The launch of ChatGPT has ignited imaginations worldwide, with visions of intelligent personal assistants managing our daily lives, virtual advisors boosting our wellness, and even chatbot friends cheering us up. In TV advertising, this translates to dreams of infinitely personalized creatives and auto-optimizing campaigns. While Generative AI grabs headlines, Vertical AI has been reshaping TV advertising for years, albeit in a more subtle way.
Truly revolutionary technologies, from the steam engine to the search engine, aren't solely about technical advancements. Their adoption hinges on when these advancements unlock significant improvements in unit economics. The steam engine succeeded because trains could carry more goods and people at a lower marginal cost than alternatives. Google dominated search less because of its vaunted Page Rank algorithm and more due to its innovative buildout of massive clusters of low-cost commodity servers, enabling it to crawl the internet more rapidly and cheaply than rivals.
The "steam engine" of today's TV advertising is not powered by some large language model, but instead by Vertical AI, which applies AI models to industry-specific challenges and data sets. In our case, Vertical AI creates efficiencies by analyzing massive datasets of ad impressions and consumer behaviors, discovering intent signals from a lot of noise, thereby enabling smarter, more profitable decisions in TV advertising.
Vertical AI Powers Data Collection and Analysis at an Unprecedented Scale
At EDO, we use Vertical AI to analyze every national ad impression on convergent TV. One of EDO’s AI models instantaneously recognizes over 2.3 million TV ad creatives at a 99% accuracy rate and connects those creatives to a taxonomy of over 32,000 brands and 115,000 products. Another EDO-developed AI model scans petabytes of data about consumer behavior related to those thousands of brands and products, discovering patterns that denote truly incremental changes driven by ad exposures versus normal baseline behavior. That has allowed us to automate, with minimal human oversight, the measurement of roughly 110 trillion TV ad impressions over the past 10 years for their impact on consumer behavior.
Trying to do that without Vertical AI would not be an option. The patterns are too complex for the human eye, and the tasks of manually updating our algorithms would be too slow, too error-prone, and too expensive if done without AI assistance. That’s a perfect example of new capabilities enabled by transformative unit economics.
Extraordinary Capabilities Create Transformative Cost Efficiencies
While it’s tempting to get caught up in the stunning capabilities of the latest Generative AI, marketers and media companies are most likely to grow their businesses by implementing Vertical AI to do the things they’re already doing — more, faster, and at a lower cost.
As marketers look to make smarter buys across an increasingly fragmented TV landscape and media companies look to control costs in a streaming-first world, these economies of scale will be increasingly vital to companies in our industry.
Those that have invested in building the necessary capabilities to capitalize on these opportunities will be the ones that continue growing through a period of rapid change and transformation — and those that haven’t will be left behind.