In 2007, Steve Jobs created a new market when he announced the iPhone. The emergence of the App Store a year later created a wave of opportunity for developers and brands eager to participate in a rapidly growing ecosystem. Some companies surged ahead while others faded as consumers learned to rely on pocket-sized computers for everyday tasks. Amid that volatility, one pattern became clear: Sustainable success in mobile depended on the ability to measure and optimize over time.
The companies that ultimately succeeded won because they have learned how to operate through uncertainty— using partial signals, proxy metrics, and experimentation to guide decisions long before perfect measurement existed.
Having lived through that period of chaos, I see clear echoes in today’s AI frenzy. Since ChatGPT-3 burst onto the scene, consumer AI in the form of large language models (LLMs) has captured the world’s attention— and it’s being adopted 15x faster than mobile was.
The leaders who recognize these patterns can apply the hard-won lessons from mobile to navigate what’s coming next.
We Flew Blind in the Early Days
In the very first days of mobile apps, we were blind to the most powerful acquisition channel: the App Store ranking page. Well before the practice of app store optimization techniques— before ranking lists even existed— companies had dedicated employees refreshing the App Store every hour to track changes in the top apps. The algorithms were invisible, and we guessed at optimization techniques for getting to the top of the charts.
While hitting refresh on the App Store may seem silly, it mirrors how brands and publishers approach AI today. Ranking and optimization techniques for chatbots remain opaque, and existing analytics ignore AI traffic as bots. The only insights companies have are simple “ChatGPT” referral codes on inbound traffic. Most companies are asking the LLM directly for their ranking— the exact same guesswork, just a different screen to refresh.
Measurement Followed the Money
As money flowed into the mobile ecosystem, rudimentary tools started to emerge. First came simple ranking tools— like SensorTower, which still exists today— then more advanced analytics to actually help with optimization. Mobile measurement partners like Branch now track attribution within the inscrutable App Store walls, and the likes of Mixpanel and Amplitude track and report on user behavior to help companies optimize and iterate product offerings.
Critically, these tools didn’t make mobile transparent. They gave marketers enough signal to test, iterate, and improve outcomes, even when causality remained unclear.
The same pattern is starting with consumer chatbots. The very first measurement solutions will help with ranking— Amplitude already has a tool to help you track your LLM rank. Eventually, new and existing vendors will build tools to track incoming AI referral traffic. Over time, these solutions will enable optimization to maximize the effectiveness of both consumer referrals and the chatbots themselves.
Chatbots won’t click on ads— advertising is based on interruption and capturing attention— but they can be influenced. A study shows that single-word adjustments can influence chatbot referrals by up to 20%. As influence becomes something teams can optimize, studies predict server-side agents will emerge to respond to AI-driven demand in real time, making chatbot optimization practical at scale.
Advertising Will Force AI to Grow Up
Not many remember that Facebook missed the first go-round of mobile. Early on, the company floundered with only a mobile web presence, and many, including myself, predicted its demise. We were very wrong. Facebook and others have exploited the continuing, massive growth of mobile and now find themselves atop the largest percentage of media ad spend.
Fast forward to today, very few businesses can afford to ignore mobile advertising. It has become a foundational part of how brands reach and convert customers. AI is following a similar trajectory. As usage grows and attention concentrates, it will become an unavoidable part of the marketing mix.
For this to happen at scale, we’ll need advertising on chatbots first. I’m one of many who view OpenAI’s advertising model as an inevitability. Google is introducing advertising this year, and others are well on their way. Like mobile, advertising will drive commerce into these emergent channels. The key to any successful advertising model is the ability to measure ad spend effectiveness. As this market grows, measurement methods will coalesce into a mature equilibrium— whether through shared pixels, CAPI-style feedback loops, or trusted third-party measurement platforms. These consumer-AI companies will need to offer tools and insights that brands can use to iterate and prove success before they scale spend.
The Lesson: Don’t Wait for Perfect Measurement
Facebook recognized it missed mobile apps and famously pivoted to become a dominant force in the mobile-first future. Meanwhile, BlackBerry rejected the concept of an app store, doubled down, and has been relegated to the dregs of history.
Today, as companies navigate the chaos of a new market and furiously cobble together rudimentary tools, they can rest assured that the efforts they make to embrace measurement now will determine whether they’re leading or catching up in 10 years. We’re entering a new era, and learning from the past will help us measure our steps in the future.
Adam Landis was the founder and CEO of AdLibertas, a mobile app data platform acquired by Branch in 2022. At Branch, he serves as Head of Strategic Growth, where his deep experience in mobile advertising and data helps foster innovation in products that increase the ability to measure marketing performance in an increasingly difficult ecosystem.
