There has been no shortage of articles, Tweets, podcast soundbites, and so on talking about moats in the age of AI.
The ‘hot takes’ vary wildly:
The moats are the same as they’ve always been.
There are no moats anymore.
There never were any moats to begin with.
Capital is the only moat.
As ever, the most hysterical soundbites oversimplify. Everything requires context.
The TL;DR of our view for Consumer is that many of the ‘old’ moats (e.g. brand, network effects, scale economies) are as relevant today as they were decades ago when Warren Buffett started using the phrase in his shareholder letters.
What’s changed post-AI is how much harder it is to establish an initial foothold than it once was. No ‘moat’ has ever come out-of-the-box on day 1 – it takes time. But time is the construct that AI is collapsing most of all.
The challenge breaks down into a couple of components:
The pace of duplication and intensity of competition is higher than ever. Your competitors are filling in the moat while you’re still digging. The timeframe you have to get to escape velocity – for those atomic networks to form, for a brand to get established, and so on – is collapsing.
The incumbents are focused on the ‘new thing’ unlike ever before. The largest, most aggressive and most sophisticated companies in history are in an existential race to own Consumer AI. They are leveraging their “unfair advantages” (scale, users, tech, compute) to own this space too. Even if a start-up gets a foothold, how can it avoid becoming collateral damage in the next feature release of the incumbents?
Here are some evolving perspectives on how to fight it.
Escape velocity
The collapsing time frame for duplicates and copycats is a real issue for Consumer founders and investors. But in our view the pendulum has swung too far when we hear soundbites that capital and distribution are the “new” - or worse, the “only” - moats.
On the capital side, the trope is that a multi-stage VC fund can ‘Kingmake’ a given player in a category. The logic is that the key resources any start-up needs – (more) capital, talent and customers – flock to the adorned ‘winner’ and effectively crowd out would-be competitors who don’t dare to get started.
In theory, it makes sense, but there are a couple of obvious potential flaws.
First, capital in-and-of-itself isn’t enough. For it to be effective, that capital has to be deployed into (1) building a value proposition that resonates at scale – a “castle” that is worth protecting; and (2) the thoughtful construction and deepening of actual moats that are fundamental to the business. History is awash with examples of companies that raised a huge amount of money but failed to convert that (temporary) advantage into anything enduring because the business model or product itself didn’t work (remember WeWork?).
Second, it is highly levered to the prevailing capital markets conditions at the time. Like most momentum-related strategies, in boom eras, it works brilliantly – there’s always another investor ready to shovel more money in – but it can unravel very, very quickly when the market turns.
Finally, it relies on ever-accelerating valuations and near-perfect execution. Any slowing of the momentum creates incentive challenges (broken cap tables, underwater option plans, the loss of ‘heat’) from which it can be very hard to recover. Getting out from underneath the weight of huge expectations can be lethal.
At best, capital gives you the materials to build something worth protecting, a bunch of big spades, and a window of time to start digging a moat. But it’s no guarantee.
On distribution, the thinking goes that in a world of product parity and instant duplication, only the ability to cut through the noise and get to end users matters.
But similar to the capital discussion, distribution alone isn’t a moat in the long run.
Getting users in the door is a great start, but it needs to be converted into something more valuable and enduring. Awareness and great product marketing gets you downloads, but product value and depth gets you loyalty – even more so in Consumer where the propensity and speed of switching to the ‘next hot thing’ is so high. Roy Lee, the Cluely founder, was the poster child for viral marketing across 2025 and early 2026, creating intentionally controversial marketing stunts to drive awareness and downloads. But when was the last time you met a Cluely user?
What can work well is using a near-term distribution advantage to enable the formation of a moat: for example, to kickstart an initial network effect that can then be built upon, now that the cold start problem has been unlocked. That’s why we completely buy into the value of having strong GTM expertise in the founding or early team of Consumer companies. But again, that makes it a spade – a tool with which to start digging a moat – rather than a moat itself.
The point is that both capital and distribution can be huge assets in the near-term process of building escape velocity – indeed, they are probably the most impactful accelerant, all else equal. But they aren’t enough: they have to be the opening chapter in a more complex narrative that translates a short-term edge into more fundamental and enduring long-term unfair advantages.
Won’t Claude do this?
The second challenge is how you avoid being steamrolled by a major feature release from global platforms. Their capital and scale can fill your moat in an instant.
For that, we think there are a few ways Consumer founders can mitigate the risk.
First, don’t stay in shallow product ‘gaps’ for any longer than absolutely necessary. Even if you hit great traction with an initial wedge that solves a temporary shortfall in existing products, the power of existing distribution and eventual bundling from incumbents is a real and existential threat. Rapidly build surface area, product depth and actual moats (e.g. network effects) to avoid being featurised.
Second, embrace messiness. Things that exist purely in the internet and software arena are much harder to defend vs. global platforms. Embrace complexity – of messy ‘real-world’ service delivery, of physical networks and assets, of vertical-specific integrations – to create a superior product experience that horizontal platforms won’t be able to match. DoorDash and Amazon are good examples where messy ‘real-world’ stuff confers an incredible moat vs. software-only competitors.
Third, lean into heterogeneity: of regulation, of cultures, of commercial structures, of use case specifications. West Coast incumbents thrive on global scalability: build it in California, ship it globally. We think founders in Europe should consider doing the opposite and focus initially on areas where that just doesn’t work. Yes, that potentially means a smaller market in the short-term, but local European markets have proven to be a large enough seeding ground to create hugely valuable businesses (e.g. Revolut) that can then amass the resources to take on global markets (e.g. Spotify).
Finally, solve a really hard problem that unlocks a fundamental new data advantage. Sava in our portfolio is a great example: hardcore hardware, biochemistry and software combining to unlock the value of biometric data. The ‘software x biology’ interface isn’t something the AI labs will do. Building the gateway to something scarce opens up enormous incremental use cases for the next generation of AI applications.
It never ends
The last point to underscore is the importance of deepening and widening moats and layering on new moats, over time. This is a concept that comes across strongly in Buffett’s original letters but tends to be forgotten in much of the commentary today.
In this context, we draw a distinction between static moats and dynamic moats.
A static moat is a hard initial step to ‘get over’, but once accomplished, has a relatively short half-life. A regulatory approval is a static moat. So is a specific technology breakthrough. Much like the capital and distribution discussion, a static moat buys time but does not – by itself – lead to long-term enduring defensibility. If the castle you’ve built is valuable enough, people (i.e. competitors) get started quickly on filling in moats that don’t evolve. A true moat is not “one and done”.
A dynamic moat is different. It compounds, grows and evolves over time – more of a system and process than a specific ‘thing’ – in the pursuit of new advantages to layer onto the business. Spotify is a great example: its initial moat was its licensing deals with record labels that allowed it to offer a legal streaming service at scale. That has commoditised over time – you can access the same content via Apple Music or Tidal or Deezer – but in the meantime they have built new moats around brand, personalisation, social features and creator-audience network effects that combine to build a fortress around their castle.
A moat is not a tick-box, it’s an evolving imperative that requires constant new thinking and strategic leadership alongside sensible sequencing and layering over time. A verb, not a noun. It’s that kind of thinking we love to see in Founders we speak to.
Pulling it together
In some ways, it’s easier than ever to build a company for all the reasons everyone always talks about.
But at the same time, that makes it harder than ever to build a company that endures. We’re still refining our thinking on what it takes in a post-AI world to build truly incredible moats and would love to speak to Founders and other investors with a strong perspective so we can continue to evolve our thinking.
If you’re building and/or investing in the Consumer technology space, get in touch at vc@true.global.