Watched is a content intelligence product helps content production companies get insights about their viewers. We help our clients to determine their target audience reactions, saving their time on production and marketing budgets and making a product launch results more predictable.

😫 🤗


Our platform provides a high-quality remote analysis of target audience reactions without the use of any additional equipment. The AI-powered system allows your company to determine the following parameters: engagement, attention, interest, comparison with expected reactions and many other things


Reflects the ability of content to cause the viewer to express emotions


Reveals the viewer's focus on the viewing process, a manifestation of interest in what is happening on the screen


Displays how much a product is liked and aroused interest


Allows to check the reactions of the audience with those that were laid down by the creators in the content

Emotions recognition process

While viewing the content, device camera scans facial expression, pose, gaze and determines user response using reaction recognition model

Step 1: Demonstrate your video content

  • Media content
  • Featuring films
  • Animation
  • Trailers
  • TV-shows
  • Commercials

Step 2: Collect reaction metrics

Reports on recognition of target audience reactions:

  • Engagement
  • Attention
  • Interest
  • Accordance with expected reactions
  • Survey

Step 3: Get better results

  • Content production
  • Investment decision
  • Marketing strategy

Most frequent use cases

Watched enables obtaining data for forecasting interest, satisfaction, and profit, improving content and marketing strategy, understanding the response of the audience of other countries.

  • Determination of the working version of the editing (A/B testing)
  • Determination of moments of the viewer’s distraction
  • Confidence in the Work/Doesn’t work moment (joke, tension, fear)
  • Product placement testing
  • Determining the response of the target audience to pilot projects
  • Testing distributed content (e.g. 1-2 episodes of tv-show)
  • Target audience segmentation
  • Comparison with reference product
  • Forecasting the desire to watch further
  • Retention Prediction
  • Getting insights and viewers response analysis on reference materials
  • Script reading analysis
  • Actors sample analysis
  • Character testing
  • Defining attributes based on viewers preferences
  • Determination of the working version of the editing (A/B testing)
  • Confidence in the Work/Doesn’t work moment (joke, tension, fear)
  • Confidence in dubbing
  • Comparison with reference content
  • Determination of the most engagement moments for trailers
  • Product placement testing

Our partners

News about us

Forbes Ukraine

Украинский стартап Watched привлек $450 000 инвестиций. Он занимается оценкой эффективности видеоконтента

Read more

Kyiv Post

Ukrainian startup supported by Amazon, NVIDIA raises $450,000

Read more