Eye Tracking Without Hardware: How AI is Redefining Consumer Insight and Beyond
This article is based on a joint webinar we held with our partners at eye square, following the successful completion of an AI project: the development of a cutting-edge eye-tracking solution that works right through a user’s mobile phone. DAC.digital developed this tool for eye square to revolutionize market research by enabling equipment-free attention tracking that operates without specialized hardware.
The result is a system that delivers in-context insights into how consumers interact with digital content in their natural environments. Let’s dive into how we built this solution, the technical challenges we overcame, and the powerful applications it unlocks across industries like healthcare, education, and retail.
If you prefer just watching the webinar, you can find it on YouTube. Here’s the recording.
1. The Evolution of Equipment-less Eye Tracking
Traditionally, eye tracking required specialized hardware, like bulky glasses, infrared cameras, and lab environments. These limitations made large-scale, real-world testing costly and inaccessible. We explored these challenges and modern solutions in our previous article: What is Eye Tracking?
SEAL, that is the eye tracking solution that eye square asked us to develop, flips the script. It offers high-precision gaze tracking using only a device’s front-facing camera and a browser. This allows for massive scalability and real-world applicability across mobile phones, tablets, and desktops. Users don’t need to visit a lab. Eye tracking happens right where they are.
This leap didn’t happen overnight. It was the result of a carefully orchestrated process of planning, data collection, model development, calibration, and testing.
2. Building SEAL: From Concept to Real-Time Precision
Step 1: Defining the Challenge
Our goal was clear: build a highly accurate, scalable, and user-friendly eye-tracking system that works across diverse environments and devices. That meant designing for real-world conditions, such as imperfect lighting, different head angles, varying screen sizes, and global users.
Step 2: Data First
One of the biggest hurdles in AI development is data (just see our masterclass of AI project management that talks about data extensively). Public datasets were insufficient: they lacked diversity, didn’t reflect real-life use, and often came with restrictive licenses. To overcome this, we developed a custom web-based data collection system and used crowdsourcing to gather a diverse dataset. We engaged over 3,100 participants, capturing data from more than 1,800 device types. We made sure to cover a wide range of use conditions (screen sizes, lighting).
Step 3: Quality and Calibration
All collected data underwent rigorous cleaning. We set up automated filters for blurriness or poor angles, followed by manual checks. Our proprietary calibration pipeline turned noisy, raw gaze data into highly consistent and accurate signals. Before calibration, gaze points showed significant variance. Afterward, they aligned tightly with target locations which is critical for actionable insights.
Step 4: AI Model Development
Once data was ready, our computer vision experts trained and tested models to maintain high accuracy across varying conditions. SEAL typically tracks gaze within 1.5 cm of the actual point, even under bright light, low light, busy backgrounds, and small head movements.
Step 5: Real-Time Processing
We optimized SEAL for real-time performance, allowing it to capture gaze and head pose data, analyze it, and deliver clear insights with minimal delay which is crucial for attention-based applications in fast-moving environments like mobile feeds or e-commerce.
3. Why SEAL’s Browser-Based Design Matters
The hardware-free and platform-agnostic nature of SEAL unlocks opportunities previously unavailable to brands, researchers, and technologists:
- No App Needed: SEAL runs in the browser—no installation, no friction.
- Natural Environments: Users can participate from their own homes.
- Massive Scalability: From a single participant to thousands in parallel.
- Privacy and Consent Built-In: Every session begins with GDPR-compliant consent, calibration, and clear user instructions.
4. SEAL’s Eye Tracking in Action
For Brands and Market Research
SEAL allows companies to:
- Optimize Content and Layout: Measure dwell time, time to first fixation, and reach for any part of an ad or interface.
- Improve Experiences on the Fly: Adjust campaigns or layouts dynamically based on live attention feedback.
A real-world example? In a 15-second Carnival Cruisers ad, SEAL identified that viewer attention spiked at second 3 and again at second 8—key moments for placing brand logos or calls-to-action. This kind of precise temporal insight helps marketers maximize impact.
For eye square: In-Context Testing
SEAL powers eye square’s in-context technology, which tracks attention across social media, e-commerce, and other digital environments. Participants are recruited via online panels and complete browser-based tests, which include:
- Consent and callibration
- Live attention tracking
- Optional survey feedback
The result? Automatically generated, high-resolution metrics: what was seen, when, and for how long.
Applications Beyond Marketing
While SEAL’s commercial impact is already significant, the underlying technology offers broad potential.
- Healthcare: Gaze tracking supports cognitive testing, patient monitoring, and early detection of neurological conditions.
- Education: Measure student focus to personalize learning and identify distraction points.
- eCommerce: Understand shopper behavior, from digital product placement to web design.
- UX Design: Improve software interfaces based on where users naturally look.
- Manufacturing Quality Control: Use machine vision systems to detect defects or monitor processes.
- Agriculture: Analyze crop health via drone camera footage.
5. DAC.digital’s Role in the Future of AI-Powered Insight
SEAL is a testament to DAC.digital’s strength in custom artificial intelligence, computer vision, and eye tracking system development. Whether it’s gaze prediction or behavior analysis, we design systems that work beyond the lab.
SEAL is a premium solution for smartphone eye-tracking that operates directly in the browser. Thanks to DAC.digital, we’ve achieved a scalable, accurate, and intuitive tool that is changing the way market research works.
6. Eye Tracking for an Era of Edge Devices
With SEAL and similar AI-powered tools, we’re entering a new era of human understanding—one where real-time attention data can inform better decisions across industries. Eye tracking no longer needs to be expensive or invasive. Now, it’s seamless, scalable, and smart.
At DAC.digital, we’re proud to contribute to building that future!
Start with POC before the full roll-out. Build a Custom Eye Tracking AI with us.
