Protecting young gamers’ privacy and their agency.
for EPIC GAMES
LEADING GAMING PLATFORM, PARENTS OF FORTNITE
LEADING GAMING PLATFORM, PARENTS OF FORTNITE
Timeframe
3 months
Role
UX strategist
UX researcher
Interaction designer
Methods
Rapid prototyping
Qualitative research
User interviews
A/B testing
Usability testing
Visual design
Tech stack
Figma
Google Slides
Google Docs
Project team
Associate Director of Strategy (Sota P.)
Senior Strategist (Lizelle G-E.)
Senior Visual Designer (Vanessa D. S.)
THE MILLION DOLLAR QUESTION
A large part of Epic’s player base is under 18. Strategizing Epic’s player profile feature begged the question – when the player is a child, who should decide how visible that profile is? The child, or their parents?
.avif)
As this was a brand new feature, there was no precedent within Epic to help answer this question. I kicked off with a competitive analysis across the platforms young gamers in the US are most likely to create accounts / profiles on: from Google and YouTube to Roblox.
The results were surprisingly polarized.
Platforms like Google afforded extensive, granular controls (all to the parent, none to the child). Others (like Twitch 👀) basically said “Your child has access to minimal privacy controls. If you’re their parent and you don’t like it, go ahead and delete their account :)”
Where on the pendulum did we want to stand? *cue tech-philosophy and tech-ethics questioning*.
Where on the pendulum did we want to stand? *cue tech-philosophy and tech-ethics questioning*.
OUR STANCE
After a good amount of philosophical and ethical pondering, we decided it should be both (note: we had no clue how that’d work). So we pitched Epic a hybrid approach to profile privacy.
They loved it – yay! BUT, this approach was clearly unprecedented, there was no standard or even slightly-beaten-path to follow here. That meant moving into design, we’d have to create:
01 / Intuitiveness that didn’t rely on familiarity
The challenge? Making two different generations (parents and their kids) feel equipped and in control... in a process they had never seen before.
02 / A balanced pace
Such a large education need meant more steps. Instead of plain speed, we had to aim for a smooth, healthy pace.
03 / A real education opportunity
For both parents and young players to feel safe while teaching them agency.
FIRST PROTOTYPE
Because we were starting from a blank canvas, we prioritized quick + dirty design rounds and frequent user testing so we could fail...fast.
Our initial expectation was “getting to a functional prototype should be pretty quick – how many different ways can you design a profile privacy flow, heh?”. How delusional we were. But I’m getting ahead of myself.
At that point, we took a first stab at a prototype to test with actual parents of young gamers (on the left). I designed, and my senior teammate reviewed/provided feedback.
Then, we tagged team with Epic’s research team to test it with 9 parents of young gamers in 1:1 interviews. Parents viewed an interactive version of this modal, and were prompted to adjust the level of privacy their child can access.
At that point, we took a first stab at a prototype to test with actual parents of young gamers (on the left). I designed, and my senior teammate reviewed/provided feedback.
Then, we tagged team with Epic’s research team to test it with 9 parents of young gamers in 1:1 interviews. Parents viewed an interactive version of this modal, and were prompted to adjust the level of privacy their child can access.
OUCH, THAT WASN’T EVEN CLOSE
None of the parents understood that they were setting a range of authorized options, and not their child’s privacy setting.
So we doubled down on education and guidance.
At that point, we realized how much of a different beast concepting a hybrid privacy setting approach was to the classical approach. So we decided to test broad before we tested narrow.
Our next prototype was an A/B test, varying both in the components themselves and the sequence they were presented in. The idea isn’t to add all of these components to the flow, but rather to identify which are the most effective in supporting parents as they make their decision, then work those into our flow.
At that point, we realized how much of a different beast concepting a hybrid privacy setting approach was to the classical approach. So we decided to test broad before we tested narrow.
Our next prototype was an A/B test, varying both in the components themselves and the sequence they were presented in. The idea isn’t to add all of these components to the flow, but rather to identify which are the most effective in supporting parents as they make their decision, then work those into our flow.
With this second round of testing, we were looking to answer three big questions.
