light sources when doing a lighting pass.

Adding a pure placement mode

Not everyone was able to get comfortable with the custom placement controls I built into the tool.  However, the prop filter features were universally used.  To accommodate everyone on the team I added an option to disable the custom placement.  This allowed teammates who did not like the custom placement to still make use of the prop filter.

​

What Went Right:

The filtering was heavily used 

The biggest success of the prop builder was that it succeeding in making it much easier to find props.  The prop filter was widely used and saved the team time when populating levels.

The placement was more efficient 

The custom placement controls were more effective then the default controls in Unreal.  AB testing found the custom placement to be faster then default placement for the team members who fully adopted the new controls.

What Went Wrong:

The custom placement was too big of a change for widespread use

The custom placement controls I built into the tool took some getting use to.  I found that for most of our team that initial learning curve was not worth it.  As a result that part of the prop builder was never widely adopted.

What I Learned:

A tool only works if it’s used

Even though my custom placement tools were found to be more effective through testing, the learning curve required to see that improvement was high.  This learning curve made that feature effectively useless since it was never widely adopted.  In the future I will recognize that a tool can only be valuable so long as its used.  A great tool that is never used is no better then a bad tool that is never used. 

The learning curve for tools is important 

A major part of the prop builder failed due to my inability to make the tool accessible to everyone.  Going forward I will make sure to not neglect how the ease with which a tool can be learned.  A tool that can be easily learned is one that will be used.  A tool that is used is one that will produce the most value for a team.