Sensimity
A neuroscience-informed player-research company serving mobile gaming studios. Founded by Katie Mattingly (CEO) after over a decade in mobile ad tech (Adjust, AppLovin), motivated by the industry’s stark retention problem: typically ~30% of budget goes to acquisition, yet 97% of acquired players churn within the first month and only ~5% ever convert to paying.
Methodology
Sensimity combines:
- Facial coding — automated analysis of players’ nonverbal facial reactions while playing.
- Eye tracking — where attention lands and for how long.
Their proprietary metric is called emotional intensity — a measure of nervous-system arousal at specific game moments. Includes a valence dimension (positive vs. negative) so they can distinguish e.g. delight from frustration spikes.
The positioning vs. competing methods:
| Method | Answers | Cost | Environment |
|---|---|---|---|
| A/B testing | ”What changed?” | Cheap | Production |
| EEG / fMRI | ”Why did the brain react?” | Expensive | Lab only |
| Facial coding (Sensimity) | “Where did players actually react and how strongly?” | Cheap-ish | Remote / scalable |
Their pitch is the cheap-scalable middle path: not the depth of fMRI but enough to localize which moment in a play session produces engagement (or doesn’t), so studios can iterate on specific mechanics rather than blindly rerunning A/B tests.
What they do with the data
Per The neuroscience of rewards - how dopamine builds game addiction (video), Sensimity benchmarks emotional-intensity peaks against game subgenre and target audience, then advises studios where mechanics are under- or over-firing. Examples cited:
- Building upgrade with no animation → no engagement spike (mechanic working but invisibly)
- Character upgrade + animation → spike + second peak from animation amplifier
- Hero acquisition from chest → sustained climbing engagement
- Facebook-connect reward → reward present but underwhelming → “delivery” needs reinforcement, not the reward itself
Why they matter here
Sensimity sits at an interesting intersection: they translate the Reward prediction error / dopamine literature into operational metrics studios can act on. Their public-facing content (this webinar, an upcoming monetization study) is one of the few sources walking through real applied facial-coding case studies in shipping mobile games.
Caveats
- Methodology details (what facial-action coding system they use, how “emotional intensity” composites are computed) are not disclosed in public talks.
- Case-study data is presented as anonymized summaries; replication is not possible from public material.
- The “abysmal 30-day post-event retention” claim is from a single client cohort and not a published benchmark.