From confusing tutorial to confident first session
Mobile puzzle onboarding
A mobile puzzle team came to us with a familiar complaint: early analytics
showed a steep drop during the first ten minutes, yet no one could agree on why.
Designers believed the tutorial was already too simple. Product leads worried
that cutting steps would hide important mechanics. Our task was to replace
guesswork with clear evidence. We ran a series of remote moderated sessions with
players new to the genre, combining screen recordings, think aloud prompts, and
short post session interviews. We tagged every hesitation, skipped tooltip, and
restart, then layered this with event data from the live build. The pattern was
sharp. Players were reading, not playing, for far too long. Key mechanics were
introduced in isolation, without the small moments of success that anchor
learning. We proposed a reshaped flow using what we call our ramp and reward
pattern: teach through tiny, winnable challenges, then explain only what the
player has already felt. After two design iterations and a follow up test, more
players reached the first meaningful puzzle, reported less confusion, and
described the game as “clever, not exhausting.”
Balancing action, information, and visual noise
PC action HUD clarity
A PC action title with a complex HUD was struggling with mid game retention.
Players who pushed past the opening hours tended to stay, but many never made it
that far. The team suspected difficulty spikes, yet combat metrics alone could
not explain the churn. We designed a mixed study: lab based play sessions
focused on combat encounters, followed by targeted surveys and a review of
anonymised heatmaps. During sessions, we noticed a recurring pattern. Players
were not failing because enemies were strong; they were failing because they
could not parse health, cooldowns, and objectives quickly enough. Eyes darted
between corners of the screen, while key audio cues went unnoticed. In our
report, we mapped eye movement, input timing, and death moments onto a simple
visibility matrix, then suggested a phased HUD approach. Early hours would show
only essentials, gradually layering advanced information once players
demonstrated comfort with the basics. We also recommended clearer feedback for
low health and cooldown readiness. After implementation and another round of
testing, players described combat as intense but readable, and far more of them
reached the narrative hooks the writers had crafted.
