These days, I feel like “sprawling” is the only way to describe the tech industry. Big Tech is a hungry beast, devouring opportunities for hockey stick growth, advancement, and profits. When I trade ideas with others in tech, it seems like the only concepts deemed “viable” are the ones that could be the next billion-dollar unicorn… or be acquired by one. But I see the seeds for change quietly being planted in FAANG’s backyard.  Ironically, it’s the AI surge that’s ushering in an era of “Craft Tech”.

Back in 1979, President Carter signed a law that legalized the practice of homebrewing beer. The moment unlocked a movement, as overnight a generation of new brewers gained the ability to experiment with the craft. These burgeoning practitioners launched a wave of small, focused breweries that ultimately came to be called the Craft Beer industry. They prized quality, local flavors, and community — ideals that tapped into a growing cultural shift. In 1979 there were under a hundred total breweries in the US, and a handful of them were considered “craft”. By the mid-90s the number of craft breweries alone had risen to ~500^[1]. Last year that tally stood at 9,683^[2].

One of the reasons Craft Beer was able to take off was its accessibility. Once it became legal, anyone could assemble a homebrew kit for a couple hundred bucks and start tinkering. Recipes were easily shared and a cottage industry of specialty ingredients providers popped up, making it painless (and cheap) for brewers to experiment with flavors and styles. But, until now, building software has been virtually inaccessible — requiring deep technical expertise beyond the reach of most home tinkerers. The No-Code family of products (like Bubble, Zapier, and Wix) initially tried to bridge this gap, but with limitations. A few years ago I built an app prototype on Bubble, but I still had to pay an engineer to write 10 lines of custom code to make my core feature work.  Fast forward to today, and AI-enhanced code editors like Cursor and Replit are the new expression of No-Code. I can describe what I want, and the tools will effectively build it for me in code — giving me virtually the same freedom of expression as an app developer.

What excites me about this new accessibility is that it changes the fundamental economics of tech. Building software has typically required hiring highly-trained teams — roughly $1m in annual salary for the average 5-person software team. That kind of financial outlay is beyond the scope of home experimentation (not to mention small business loans), which means venture funding and its associated growth pressure. But AI “no-code” editors enable anyone with an idea to tinker for a $20 monthly subscription.

To be clear, this doesn’t mean that every homebrewed app will be good. Just like your neighbor’s way-too-hoppy “IPA”, there will be terrible products. But the ability to cheaply experiment will enable a whole new generation of software builders. A lot of them will fail, but this is a numbers game — more experimentation will ultimately enable more great products.

With reduced financial pressures and increased creative freedom, creators no longer need to myopically focus on massive markets. When you don’t have to justify the potential for a $1B valuation to build a product, you’ve unlocked a whole new strata of user needs that haven’t been addressed by software to date. You can build focused, thoughtful products that are specially tailored to niche users — much like craft breweries prioritized local flavors and tastes.

But what do these opportunities look like? I’m excited by the fact that we don’t really know yet, but there are some places we can look for inspiration. One is the “1000 True Fans”^[3] wave of content businesses that came in the wake of publishing platforms like Wordpress, YouTube, and Substack. They lowered the barrier to content creation and allowed creators to build viable businesses by laser-focusing on niches like parenting data^[4], pop culture deep dives^[5], and product skills^[6]. There are likely solutions to key daily problems that could convert these fans into “1000 True Users”.

Another place we can look is niche power users. Readwise^[7] is one of my favorite examples of this — they successfully bootstrapped a company by deeply serving nonfiction readers who wanted digital solutions (hi, it’s me). Ravelry^[8] also became a self-funded success story by becoming the go-to platform for sharing and selling knitting patterns. With a small team (1 coder), they have become the undisputed home of the online knitting community (hi, it’s my wife).

We can also take inspiration from previous unbundling cycles. One of the classic tech biz articles^[9] discussed the unbundling of Craigslist — outlining how different sections of the do-everything Craigslist homepage were turned into more focused platforms like Zillow and Indeed. This cycle is primed to repeat itself with the proliferation of large, sprawling SaaS platforms (especially in B2B). There are likely opportunities to carve out viable small tech products by specifically serving the commerce or operational needs of specialty small businesses — craft breweries, perhaps?

Taken against the backdrop of a growing distrust of Big Tech, these fundamental changes in software creation are opening up an exciting new time for entrepreneurs. It’s a specific type of person who aspires to run a billion-dollar company, and there’s a large swath of entrepreneurially-minded creators who want to make something meaningfully useful, not massive. Focusing on the craft, and building companies around niche software products that “do one thing well” is now a viable path to being a thriving entrepreneur. Just like homebrewing helped nurture a diverse culinary landscape, Craft Tech has the potential to open up a rich ecosystem of digital products designed to serve specific humans over mass appeal.