Adsvision
Adsvision delivers digital signage analytics that measure who actually sees your screens — impressions, dwell time and audience demographics — turning displays into accountable, measurable media.
Request a demo
Adsvision is a digital signage analytics platform from Alfabeta Solusi Nusantara that uses computer vision to measure real viewership instead of estimating it. A camera at each screen anonymously detects and counts the people who pass and the people who look, so advertisers work from actual reach rather than assumptions. As a DOOH audience measurement tool, it brings the accountability of online advertising to out-of-home and in-store screens, where reach has traditionally been guessed.
The platform combines several vision capabilities — people counting, vehicle counting, face detection, face recognition, and gender, age and emotion recognition — to profile who a screen reaches. This demographic ad targeting turns raw traffic into an audience picture: how many viewers were reached, their gender and age range, and their emotional response to what was playing. Data is processed in real time, and Alfabeta reports the accuracy of Adsvision's data at 95%, giving media owners a dependable basis for planning and billing.
Adsvision correlates audience data with the exact content playing at that moment, revealing which creatives, screens and time slots perform and which do not. That makes it possible to adjust billboard and in-store content dynamically, display specifically targeted ads for the audience actually in front of a screen, and even push notifications triggered by events on the digital signage. Beyond advertising, the same measurement helps retailers promote favourite products and broadcast area-specific offers based on who is present.
Analysis is designed to be privacy-conscious: gender, age and emotion attributes are estimated and reported in aggregate to profile an audience, rather than to build a named record of any individual. Results feed dashboards that present reach, engagement and demographic breakdowns, so signage inventory can be valued and priced on genuine viewership. The platform can also incorporate beacon-based data collection from visitors' smartphones, and integrates naturally with both advertising and retail operations.
Impression counting
Counts how many people actually pass and view each screen, giving a real reach figure for every placement. Automatic people counting means every viewer through the camera is detected without manual tallying.
Dwell-time measurement
Measures how long viewers engage with content, distinguishing a passing glance from genuine attention. This behavioural signal shows which creatives hold an audience rather than merely being seen.
Audience demographics
Estimates aggregate attributes including gender, age range and emotional state to profile who a screen reaches. Age is inferred from facial geometry and emotion from movement around the eyes, nose and mouth.
Content correlation
Ties viewership to the exact creative playing, so you can see which content, screens and time slots perform best. That insight drives specifically targeted ads and dynamic adjustment of what each display shows.
Privacy-conscious analysis
Reports anonymous, aggregate metrics rather than identifying individuals, keeping measurement respectful and proportionate. Demographic attributes are used to profile an audience, not to build a named record of any person.
Campaign dashboards
Presents reach, engagement and audience data in real-time dashboards that make signage as accountable as digital media. Event-based push notifications can also fire from activity detected on the signage.
Frequently asked questions
How accurate is the audience data Adsvision produces?
Alfabeta reports the accuracy of Adsvision's data at 95%. That level of accuracy is what allows it to be used for commercial purposes such as proof-of-play reporting to advertisers and pricing signage inventory, where estimates alone would not be defensible. Data is processed in real time so figures reflect current, not historical, audiences.
What audience attributes can it measure?
Adsvision combines people counting and vehicle counting with face detection and recognition, plus gender, age and emotion recognition. In practice that gives you the number of viewers reached, their gender and age range, and their emotional response — happy, sad, angry, surprised and so on — inferred from movement around the forehead, eyes, nose and mouth. Together these profile the audience in front of each screen.
Is Adsvision privacy-conscious, given it uses cameras and face recognition?
Yes. The analysis is designed to report anonymous, aggregate metrics that profile an audience rather than identify named individuals. Demographic attributes such as gender and age range are estimated and summarised for measurement, not used to build a personal record, keeping the deployment respectful and proportionate for advertising use.
Can Adsvision target or change ads based on who is watching?
It can. By correlating live audience data with the content schedule, Adsvision supports demographic-based advertisement display and specifically targeted ads for the viewers actually present, and it can push notifications triggered by events on the signage. This lets media owners adjust billboard and in-store content dynamically instead of running one fixed loop.
Who benefits most from Adsvision?
It is built for two main audiences. Out-of-home and DOOH advertisers use it to prove reach, measure engagement and price inventory on real viewership. Retailers use the same measurement to promote favourite products, tune in-store screen content, and broadcast area-specific discounts based on who is present, making Adsvision effective across both advertising and retail operations.