👋 Good morning. For years the advice was build a great product, nail your brand, write better copy than the competition. That was the playbook. The problem is AI just turned all of it into table stakes. Great product? Commoditised. Strong brand voice? Replicable in an afternoon. Clever copy? Generated in seconds. None of those things are moats anymore. They're entry requirements. The real question is what you have that actually can't be copied - and most founders haven't thought hard enough about the answer.

THE PLAYBOOK

Somewhere in the last two years, a lot of what founders thought protected their business stopped being protection.

Fancy branding. Clever copy. Solid code. AI ate all of it and asked for seconds. The playing field didn't just level - it tilted. And a lot of businesses that felt safe suddenly don't.

But moats didn't disappear. They just got narrower. More specific. Harder to fake.

There are exactly 10 that still hold up in the AI era. Think of this as a checklist. If your business doesn't hit at least a few of these, you're not protected. You're just profitable for now.

Moat 01 / Proprietary Data Nobody Else Has

Spotify has a decade of your listening behaviour. Not just what you played - when you skipped, what you replayed, what you listened to at 2am on a Sunday. No competitor can buy that. No AI can generate it. If a well-funded rival can't recreate your dataset from scratch, every additional data point you collect widens the gap. AI tools are becoming commodities. Unique, compounding data is not. The question to ask: do you have data that gets more valuable the longer you collect it?

Moat 02 / Embedded in Daily Operations

NetSuite scores close to a 1 on this. Zendesk scores around 0.5. The difference is how deep inside daily operations the product lives. Once your tool becomes part of how a company runs every single day - touching payroll, client records, reporting, internal comms - ripping it out means broken integrations, retraining costs, and real downtime risk. That switching pain is a moat. Pure AI features can't easily puncture it because the pain of leaving isn't about features. It's about disruption.

Moat 03 / Regulatory

Ugly. Slow. Incredibly powerful. Coinbase holds money transmission licenses in 50 states. That took years and millions of dollars. A competitor with better technology still can't just show up and compete tomorrow. Multi-year contracts, capital requirements, and compliance burdens create walls that take years to climb - even with unlimited funding. Not the sexiest moat. One of the most durable.

Moat 04 / Distribution

Intuit trained every CPA to default to QuickBooks. Not because it's the best accounting software. Because they owned the channel. When your distribution is proprietary and exclusive, competitors can build a better product and still lose. The channel becomes the product. Control the path to the customer and what you're selling becomes almost secondary.

Moat 05 / Ecosystem

Shopify has 100K+ app developers locked in. When developers, agencies, and partners build their entire livelihoods on your platform, you've created gravitational pull that defends itself. Leaving doesn't just mean switching products. It means an entire ecosystem collapses around you. That's a moat that gets stronger the more people build on it.

Moat 06 / Network

DoorDash wins on courier density and restaurant liquidity. More couriers mean faster delivery. Faster delivery attracts more restaurants. More restaurants attract more customers. More customers attract more couriers. Network moats strengthen with every new participant. Once the flywheel is moving fast enough, it becomes nearly impossible to replicate from zero - because zero is exactly where every competitor has to start.

Moat 07 / Physical Infrastructure

Atoms are harder to displace than bits. Wherever you have physical infrastructure - warehouses, manufacturing, logistics - displacement is structurally harder. Software competitors can clone features overnight. They can't clone a fulfilment network built over a decade. Humanoid robots might close this gap eventually. But not yet. The real world still requires real assets.

Moat 08 / Scale

Amazon and TSMC made scale a permanent advantage. Their cost per unit dropped so far below what any new entrant could achieve that the economics alone became a defence. Here's the shift though - in the AI era, pure software companies can no longer claim scale as a moat. AI makes software production equally cheap for everyone. This moat now belongs to hyperscalers and physical-world companies only. If you're a software business claiming scale protection, check again.

Moat 09 / Speed

Execution velocity compounds. Daily shippers outrun annual planners. In the AI era, speed is structural. If your team ships faster than competitors can even scope the problem, your iteration cycles become your moat. Every day you're live and they're still planning, the gap widens. Speed doesn't just mean moving fast - it means building a culture where shipping is the default and planning is the exception.

Moat 10 / Talent

Anthropic and DeepMind have researchers no amount of money can buy. When your team has irreplaceable expertise, mission alignment, or domain knowledge that took years to accumulate, competitors can't simply hire their way to parity. The talent IS the technology. And the best people don't leave for a slightly better salary - they leave when the mission stops feeling worth it.

So where does your business sit?

Go through the list. Be honest. One strong moat is good. Two or three is a real business. Zero means you're running on momentum - and momentum runs out. The founders building for the long game in the AI era aren't the ones with the best product right now. They're the ones building something that gets harder to compete with over time. That's the whole game..

NUMBER OF THE DAY

47

The percentage of S&P 500 companies from 2000 that no longer exist. Most had what felt like moats at the time.

AROUND THE WEB

P.S. Know a founder/CEO who needs this? Send it their way. Consider it your good deed for the week.