top of page

Startup of the Month: Neuraflow Predictive Media (Brazil)

PackTech Ventures - Newsletter #2 - December 2022

How much is predicting the future worth? How to anticipate consumer engagement with your packaging, brand, and/or advertising, even before launching it to the market? The “Attention Economy” shows us that the task of retaining the consumer's attention is increasingly difficult. To know if your ad, website, app, packaging, label, indoor and outdoor visual communication will perform as expected, you need to invest and launch them “officially”, and do type AB tests, for different audiences and with significant samples. And that always ends up consuming time and costing even more than expected, with no guarantee of success.

In 2022, its co-founder and expert on the subject, André Gazineu, faced the challenge of predicting this future, using Artificial Intelligence (AI), and incubated the idea in partnership with ProjetoPack. Neuraflow emerged with the pioneering purpose of solving the problem of uncertainty about the success or cognitive deficit of the consumer to a new or already established idea or design, using Artificial Intelligence, specifically Machine Learning and Deep Learning.

The technology employed allows the company to provide valuable insights, with visual representations such as heat and attention maps, which will save time and money in the ideation and design phase. This will make the concepts for visual communication better worked before spending financial resources on production and dissemination.

Unlike research services with Eye Tracking, which require panelists and the use of special cameras that follow the direction of gaze upon a visual stimulus, Neuraflow's technology makes use of a predictive eye tracking system based on deep learning and trained in more than 30,000 images from studies of this type of tracking. This predictive system – called deep convolutional neural networks (CNNs) – is the same type of technology used in image and video recognition in recommendation systems (such as that used by Netflix and YouTube). However, Neuraflow's model was built from data collected using specific, high-engagement social media advertising images that cause high visual arousal. With this, it is possible to perform faster and more economical tests, with immediate predictive feedback and a high degree of accuracy.

Learn more about Neuraflow by visiting the website:

Copyright © 2022, PackTech VenturesTM. All rights reserved.


bottom of page