You spent a year building a feature. Someone just replicated it in a day using AI.

This isn't hypothetical. Roei Samuel is watching it happen in real-time. As founder of Connected - a marketplace helping 5,700 fractionals work with scale-ups - he's spinning up products daily that took his team a year to build in 2020.

His conclusion? Unless you're building quantum computing or genuine deep tech, your technology moat is dead. AI killed it.

Here's what makes this different:

Roei isn't being dramatic. He built and sold a media company that scaled to 9 million monthly users, worked with the Premier League, NBA, and NFL, and joined the senior management team of a PLC at 26. He's seen what creates lasting value.

And his take is clear: product doesn't create defensibility anymore. Network effects do. When every feature can be replicated in weeks, the only moat is how your users create value for each other - and how hard that is to reproduce.

You'll learn:

Why AI just eliminated technology moats. What took a year to build in 2020 now takes a day. Your 10% optimization? It'll be copied in months. The only defensible businesses are built on network effects and brand—mechanisms competitors can't easily replicate.

What network effects actually mean. It's when one user's participation improves the experience for all users. Could be data (more users = better matching), could be multi-sided supply (Roei's fractionals average 3 roles each, solving the liquidity problem), could be customers becoming promoters.

How most businesses can access network effects. You don't need to be a marketplace. If you're good at turning customers into promoters—testimonials, LinkedIn posts, word-of-mouth - you're building network effects. The best businesses layer multiple mechanisms.

Why hiring full-time is becoming the last resort. Smart founders now think: (1) What can I automate? (2) What requires a fractional specialist? (3) Only then, do I need full-time? This isn't theory - startups on Connected average 3.7 fractionals each.

How to solve marketplace liquidity problems when starting. Don't try to build both sides simultaneously - it kills companies. Use SaaS-enabled networks: give one side free tools (dashboards, benchmarking) while you populate the other side. Roei did this launching Connected in the US.

Why you shouldn't scale until you nail cohort metrics. Don't worry about growth. Start with 150-200 users. Measure daily active usage, retention, behaviors that drive engagement. Roei invested in Lapse based purely on cohort analysis—they raised £8M seed, then £30M Series A from Greylock. Zero monetization. Just strong network effect metrics.

How to identify your specialty if going fractional. Lean into where you deliver tangible results fastest. Not what you're best at. Not what's most fun. Where can you prove ROI in 6 months? That's your first case study. That's how you build track record.

Why living out of alignment destroys everything. Roei's real mission isn't about fractional work - it's about helping people live authentically.

The reality check:

This isn't anti-product. Product still matters. But product alone won't save you when competitors can replicate features in weeks. Network effects create the compounding advantages that turn good products into defensible businesses.

If you're building a business in 2026 and you haven't thought about network effects, you're building on sand. AI just raised the stakes.

One action: Listen to the end for Roei's hiring sequence every founder should use immediately.

More from James:

Connect with James on LinkedIn or at peer-effect.com