Welcome to, we're so glad you're here! We'd like to welcome you with a short video from our CEO, Coen Olde Olthof:

What can Junbi do for me? is an AI-powered YouTube creative insights platform. It is powered by - an attention prediction platform that allows you to analyze any creative with amazing accuracy, without needing participants. enabled you to get second-by second insights as to how your ad is able to grab the attention of your viewers on YouTube - in seconds! So you can make sure to test each and every YouTube ad you create before you go live. Ensuring you don't waste your precious ad spend on ineffective YouTube ads.

How does it work?

It works by first creating an attention prediction (similar to eye-tracking, but predictive) heatmap of your ad using our attention prediction platform Our software predicts eye fixations on natural images and videos by highlighting areas that have high salience. After the heatmap is created, Junbi is able to create your effectiveness scores by analysing several visual elements of your ad which have been found to be key drivers of effective YouTube video ads. This includes how much attention is on your brand, how much ‘visual clutter’ is in the ad, and how much attention is given to the ad relative to its surroundings in the YouTube environment

Develop, test and improve
As you don’t need any participants, you can test for attention while working on your ad. Results are generated in a matter of minutes. 
All of that, with the simple click of a button, so you can focus on what really matters.

An often-overlooked resource
While eye-tracking is a great resource, traditional lab-based research is time consuming and therefore rather costly. With predictive eye-tracking we are changing the game, as we cut out the need for participants and an eye-tracking lab.
Our platform isn’t just cheaper, it’s also a lot faster than traditional lab-based research. So, don’t have the research budget or the time available to do traditional eye-tracking? 
Think again!

Predictive or traditional eye-tracking?
While both predictive and traditional eye-tracking have its pros and cons, they have one clear overlapping metric; the accuracy of the results. 
MIT has developed a benchmark specifically for this purpose and the results don’t lie. 
Our platform scores a 0.87. Compared to real eye-tracking that’s an accuracy of 94%.
Don’t believe numbers? Check out the video below.

Still need help? Contact Us Contact Us