Under the Hood of Ad Benchmarking: Building the Database

23 August 2023

2 min read

Learn about the database powering our AI apps and how they make industry benchmarking a breeze.

In the rapidly evolving world of advertising, it is vital to measure the success of your advertisements against your intent, your brand, and various industry standards. Quilt.AI’s Ad Benchmarking Tool provides an essential tool to predict the performance of your ads and compare them with high performing ads across different industries.

Built on an extensive database comprising more than 13,000 high-performing industry ads from 17 diverse industries, this broad spectrum of data ensures an accurate and multi-dimensional comparison, allowing brands to gain insights and inspiration to create better ads.

In this installment of “Under the Hood,” we unveil the construction behind this extensive database — a deep dive into how this industry database was developed in less than 6 months.

1. Curate a List of Industry Leaders

Initially, a comprehensive list of 600 industry leaders across 17 different industries was curated. This selection provided a wide range of examples and served as a robust foundation for the database.

2. Extract Videos

The next step involved extracting the most popular and recent videos from various YouTube channels of these industry leaders. This ensured that the data was up-to-date and reflective of current trends.

3. Data Cleaning

To enhance analysis integrity, we performed data cleaning by removing duplicate videos and non-ad content such as CSR efforts, employee interviews, and product demos. This process eliminated redundancies and noise, enabling more accurate comparisons.

4. Curation of Top-Performing Ads

The final step was to curate a list of top-performing ads, identified through a meticulous analysis of metrics like the like-to-view-to-comment ratio. This fine-tuned approach ensured the selection of truly high-performing ads, reinforcing the credibility of the benchmarking tool.

By harnessing automation, our goal is to elevate this industry database to 100k by year’s end, paving the way for even more insightful analytics.

In the upcoming installment of “Under the Hood”, we’ll delve further into the mechanics of constructing our analytics platform atop this database, offering a behind-the-scenes look at the innovation that fuels Ad Benchmarking.

Sphere’s Ad Benchmarking app uses predictive AI to rapidly analyse video data frame by frame and compare how your ads perform against various industry benchmarks

Transform the way you understand online data

Try now

arrow_upward