In 2022, Craig Campbell made a decision that baffled many of his peers in the tech startup world. Fresh off selling his e-commerce tool business for Shopify merchants, venture capitalists were practically throwing money at him to dive into artificial intelligence. The AI boom was in full swing, and experienced founders with a proven exit were prime targets for blank checks. But Campbell had other plans. Instead of chasing the hype, he walked away from the river of investor money to start, of all things, a website. Not just any website – a platform dedicated to historical maps, an idea that seemed quaint in an era dominated by generative AI and massive language models.

Campbell’s background made the decision even more surprising. He had spent years as an engineer at Meta and had honed his entrepreneurial instincts building tools for Shopify businesses. When he sold his last venture in 2022, he could have easily launched an AI company and secured millions in funding. Yet he felt a pull toward something more personal and, in many ways, more authentic. His new creation, Past Maps, allows users to view historical maps of a specific region overlaid on a modern-day map. By adjusting a simple opacity slider, explorers can fade between the past and present, revealing how landscapes, rivers, and cities have transformed over decades or even centuries.

From Hobby to Business

The idea for Past Maps came from a very analog pursuit: metal detection. Campbell had taken up the hobby and found himself frustrated trying to locate old structures, trails, and potential artifact sites. He began building custom map tools to overlay vintage US Geological Survey maps onto current satellite imagery. The results were eye-opening. By pinpointing the modern-day locations of abandoned farmhouses, forgotten roads, or old mining claims, he could identify promising spots to search for relics. When he shared his tool on Reddit among fellow metal detection enthusiasts, the response was immediate and overwhelming. People wanted access to his creation.

That community feedback crystallized Campbell’s next venture. He decided to turn his personal project into a publicly accessible website. While the broader tech world was obsessed with chatbots, image generation, and autonomous agents, Campbell focused on something far simpler: giving people a way to see how their local area used to look. The maps come from freely available public sources like the USGS, but the interface and tools to explore them seamlessly were built entirely by him. The result is a resource that appeals to historians, genealogists, real estate researchers, outdoor enthusiasts, and the merely curious.

Growth Through Organic Search

Past Maps’ growth story is a testament to the enduring power of search engine optimization. Campbell saw early on that his pages were climbing Google’s rankings for queries about specific locations and historical features. Someone searching for “old maps of abandoned mine sites in Nevada” or “historical map of the Duwamish River” would find Past Maps near the top of the results. This organic traffic became the lifeblood of his business. In year one, the site averaged 20,000 active users per month. By year three, that number had surged to over 300,000. Campbell attributes this success to careful tagging and structuring of his map pages, making the data easy for Google to index and surface.

“This is how the web is supposed to work,” Campbell reflects. “This is actually the old school web. It is alive and well, but only in these really, really small niches.” His approach runs counter to the prevailing wisdom that the open web is dying under the weight of AI-generated content and search engine changes. By focusing on a specific, high-quality, and useful product, he has carved out a sustainable niche. He doesn’t rely on display advertising or venture capital; instead, he offers a freemium subscription model. Users can try the service for free, but deeper exploration costs $9 per week or $52 per year. This subscription revenue provides stable income, insulating him from the volatility of ad markets and the grip of Google’s ad monopoly – which the US Department of Justice ruled as an illegal monopoly in 2025.

Embracing AI on His Own Terms

Ironically, Campbell has fully embraced artificial intelligence to run his business – but on his own terms. Rather than building an AI startup, he uses AI as a tool to streamline operations. Customer service is a prime example. He used to spend one to two hours a day personally responding to every support request, typing lengthy emails instead of sending canned responses. Now, a local language model runs on his desktop at home. It checks his email once every hour, filters out spam and marketing messages, identifies urgent requests, and drafts replies. When a customer asks for a refund, the AI even initiates the refund process through Stripe and cancels the subscription, then pings Campbell for approval. This has slashed his daily customer service time from hours to about ten minutes.

Campbell is also applying AI to a more complex problem: optical character recognition for old maps. Historical cartography presents unique challenges. Labels often curve along rivers, use irregular spacing, and crowd together in dense areas. Off-the-shelf OCR tools typically fail to parse these intricate layouts. Campbell discovered that modern large language models, when used with reasoning steps, perform far better. But he emphasizes that it’s not a matter of simply prompting an AI to extract text. “You have to still bring that human spark into the mix,” he says. He combines his own intuition and decades of experience with the LLM’s capabilities, experimenting and tweaking to get the best results. This hybrid approach – human creativity augmented by machine efficiency – exemplifies his philosophy.

The Old School Web Lives On

Campbell’s story challenges the narrative that the only way to succeed in tech is to chase the latest trend. He walked away from an AI gold rush that has minted billion-dollar valuations overnight, and instead built a modest but profitable business that serves a passionate community. He makes roughly the same income as a mid-level engineer at Meta, but he runs his own company, sets his own schedule, and works on something he genuinely loves. His customers range from genealogy researchers tracing family history to hobbyists mapping old oil wells. One daily user simply enjoys watching the Duwamish River straighten out over time. For Campbell, that kind of engagement is richer than any venture capital return.

The tools and tactics he uses reflect a blend of old and new. He relies on organic search traffic, just like publishers did a decade ago. But he also uses AI agents for customer service and OCR, and he runs a subscription model that guarantees recurring revenue. The foundation, however, is thoroughly human: a passion for history and exploration, a willingness to share knowledge, and a focus on building something genuinely useful. Campbell admits that cartographers can be “assholes” when it comes to laying out labels on old maps, but that human imperfection is exactly what makes the work interesting. In the end, his bet on the old school web is paying off because he understands that the web, at its best, is about connecting people with information and each other – and that never goes out of style.

Source: The Verge News