customers.ai

Larry Kim had already built and sold WordStream, an AdWords and Facebook management tool, for $200 million in 2018. But he wasn't done building. In 2018, he launched MobileMonkey, a chatbot automation platform built on Facebook's ecosystem. The growth was explosive—zero to $1M ARR in under a year. By 2019, Larry thought he'd cracked the code again.

But MobileMonkey hit a wall. The problem: being a Facebook partner meant playing by Facebook's rules. "When you build in an ecosystem like a Facebook partner, you're not really master of your domain," Larry explained. "You're kind of at the whim of some product manager at Facebook who decides to kill some functionality." Facebook made policy changes that made the business unviable. The pandemic hit, SMB customers evaporated, and revenue flatlined around $1M. By 2020-2021, Larry's cash balance had dropped to "a couple hundred thousand dollars."

He faced a choice: pour his own wealth into a struggling business, or let the market reprice it in a painful down round. Instead, he chose a third path: non-dilutive capital.

Larry realized he had something valuable: the data infrastructure and hundreds of thousands of websites running MobileMonkey code. He decided to pivot from social messaging automation to B2C sales outreach. The new thesis: help companies identify website visitors by email and conduct automated sales outreach—a completely different use case, but using the same underlying tech stack.

The key innovation was building a proprietary large language model (LLM) trained on MobileMonkey's unique dataset of consumer behavior and website interactions. Using device fingerprinting, IP tracking, and browser signals (mouse type, plugins, screen resolution, language settings), combined with AI inference, customers.ai could tell companies: "This person visited your site, their email is [X], and here's what they're interested in."

In 2020-2021, Larry took a $400K loan from FounderPass (a non-dilutive capital platform) and matched it with $400K of his own capital. This gave him roughly a year of runway to prove the new use case without selling equity. He praised the speed: "Five business days from start to end." The downside was the interest rate—"credit card rates"—but he saw it as a no-brainer given the alternative.

With that runway, Larry and his team fully developed the customers.ai product, moved upmarket, and demonstrated strong traction. By the end of the pivot period, he'd grown the business back to over $2M ARR and attracted serious Series A interest.

Customers.ai reached $2M+ ARR last year and is targeting "mid to high single digit millions" (approximately $5-7M) by year-end. The team has grown to 40 people, with 15 engineers. Several customers are now paying over $100K annually. The business is remarkably efficient: burn-to-ARR ratio is below 1.0, meaning Larry is adding more in ARR than he's burning in cash. They have over two years of runway post-Series A.

Larry's bigger vision mirrors his time at WordStream: provide both the tools and the data. At WordStream, customers would spend $500 on the tool but $10,000 on ad spend—a small wallet share. With customers.ai, Larry is providing the leads, the data, and the outreach automation in one platform. "We're providing the tools and also the lead data, which is a much more interesting value prop," he said.