by Jonah Goodhart
Digital advertising turns 30 this year. In those three decades, we have gone from $0 to $600 billion in annual global digital advertising spending. More than two-thirds of all money spent on advertising itself now goes to digital formats and channels. Of that $600 billion, more than two-thirds of that goes to five companies – Google, Meta, Amazon, ByteDance (owner of TikTok), and Alibaba. Those are remarkable facts. What is also remarkable is there is still no consistent way for brands to judge the success of their digital spend. Sales, brand health, and market share may stand as ultimate goals, but seamlessly and independently linking these to actionable metrics is a challenge the industry has yet to overcome. In a landscape where advertising’s role in brand growth is ever-expanding, perfecting the art of measurement is more critical than ever.
Measurement is not devoid of innovation — quite the contrary. There are more than 200 measurement companies, commanding over $20 billion in annual revenue, offering services to help marketers answer pivotal questions about reach, frequency, audience type, contextual environments, attention, emotion, carbon footprint, brand lift and even purchase-based outcomes. However, this fragmented approach has created significant barriers, with independent systems plus difficulties in integrating data from dominant platforms. The rise of AI, however, signals a transformative era where the data powering AI models will open the door to unprecedented innovations and optimizations.
To grasp the importance of measurement, let’s consider its far-reaching implications. Advertising not only powers the media landscape but also drives societal innovation, shaping culture, economics, and beyond. Last year alone, Google’s ad revenue, a staggering $225 billion, not only underwrote its suite of leading services but also propelled innovation across the digital realm — including leading generative AI capabilities such as Bard. Regardless of that innovation, Google is in the business of selling ads and marketers must have independent measurement capabilities when buying through Google.
Meta of course enables billions of users to use Instagram, Facebook, WhatsApp, Threads, and more by generating more than $100 billion in annual digital advertising revenue. More than 95% of Meta’s total revenue is from advertising, so their incentives are clear — and the need for independent measurement is also clear. It is also a simple fact that Meta’s creation of the open-source AI platform Llama stands as a testament to the innovative potential of ad revenue streams.
Amazon used to be the classic scaled non-advertising example on the internet. After all, Amazon generates hundreds of billions in e-commerce revenue and around $80 billion in AWS cloud services. It may surprise some though that not only is Amazon‘s $40+ billion advertising business its fastest growing unit, it is also the most profitable one (more profitable than AWS, which itself is used by most ad companies!).
Netflix famously said they will never sell ads. Netflix now sells ads and is focused on advertising as a core part of their future success. X (formerly Twitter) has infamously announced it will fail without ad spend. Zillow is the most popular residential real estate application in the U.S. More than 70% of their revenue comes from ads. Apple not only generates billions in direct advertising revenue via their App Store search ads, they also apparently sell the “default search engine” slot on Safari to Google, generating a reported $18 billion in annual revenue for Apple. This works for Google because they monetize those searches with ads. The fact that four of the top five U.S. companies by market cap are major advertising players is a clear indication of the industry’s foundational role in the economy.
This begs the question: Why, in this digital age, is there still no universal measurement framework for brand advertising success? The answer lies in the intricacy of digital itself. Unlike the consistency found in traditional media, digital advertising is a chameleon, constantly evolving with limitless variations. These complexities have compounded challenges such as ad fraud, brand safety, and the gamification of metrics, rendering traditional concepts of reach somewhat meaningless.
Yet, in the digital-first world where streaming has surpassed cable, and online platforms outpace traditional media, hope gleams on the horizon. The collective intelligence of hundreds of measurement companies presents an opportunity to converge in a powerful new approach. By harnessing the diverse strengths of a broad marketplace, a suite of comprehensive measurement capabilities has the potential to recalibrate creative and media to better align with outcomes brands seek.
The data from these measurement capabilities is where this gets quite interesting — serving as a critical input to the AI-powered systems of the future. AI is poised to revolutionize advertising measurement, bringing unprecedented precision and efficiency to our world. One of AI’s near-term potential applications lies in automating creative and media ad decisioning, a process that hinges on the availability of extensive training datasets. By consolidating diverse measurement data into a unified, harmonized dataset, brands gain access to predictive decisioning capabilities. These AI-driven systems can optimize advertising strategies on a quantum scale, far surpassing the limitations of traditional machine learning techniques.
One of the most transformative use cases of AI in advertising is the generation of creative content. Advertising is about storytelling — presenting products and services to consumers that resonate on practical and emotional levels. AI’s capacity to predict and assemble effective combinations of visual and video elements marks a significant leap forward. Not only will this reduce the costs associated with creative production, but it will also enhance performance by leveraging AI’s dual capabilities in decision making and content creation. Imagine a scenario where AI not only predicts that a certain mix of visuals will excel but then proceeds to construct and test these elements autonomously.
At my company, Mobian, we are on a mission to democratize measurement, recognizing that the integration of disparate measurement signals is fundamental to developing robust AI-powered decisioning systems. Our ultimate goal is to empower brands to harness their full potential, ensuring that advertising strategies are not only data-driven but also dynamically responsive to an ever-evolving landscape.
Envision an industry that provides true end-to-end, integrated measurement that can leverage the full power of AI and evolve as the world evolves. We believe this world is not far off from today, but it will take collective effort. We believe it is worth it and that we have an obligation as an industry to make this happen. It is time that we get measurement right.
-Jonah Goodhart is the co-founder of Mobian and Montauk Labs. Previously, Jonah was the co-founder and CEO of Moat (acquired by Oracle) and founding investor in Right Media (acquired by Yahoo). Jonah is also a partner in WGI Group, an investment group with over 100 investments in the technology space
