The future of content optimization: Speed up and scale ad testing with AI

Published in
6 min readJul 26, 2023

Sphere’s Ad Evaluation app uses AI to rapidly analyze video data frame by frame to identify symbols, emotions, cultural nuances, and whether brand assets convey intended values and meaning.

Remember Pepsi’s infamous Live For Now ad that seemingly trivialized ground up movements for Black Lives Matter and against police brutality?

How about Dove’s unintentional diversity backfire that sparked accusations of racial insensitivity?

Advertising cringe fests and controversies come and go quickly in this age of media ephemerality. But the ubiquity and speed of social media means that negative backlash hits brands harder than before and a single misstep can quickly snowball into a global PR disaster.

Ad testing helps companies avoid such mistakes by identifying potential issues before the ad is released or even while it’s airing. Brands can get critical feedback on how their content will be perceived by audiences and refine their assets according to what works and what doesn’t.

The downsides to traditional ad testing

Traditionally, ad testing is done through focus groups, surveys, or in-person workshops where a sample of the target audience is shown the ads and asked for their opinions. But as you can imagine, this takes a significant amount of time, effort and resources.

Survey respondents have to be recruited and incentivized, moderators hired, surveys written, locations booked, responses compiled, cleaned, and analyzed. The entire process can take up to a month or more, depending on the complexity of testing and number of assets being tested.

Response bias adds another stumbling block to the process and can hinder the accuracy of your findings. Answers from respondents may not fully reflect their actual perceptions if they have preconceived notions of a brand, product or other factor.

They may also respond with what they believe to be socially desirable answers when asked leading questions or about sensitive topics like religion, health and alcohol consumption.

Transforming the content evaluation process

Using AI-based creative testing instead of traditional surveys avoids the problems mentioned and automates a month-long process with the click of a button.

Sphere’s Ad Evaluation app harnesses the power of the internet and crowdsources mass opinions for an objective picture of how creative content will be received. Using machine learning, it determines if a brand’s messaging aligns with its desired values and provides the insights needed to avoid marketing mishaps.

Here’s how it’s a step up from conventional testing:

1. Speed

The app analyzes multiple pieces of creative material in as little as 5 minutes, giving instant insights in the form of user-friendly data visualizations.

Instead of taking weeks or months to ad test and risk releasing outdated content, Ad Evaluation allows brands to quickly analyze their strategy and make informed decisions.

2. Scale

The app is designed to test large volumes of content — whether it’s every single asset a brand creates, or doing a comparison analysis of all competitor content.

Instead of using a small sample of consumers to evaluate campaigns, the large scale AI model is able to computationally use the amount of information available on the Internet today. This gives brands the largest sample size they can tap into with convenience and cost-efficiency.

3. Reliability

Another advantage of the Ad Evaluation app is its lack of preconceived notions of brands or concepts. Unlike focus groups who may have biases based on personal experiences, the app gathers opinions from a diverse and impartial audience that provides a clearer understanding of how content will be received in the real world. People are more truthful in online discourse than in surveys and focus groups where social norms and basic decency come into play.

Harnessing the digital conscious of society

Human culture is built on a unique series of associations between images and words. Instead of using a small sample of consumers to evaluate campaigns, we wanted to tap into the entire digital society of human experience to do so.

The AI model behind Ad Evaluation was trained on nearly half a billion image and meaning associations to understand how we see the world. For example, a balloon or a smile may represent happiness.

By using training data from a vast array of digital mediums from historical books to stock imagery to Reddit threads, we built a new communication testing technology that takes into account cultural and generational nuances from any cohort.

Ad Evaluation is able to analyze each video frame, image, or text and calculate the proximity of these assets to the contextual meanings they represent. The results are a thorough analysis of brand values that come through in an advertisement and how closely each frame conveys a particular value.

Elevating your creative strategy

With an average person receiving 10,000 brand messages a day, your branding needs to stand out to create an impression in a crowded market.

This app helps you to:

1. Optimize advertising materials to best convey specific values

See all images in the ads that signal each brand value. Discover the types of objects, backdrops, colors and symbols that represent different values.

2. Analyze high performing videos to see what drives views

Analyzing Competitor’s Top 4 high performing assets, we can distill what drives views across countries.

Eccentricity, sensorial product shots, and daytime urban scenes may work well in Mexico, for example. Meanwhile, American viewers may prefer slice of life scenes, urban nightscapes, and aspirational imagery.

Using the app

  1. Select Data Source: Choose to analyze video commercials or short brand clips
  2. Enter Ads to Analyze: Link multiple videos to compare across campaigns
  3. Define Your Brand Values: Put in your own brand values or let Sphere’s AI detect relevant values

Review the results through visualized insights

  1. Get a sense of how much each ad signals different brand values

2. Explore how brand values are expressed in each video frame

3. See all images in each ad that signal the respective brand values

4. Learn about the personality archetype that your brand represents and the traits that the ad conveys to audiences

Request a Sphere demo to see how artificial intelligence can radically transform the way your company finds consumer insights and market trends.




We are a culturally rooted, AI powered insights firm that converts millions of data signals into human understanding. Visit us: