TechFlowPost on Sept. 11 published a translated essay from Not Boring arguing that the sharpest early weapon available to startups competing with incumbents is not merely being better. It is building a business model that incumbents cannot copy without hurting themselves. The essay calls that dynamic counter-positioning.
The piece places the idea inside a broader discussion about strategy and moats. Its central claim is straightforward: startups do not get to ignore strategy just because they are young. Before they develop longer-lasting forms of power such as network effects, scale economies or brand, counter-positioning can buy them time.
What counter-positioning means
The article cites Lindy CEO Flo Crivello, writing in Mind the Moat, an introduction to Hamilton Helmer’s 7 Powers. He defines counter-positioning as building a business model in a way that prevents incumbents from competing effectively because doing so would conflict with their own interests.
That is the core reason the author values it so highly. Large companies usually have bigger businesses, more resources and stronger distribution. A startup that plays the incumbent’s game on the incumbent’s terms is often trapped from the start. Counter-positioning changes the game. It does not automatically make the startup superior. It makes the incumbent’s existing strengths act like shackles.
The author revisits a comparison first used in December 2020 while writing about Ramp: counter-positioning is like the Five Point Palm Exploding Heart Technique from Kill Bill. A startup can strike an incumbent in a way that does not look fatal at first. Then the moment the incumbent steps toward the startup’s model, it collapses.
Ramp: make money by helping customers spend less
Traditional corporate card providers encourage customers to spend more through points and rewards. Ramp took the opposite path. Its business model is built around helping customers spend less.
That is why the author sees Ramp as a clean counter-positioning example. For incumbent card issuers, lower customer spending means lower revenue, which makes following Ramp’s model deeply unattractive. The essay notes that other startup challengers such as Divvy and Brex played closer to the old rules by using rewards and incentives to drive spend. They still had successful outcomes: Bill.com acquired Divvy for $2.5 billion, and Capital One acquired Brex for $5.15 billion.
Ramp, though, followed a different trajectory. The article says Ramp is now valued at $44 billion and is rolling out more products meant to save customers time and money, including a model router. The lesson attached to that example is blunt: if you want to challenge the king, do not bring better technology just to play the king’s own game.
Base Power: do not sell the battery, sell the electricity
The essay says the author was reminded of counter-positioning after hearing Base Power Company CEO Zach Dell on David Senra’s podcast. At 16:18 in the episode, while discussing how he and Justin settled on a business model, Dell framed the issue directly: if you want to challenge an incumbent, the best way is to have a counter-positioned business model.
He used home batteries to make the point. If a startup enters the market saying it has the best home battery and it sells it for 10% less than established manufacturers, those manufacturers can copy the product, start a price war and squeeze the entrant out.
But if the startup says it is not selling batteries at all, it is selling electricity, the frame changes. Base installs a battery in the home and offers the service at one-twentieth or even one-fortieth of the price of full ownership, according to the quote included in the piece. At that point, incumbents cannot respond without overhauling their entire business model, something Dell says is particularly hard for a public company.
The essay is careful not to overstate the durability of that advantage. Base cannot rely on counter-positioning forever. To protect margins over time, it will need other forms of power, including scale economies, cornered resources, switching costs, brand and possibly network effects and process power. Counter-positioning gets the challenge started. It does not finish the job.
Helmer’s point: powerful at takeoff, incomplete over the long run
The article says Ben and David of the Acquired podcast return to this theme often, describing counter-positioning as a form of power that matters especially in the takeoff stage. Hamilton Helmer himself has said the same thing on the show: it is the only incomplete source of power, and if a company wants true durability, it needs another source as well.
The author condenses that into a simpler line. Counter-positioning is there for a good time, not a long time. It behaves like a trickster’s power. It buys a startup time against incumbents whose own positioning makes them vulnerable, but it does not necessarily protect that startup from the next startup that comes after it.
That is why the window matters so much. Once the opening appears, the company still has to build a real moat.
Dell, Android and Amazon: incumbents trapped by their own structure
Helmer has used Dell as an illustration. Dell counter-positioned Compaq because Compaq’s dealer channel made direct sales painful. But direct sales eventually became available to everyone. By then, Dell had used the opening to build scale economies around its direct model and just-in-time operations.
The same logic appears in the Android example. Google could offer Android for free because the parent company made money from search. More mobile search meant more revenue. Nokia, by contrast, needed to make money from the mobile operating system itself. Giving it away would have cut directly into its own business.
The article turns that into a practical question: can challenger A do X because it makes money in a completely different way from incumbent B?
Amazon and Barnes & Noble present another version of the same problem. Amazon was born on the internet and built around direct-to-consumer fulfillment. Barnes & Noble optimized around physical stores. For Barnes & Noble, fully embracing Amazon’s model would have meant reallocating capital, shifting management attention, rebuilding distribution infrastructure and accepting lower short-term profits from existing stores. In the phrasing quoted in the essay, Amazon’s model was simply less profitable for them than the stores they already had. That is why Amazon counter-positioned rivals whose cost structures were designed for physical retail.
The article offers a second way to think about this. The more money incumbents have sunk into existing infrastructure, the harder adaptation becomes.
Somos Internet and Microsoft: having no baggage is a form of freedom
The essay stresses that this strategy is not reserved for companies with Google-sized resources. One of the author’s favorite examples is Somos Internet. As described in the article, Somos designed a new network architecture and built its own hardware to run it. Incumbent telecom operators could not respond easily because doing so would require undoing billions of dollars in capital expenditures tied to older network configurations and third-party hardware.
Those companies were also constrained by vendor dependence, limited technical flexibility and debt used to finance network buildouts. That made it hard to lower prices to attack Somos, let alone fund an uncertain new build.
The article then moves to the PC era. When Microsoft challenged IBM, Ben’s explanation was that Microsoft had almost no baggage. IBM sold integrated computing systems through one of the strongest enterprise sales organizations in the industry. To launch the PC quickly, IBM broke from its normal pattern by using off-the-shelf components, Intel processors and an operating system supplied by Microsoft. Microsoft, however, did not give IBM an exclusive DOS license. It licensed DOS non-exclusively because it wanted PCs to become as cheap and widespread as possible so its software could reach as many desktops as possible.
The point is not that Microsoft wrote better software than IBM. The point is that IBM needed software to sell high-margin machines, while Microsoft benefited as hardware became commoditized. Once the hardware layer turned into a commodity, every hardware provider had to remain compatible with the software layer. In the line quoted in the essay, Microsoft only had to send bits to become the integrating core of the ecosystem.
This, the author says, is an especially aggressive form of counter-positioning: the entrant benefits not just from taking share from the incumbent’s profit pool, but from destroying the incumbent’s profit pool altogether.
Facebook versus MySpace: choose the opposite growth logic
The social media section looks back at Facebook’s earliest years. When Facebook launched in 2004, it faced MySpace, which already had 1 million users. The timeline in the article says MySpace launched in 2003, reached 1 million users in February 2004, surpassed Friendster the next month and grew fivefold to 5 million by November.
Facebook’s counter-positioning move was not to chase the same growth path. It argued, in effect, that cramming a huge number of users into one network from the beginning was not the best way to build a social network. Starting smaller, seeding the network with Harvard students and expanding outward gradually could be stronger in the beginning.
If you were MySpace, growth looked like success. Why would you stop to imitate a slower model built for a smaller, more concentrated network? The article argues that Facebook used that window to form an intensely engaged community before opening up and riding the next phase of network effects.
In AI assistants, “better” is not the same as protected
The essay eventually brings the framework into the AI assistant market. It cites a post from a16z partner Josh Elman earlier in the week: the new moats are the same as the old moats, and every few years people fall in love with shiny new technology and forget the basic physics of consumer software. The author extends the point to business strategy more broadly.
Instinct is described as one of the most talked-about assistant products right now. The product experience is strong and, in the author’s view, better than earlier assistant products. But the essay states the distinction plainly: better is not counter-positioning. It quotes Michael Porter’s line that operational effectiveness means performing similar activities better than rivals, while competitive strategy is about being different.
That distinction matters because product quality alone is not a moat. A moat exists when incumbents can see what makes the startup better yet still cannot copy it without damaging their existing business. If that constraint is absent, the startup may have an advantage, but not a barrier.
The article includes several specifics on Instinct. By the end of August, it had raised a $350 million Series B at a $2.5 billion valuation. One theory mentioned in the essay is that the company raised that much so it could keep offering the product for free while trying to reach a new business model, perhaps transaction fees, and overwhelm competitors along the way.
The author’s response is direct: better is not a moat, and drowning Meta’s money machine is a difficult bet.
Meta’s Muse: copying the model does not hurt the incumbent
The article says Meta this week released its own assistant, called Muse. Some early reviews have said it is even better than Instinct. The author says that is not the main point.
The real issue is that copying Instinct’s model and product is not painful for Meta. It may actually fit Meta extremely well. The essay lists several reasons. Meta knows an enormous amount about its users and can recommend things users may not even know they want, which helps with the cold-start problem. The article also cites Ben Thompson’s point that Meta can give each user “a virtual machine with 8GB of memory and 8GB of storage” — a real computer, offered first in the United States and eventually more broadly.
Meta can also integrate the assistant directly into WhatsApp, into FB Marketplace and into Meta Ray-Bans, while training it in its own large data centers. The article’s conclusion is that Meta can keep offering the service for free for a very long time while drawing on network effects, scale economies, brand and any other moat it needs, because it already went through its own counter-positioning period earlier and has since built enduring defenses.
The author notes that Mark Zuckerberg has failed before when trying to copy or launch new products. Threads, for example, did not cure the author’s habit of using Twitter. But Twitter has network effects. It has a moat. That, in the article’s framing, is the whole point.
Being acquired is not a strategy
The essay says many startups still believe that in a new technology cycle, speed, taste, focus and flexibility can replace a moat. The author does not dismiss the possibility that some of them will do well anyway. Large, well-defended companies are often willing to pay heavily to bring that hard-to-define startup energy inside the castle and keep it away from rivals.
Several examples follow. The author says it is hard to identify Cursor’s moat, but it is a great product and was worth $60 billion to SpaceX. The article also says Nvidia acquired Hugging Face for $12.9303 billion and notes that the author previously used Hugging Face as an example of how an AI company might use a period of complexity and uncertainty to develop early network effects. The author adds a personal anecdote about speaking with a founder who said, in effect, that the company did not care about moats and only wanted to build a great product. The author gave up on the point, and that company is now valued at roughly 50 times what it was at the time.
Still, the line the essay refuses to cross is this: hoping to get acquired is not strategy, and “better” is not a moat. The piece even says there were rumors that Meta, along with others, had made ten-digit offers for Instinct. When Instinct turned them down, Meta’s answer was simple: then we will build our own. For Meta, that only requires spending money, and it has plenty of money.
The author adds that AI can now write code, which may make copying products without moats even easier than before. But the broader point does not depend on that. Any strong company without a moat can be copied, whether by code generated through AI or by people doing the work manually.
Instinct is trying to build the next moat now
Despite the skepticism, the author does not dismiss Instinct. The essay says the author wants Instinct to win, just as he wants many startups to beat the Goliaths in their industries.
One sign of that effort came the day before publication, when Instinct launched a trusted contacts network that lets a person’s assistant talk with the assistants of partners, family members and friends. The article describes that as a direct attempt to create network effects.
The open question is whether Instinct has enough time to build those effects before users begin adopting Muse at scale, especially since Muse instances will also be able to work with one another.
That returns the argument to its starting point. Time is the scarce resource. Counter-positioning matters because it can make copying painful, even self-destructive, for a stronger rival. A startup’s job is to use that period to sprint until a lasting moat exists.
