Customer Background

Our customer is a leading advertising house in California that caters to the requirements of established companies, organizations, and digital agencies. They are known for being the first to implement automation in their line of business and servicing the customers exceeding their expectations.

Project Background

The customer started facing a decline in the viewership of the commercials they make and this resulted in unhappy customers and reduced business. Competitors had better ideas to entertain the audience which made them win deals. The customer wanted to analyze the issues behind this decline of business and to facilitate that they wanted to understand what their audience like and dislike. 

Challenges:

The audience was spread across different localities and it would be pretty difficult to understand their tastes until they were monitored. It was near to difficult and not viable for executives to physically visit each customer and ask what their interests were or run a digital campaign understanding the taste of the audience as other media platforms would do. 

Solution:

Cubet had multiple sessions with the customer to understand:

  1. Business models being used to approach the audience.
  2. Feedback mechanisms currently adopted.
  3. Channels through which the audience could be approached.
  4. Channels with highest retention rates.
  5. Metricesrequired to analyze the customers. etc.

A plan was then created that would divide commercials into different segments each being assigned a specific rating and evaluating and assigning values to emotions that were expected after viewing each segment. The values were then computed and integrated into an algorithm that could calculate a user’s overall interest using probability and similar occurrences recorded as data values, and rate the involvement. A machine learning model was created on top of this so that the data values are constantly analyzed and reported.

The channel that best suited this study was cinema theatre where more audiences could be monitored and analyzed. A camera was mounted that would record visuals of multiple audiences when an advertisement was displayed. These visuals were then separated based on the age group, gender, etc. and evaluated against the segment of advertisement already calculated. An Artificial Intelligence based engine was developed to analyze the visuals and calculate the emotions they experienced during a segment. These emotions were assigned values and fed to the main application which would analyze the trend.

Result:

This enabled the customer to predict which commercials the audience liked and disliked based on their age, group, gender etc. Our customer used this study to analyze gaps and make effective advertisements.

Latest Case-study