Lessons learned from IRIS:VIDEO and where we go next

Webcam eye tracking won’t replace research grade labs — but with the right setup, it can deliver fast, actionable insights into attention and perception processes as key indicators for successful learning. Our recently finished adesso Research project together with the University of Bochum, here are our biggest takeaways from IRIS:VIDEO, and how they translate into use cases for adesso and our partners.

What we set out to build

IRIS:Video started with a clear goal: extend a virtual based learning setup for flipped classroom scenarios. In flipped-classrooms students first learn new content on their own (e.g., with short videos) and then use class time for practice, discussion, and problem solving. Webcam eye-tracking can be designed to support students during self-study. Eye-tracking data can for example inform the creation of new learning materials or delivery of already existing materials. Along the way to this goal, we focused firstly on analyzing which criteria are suitable for creating and steering the best learning experience. Secondly, we looked for the best way to integrate the webcam eye-tracking into existing and future platforms. Lastly we generated an outlook on use-cases that leave the academic purposes behind and harness their commercial potential.

What we actually built

• An embeddable, research ready Single Page Application (SPA) for browser based eye tracking: webcam capture, neural processing, synchronized logging, and export. Designed to be run as standalone and be integrated into different platforms, i.e. MathAdventure — a platform designed to teach students mathematical concepts through engaging, story-driven, or game-based approaches.
• A harmonization layer for multi source gaze data: aligning coordinates, sampling rates, and timelines across Tobii Pro (State of the art Eye Tracker), WebGazer (Webcam Eye Tracker) and AI outputs to enable comparability and publication ready analyses.

From lab to product: unlock everyday UX

Real world constraints (no lab visits, no Tobii at home) make webcam approaches appealing. With consent, eye tracking can piggy back on regular websites and apps to enable rapid, iterative UX diagnostics (e.g., regression testing for changes).

Future Use Cases for adesso and partners

Eye movements are widely recognized as an indicator of what learners understand and how well they process instructional material. However, most previous studies have been conducted under laboratory conditions, and aligning gaze data with experimental materials remains time-consuming and methodologically demanding.

Thanks to technological advances, findings from the laboratory can now be transferred to everyday study and workplace contexts. What is relatively new is the possibility of capturing the relationship between eye movements and learning outcomes in real time via webcam and adapting the learning experience on the fly. Our study therefore examines the extent to which laboratory findings on eye movements can be transferred to real-world settings.

IRIS:VIDEO shows that browser based webcam eye tracking—delivered via an embeddable Single Page Application with synchronized logging and a harmonization layer—can produce fast attention insights from afar (e.g., orientation, scanpaths, Area of Interest (AOI) misses) that translate into concrete UX and learning improvements; for adesso this offers a differentiating, reusable capability, and for customers it lowers the barrier to use evidence based iteration without lab hardware.

These results are immediately applicable to multiple sectors:

• recruiting and career portals (finding goal matching job listings, detecting Call to Actions (CTA), locating necessary information),
• digital learning in higher education and corporate Learning and Development (pacing, signaling, attention aware content),
• e commerce and content discovery (findability, banner blindness, add to cart visibility),
• enterprise app onboarding and support (first run guidance, form clarity, error recovery),
• and public sector citizen services (wayfinding and accessibility in online forms).

The project explicitly targets transfer from education into website/app UX validation and e learning analytics, with the SPA designed to integrate into diverse web platforms—groundwork that supports pilots and future integrations across industries.

Bonus: lessons learned from our material design research for sustained audience attention

• Keep one core idea per slide and remove anything that doesn’t serve your main goal.
• Group related elements and use generous white space to make the intended reading path obvious.
• Guide focus with consistent signaling (colors, contrast, minimal purposeful animation)
• Define clear Areas of Interest (the 1–3 must see elements) that you highlight visibly.
• Pace for understanding with step by step reveals and brief pauses so key points land.
• Design a clear hierarchy so attention flows through title → key visual → takeaway.
• Add light interactivity where it helps (quick prompts) and always favor accessible choices (readable fonts, sufficient contrast, clear labels) to restore attention.
• Before presenting, do a quick dry run to check that a first time viewer knows where to look first, second, and third—if not, simplify or strengthen your signals.

Key takeaway

By moving from a prototype to an embeddable Single-Page Application with harmonized gaze data, we made webcam eye tracking practical for both education and industry. The method thrives when tasks and materials are designed for it — and it delivers fast, explainable signals that help teams improve experiences where it matters most. Next up: pilots, benchmarks, and co authored publications to turn these lessons into repeatable outcomes.

Interested in a Discovery Sprint for your careers site or product funnel? We’ll map attention, fix friction, and validate improvements. Don’t hesitate to contact.