01 / Should we visually signal the nested nature of the different settings?
Or have parents set a single limit to how public their child's profile can be?
02 / At what point do we surface the child’s preference?
In the modal where parents set permissions, or before they do so, in a modal that tells them “hey, your child requested this setting, do you agree or do would you like to set different permissions?”
03 / Where do we confirm the parent’s selection?
i.e., where to explicitly state how the parent’s selection translates in terms of options their child can select: in the modal itself, or in a dedicated confirmation after they save their selection? Here, we wanted to learn how parents processed that information: is it most helpful as they’re taking action, or as its own focus?
OKAY, MUCH BETTER
We A/B tested these directions with 10 additional parents, through 1:1 interviews.
The interactive prototypes allowed parents to set a cap on the profile privacy options their child could select in 2 different ways – one after the other.
Here’s what we learned.
Here’s what we learned.
“It’s thoughtful that you’re not setting the setting for them but giving them the option [...]. I don’t see that a lot in other places.”
– Parent 4
– Parent 4
FINDING
Our hybrid approach to parental controls felt like a valuable education opportunity to parents.
OPPORTUNITY
The hybrid approach was worth figuring out. We were on to something good.
“I have a huge amount of trepidation about sharing anything about my kid.”
– Parent 8
– Parent 8
FINDING
Feeling informed and in control is paramount – to the point where clarity prevailed on speed. Parents were willing to take more time to access information that would help them make an informed decision.
OPPORTUNITY
Provide context and clarity in each step of the flow. Explicitly note the scope of the parent’s selection, repeteadly.
“It was not clear that the options are connected, even after I read the copy. Seeing the visual of the choices together gave me that immediate clarity.”
– Parent 2
– Parent 2
FINDING
Showing information graphically was much better understood than telling it through text.
First, parents didn’t really read the text on screen. When they did, they didn’t necessarily register the information. Instead, they tended to look at the UI, and infer the action to take from that.
Information displayed graphically also captured their attention better, and was more easily understood.
First, parents didn’t really read the text on screen. When they did, they didn’t necessarily register the information. Instead, they tended to look at the UI, and infer the action to take from that.
Information displayed graphically also captured their attention better, and was more easily understood.
OPPORTUNITY
Pair all key information with visuals to direct parent’s attention, and to use UI elements that really drive home that parents are setting a range for their child to choose from.
TOES & FINGERS CROSSED
Our next iteration was a direct response to these sticky points, integrated the modules that tested best into a unified flow. We couldn’t wait to get the “yay” or “nay” from the parents.
GETTING WARM
Version 3 tested really well – parents understood the scope of their action and felt adequately supported along the way.
This time I ran the interviews myself – it felt super rewarding to see actual parents of gamers interact with what we had made.
Although the flow was functionally validated, testing showed that copy could be more explicit in the first two screens. We tweaked the copy, I ran what ended up being our fourth and final round of testing, and there we had it: our unanimous “yes” from testing participants.
Although the flow was functionally validated, testing showed that copy could be more explicit in the first two screens. We tweaked the copy, I ran what ended up being our fourth and final round of testing, and there we had it: our unanimous “yes” from testing participants.
Behind the scenes, documentation was key (given the size and breadth of our stakeholder group).
I created reports to capture the insights from each test, translated those insights into action items to guide visual design and copy updates, and packaged key learnings and recommendations into slide decks for our weekly client shares.
NOTES
If I could turn back time, here’s what I’d do differently.
01 / Expect to iterate
We tested version 1 expecting minor updates – not a complete overhaul. We were in uncharted territory, and our design process had to reflect that.
02 / Protect the integrity of the research when outsourcing it
Make sure the questions aren’t asked in any leading form.
03 / Give this feature more of a moment
Beyond setting a new standard for privacy, this feature helps parents equipping their children to exist safely internet and build agency. This could have been a very cool brand moment for Epic.
OVER TO YOU
Let’s be real, case studies can be an absolute bore. How did this one feel? Let me know here in under 2 mins. I read every single one and implement on the patterns that ~reveal~ themselves.
ROOM FOR ONE MORE?