Advertisers are spending millions of dollars on media that doesn’t deliver outcomes for their brands.

Why? Because there is a problem with underlining measurement. Time-in-view (one of only two critical input variables in our currency) makes the assumption that humans are focused, but that is far from reality.

We’ve proved through robust, global research that 75% of viewable inventory gets zero human attention. 

MRC impressions


Attention is the human-centric principle that determines whether people are really watching ads, both in terms of how long and how focused.

And by understanding real-world behaviours, you can develop more successful campaigns with tangible results.

Brand Growth: Attention has a strong correlation with business outcomes and driving brand growth.

Effectiveness: Are humans actually watching? Attention plays a pivotal role in advertising effectiveness and its impact on the bottom line

Mental Availability: Better understand and cater for how ads are processed to help reveal how much attention is needed to create and place distinctive brand assets in order to deliver mental availability.

Compete for Attention: By knowing how audiences choose to pay attention helps you understand how ads can better compete for attention e.g. addition of auditory attention to visual attention.


Incorporating attention measurement into media strategies offers a host of campaign and commercial benefits. Attention metrics can greatly enhance advertising effectiveness by providing insights into which platforms and ad formats capture attention and as a result, resonate with viewers, enabling more informed creative decisions. Attention metrics also improve return on investment (ROI) by helping marketers and strategists optimise media spend, focusing on platforms and formats that deliver higher levels of active engagement. It can become a powerful tool in your arsenal, providing greater control of media planning and buying decisions, so that you can ensure you’re paying for the ads that are served in environments where consumers are more likely to be receptive, in-turn increasing the likelihood of capturing valuable attention and driving desired actions.


Person-Level Human Data: We collect audio and visual data via computer vision techniques (including gaze tracking, facial detection and pose estimation) while viewers are using real-time platforms. Privacy safe facial landmarks are parsed through 3 models of attention, active (looking directly at ad), passive (looking nearby but not on ad), non-attention (nowhere near ad) to understand format-level ad effectiveness. All technology developed and used is proprietary and the data collection process adheres to the strictest global and local regulations, with participants opting in and providing full consent throughout their experience.

Impression-Level Device Metadata: Media consumption and media placement data is collected via JS Tag (VAST tag 2-4.2) appended to creative (Web, CTV and Audio Streaming: PC, Mobile, Tablet, TV). We use the accuracy of the observed data and the scale of the impression data to train machine learning models that can predict human behaviour in-flight and at the point of transaction.

For truly effective and trustworthy attention measurement data, attention vendors should be utilising a combination of both person-level and impression-level data in their methodology and offering. Person-level human data collected via gaze tracking and/or facial detection when they are consuming media in real time tells us exactly how much attention-time, and attention-focus, a human pays to advertising. While impression-level data collected via tracking pixel to collect data on a user’s scroll speed, time-in-view, ad pixel load, and ad coverage tells us how the ad loaded during their session and how it was displayed on the screen but the quality of the attention values is based solely on the quality and quantity of the ground truth data upon which the estimation is based.

One is quality rich but low scale with millions of data points, and the other is quality poor but high scale with billions of data points.


We have built a host of robust and sophisticated machine learning models that are able to predict active, passive and non-attention from real, in-the-wild audience footage. This has been built beyond iris detection, face detection and pose estimation and is supported by an age estimation model for our TV collection to help with demographic profiling and ethical data capture.

There are then additional proprietary machine learning models that predict both attention time and attention focus, to specifically support the planning and buying of media as well as aiding in-flight measurement. These models are drawn from our person-level and impression-level data to provide unparalleled depth, granularity and accuracy to our human attention dataset.

How we collect attention

The only way to capture real, meaningful human attention data is to collect it from genuine environments in a natural and non-disruptive manner. Some attention gathering practices involve the use of in-lab recording, require the user to wear clumsy goggles or place the viewer in environments or situations where their viewing behaviour can be influenced. With Amplified Intelligence, we ensure at all times that there is no distraction, disruption or discomfort to ensure the most effective data capture possible.

Attention signals can be captured, and harnessed, at scale across the entire advertising lifecycle to drive effectiveness and efficiency. From TV, CTV, BVOD and streaming. To digital environments like social media, open web and gaming environments, and a host of offline mediums such as cinema, outdoor and audio.


Participants recruited following strict data and privacy processes


Natural platform use, intercepted ad load,
5 frames/sec


Footage parsed
through ML models


3 levels and a multitude of attention behaviours assessed

Our Products


Understand how people are viewing your advertising.

Test your branded creative and messaging using real human attention in real environments. Ethical collection of human attention to advertising through GDPR compliant triple opt-in panels. Machine-learning models designed to create true omnichannel measurement and true comparison across environments, devices and formats.

Learn more...


Plan your next campaign with the world’s most accurate and rich attention-based media planning tool.

Predict the amount of human attention your channel mix will attract and plan every detail of your campaign to increase the long-term lift of your brand. Use attentionPLAN® as a standalone web tool, or integrate it into existing systems through our advanced API offering.

Learn more...


Using real human attention for in-flight campaign optimisation.

Attach an attentionPROVE® tag when you purchase your media to unlock valuable attention insights. Replace performance assumptions with a complete view of audience viewing behaviours, fuel instant optimisation by knowing which ad creatives, messages, and environments generate the highest quality attention.

Learn more...

Why Amplified?

We have the largest database globally of real human attention data collected from over 110,000+ people viewing all major media with 860M+ individual data points across 15 markets, deployed across 26 markets.

We have built our own methodology and technology for collecting human attention data via app based gaze estimation, natural viewing environments and market-leading machine learning.

We’re able to offer second-by-second views of attention performance of media and creative by combining real human level attention data with inferred impression level device data for accurate modelling and prediction.

Our end to end product suite measures, evaluates and optimises attention across all the major media channels to enable optimal media planning and activation, then maps this to brand outcomes.

The 9 Principles of Amplified Attention

Our work is guided by connected principles


Academic foundation;
methodology; control.


Accurate collection of gaze.
Test and improve error rates.


Tech developed to reach a
valuable level of granularity.


Leading the way in combining
ML with marketing science
theory for attention.


Challenging the efficacy
of accepted norms.


Insight-led tech and tools
development feeds
industry application.


Delivering honest results
in the face of client
and industry expectations.


Natural viewing environments.
Real human views.


Ensuring that platforms are
treated fairly in methodology.