Post

User research is critical for improving internal tools. You have to know who is doing what, when, and why. You cannot make proper improvements without knowing what the issues are. Seems basic, but is often still forgotten.

Especially because Tool Design is still a nascent field, most tool engineers have not worked with a designer before. So, they expect you to perform magic right away and 'do your design magic' to remove problems right away. This is unrealistic. You have to do research first.

88% of their userbase used the tools on a weekly or daily basis. So improving the internal tools is a huge boon to the entire company.

You have to observe, listen, and ask your users about issues. This gives you a huge amount of data to work with. For example segmentation, flow, hidden pain points, insights, and iterative optimisations.

Flying through great examples of improvements here! Check the recording on youtube once it goes live on the @gameuxsummit.com channel. Every upgrade wasn't just visual, but an active improvement to daily workflow.

Cross studio at Ubisoft, the amount of people that use the internal tools is immensely huge. 20000+ products, 1600+ games, these tools get used all over. This makes it difficult to find a mapping of how users use these tools, where the complexity is, and finding patterns.

Example of product names/tags, being organized in different ways Makes it very hard to find what you need.

The solution here was hard work: Going through loads and loads of games, understanding how they are set up, so the tool designers can understand how names and tags should be organized. So you have to understand the content the tool is for, otherwise, how can you design a better workflow?

And with that new naming structure, it also became much easier to design new workflows. Because now the tags could more easily be used for filtering and organizing. So the foundation of proper research and design makes future work much easier.

(hell yeah dude)

New problem space: Discussions between release managers and external partners. This involved many emails. Issues were: Huge risk of human error, inefficient repetitive work, and so much time. So how about instead of manual emails, there is a new service?

Again lots of great stuff here to catch on the recording, essentially the gist is: They created a clear step by step system, with clear status on each step, for everyone to look into at any time internally. This removed manual emails, missed information, and status became clear.

Another issue: Internal Ubisoft folks get free games to play. Nice! But getting them was problematic. There were a huge amount of support requests all the time! It was causing way too much time to resolve, and users kept having issues. It seemed straightforward. But it wasn't. A design problem.

Classic design process: Go over it with colleagues. Figma up some different workflows. Find out what works and doesn't work. Remove clutter and options which are not critical for user flow. And then, with the new design, ran through it live with actual employees over a call. And it worked!

This cut down the support ticket load by 50%, and improved the process overall. So, for takeaways, start with empathy, identity complexity, streamline flows, guide user actions, and so much more.

Next, Mike Ambinder is talking about the science of Game Design. It is about using scientific approaches to confirm or disprove the decisions you have made. He sees science not at odds with design and art, but to help assist it.

What Mike wants to put forward is certainty in a place that he often sees as uncertain.

Could you make game design as certain as physics, chemistry, and biology? Could you as effectively describe what rewards you can give players, with certainty?

That would mean creating theories for many things that we currently do, such as: Fun, Matchmaking, UIs, difficulty, retention, rewards, and accessibility. We are good at post-launch rationalizing, but can we predict what happens before that?

This would require frameworks that allow us to gather this data, compare, contrast, and share knowledge. Of course, the usual barriers would stop us: Time, and money. But wouldn't it be worthwhile to do this work, and create more certainty?

Another example: ELO distributions: Luck vs Skill vs Combination. There is data, so gather it, compare it, create theories, put them to the test.

Other data you can gather, through surveys: Quality of match, quality of teammates, Quality of yourself. By adding that last question, 12.5% of millions of complaint reports went down, because it made players realize that sometimes, they were a problem themselves. So they would try to improve.

So basically: Try and apply theory. See what you can learn to create your own theories. Save the data and things you learned, and share them.