ExpertsFMPlatforms and Power (with Hamilton Helmer and Chenyi Shi)
Acquired · April 6, 2022
5 Selected Lessons:
heuristic · 3 case · 2
Multi-Homing Arbitrages Away Platform Power
Source
Acquired — Platforms and Power (with Hamilton Helmer and Chenyi Shi) (April 6, 2022) · at 20:47 ▸ · at 1:08:48 ▸ · at 1:16:17 ▸
When this applies
when any business's competitive advantage depends on customers being 'locked' to it, but customers can trivially switch or use alternatives in parallel
The takeaway
A scale-based advantage only converts to power if the cost of using a rival in parallel (multi-homing) is high; low switching/parallel-use friction destroys the link between scale and profit capture.
Why it works
If customers on either side of a platform can costlessly use multiple competing platforms simultaneously, any differential scale advantage gets neutralized because customers simply contribute to whichever pool is locally larger in the moment, eliminating the platform's unique capture of value.
One level up
Any competitive advantage built on customer scale must be stress-tested against how easily customers can defect to or simultaneously use a competitor.
Where it stops holding
This stops working when there are real frictions to multi-homing — unique inventory, contractual lock-in, high switching costs, or sunk learning/data — that prevent customers from costlessly using a rival in parallel.
Detail
rule: Check whether both sides of a platform can frictionlessly use a competing platform too; if so, relative scale stops mattering because all parties effectively pool into whichever match is best at that moment.
threshold: Power requires friction to multi-homing high enough that relative scale differences persist and matter; frictionless multi-homing (e.g., opening two ride-hailing apps) erases the advantage entirely.
rationale: Ben's own startup built a meta-search tool across ride-sharing apps that was effectively a friction-reducer enabling multi-homing — Uber shut off API access specifically because this threatened their power.
In their words
So this is a scenario that we call multi-homing. And essentially what we're saying is a lot of the differential value a platform can generate is a result of differential scale they have with their participants. And multi-homing is what arbitrage out all of that differential value.
Build Switching Costs Through Value, Not Lock-In Tactics
Source
Acquired — Platforms and Power (with Hamilton Helmer and Chenyi Shi) (April 6, 2022) · at 58:00 ▸
When this applies
when any business wants to deepen customer retention and is tempted to design lock-in mechanisms directly
The takeaway
Pursue switching costs as an emergent consequence of delivering real customer value, not as the primary design goal.
Why it works
Designing primarily for lock-in creates a win-lose dynamic that customers resent and that tends to erode over time or invite backlash/regulation; designing for genuine value first, where switching cost is a natural byproduct, creates a durable win-win relationship.
One level up
Advantages built on withholding or coercing (win-lose) are less durable than advantages built on genuinely creating value (win-win).
Where it stops holding
This stops working when a competitor can trivially replicate the exact value-add (e.g., a similar insurance program), in which case the 'switching cost' from value never becomes a true differential barrier.
Detail
rule: When seeking to build switching costs, ask 'what can I do for my customers that creates value?' first; if a lock-in effect emerges as a byproduct, that's acceptable, but don't start from 'how do I lock customers in.'
threshold: N/A — qualitative design principle.
rationale: Companies with high switching costs face an ethical/strategic tension: as long as customers can't easily leave, the temptation is to extract more and treat them worse over time, which Helmer frames as a 'bad marriage' dynamic to avoid.
In their words
My advice always is: don't think, how am I going to lock in my customers? But actually think of it a different way. What is it that I can do for my customers that create value for them? And then if that proposition has with it a way that they're tied to me, then so be it.
'Run For The Hill' — The Flywheel Logo-Swap Test
Source
Acquired — Platforms and Power (with Hamilton Helmer and Chenyi Shi) (April 6, 2022) · at 1:08:48 ▸
When this applies
when any team presents a self-reinforcing growth loop as proof of competitive defensibility
The takeaway
A growth-loop narrative should be tested by substituting a competitor's name into the diagram; if the logic still holds, the flywheel proves growth, not defensibility.
Why it works
A flywheel diagram only illustrates how different parts of a business reinforce each other to drive growth (product-market fit); it says nothing about why a competitor couldn't build and benefit from the identical loop, since the same enabling technology that created the loop often also enables easy multi-homing/replication by rivals.
One level up
Treat visually compelling strategy narratives with suspicion if they only describe positive feedback loops without addressing why rivals can't replicate the same loop.
Where it stops holding
This stops working when the flywheel genuinely depends on a differential, non-replicable scale advantage (e.g., proprietary data or direct network effects) that a swapped-in competitor could not reproduce.
Detail
rule: When presented with a flywheel, swap the central company's logo for a competitor's and check if the described loop still 'works' logically — if so, the flywheel demonstrates product-market fit mechanics, not a competitive barrier.
threshold: N/A — qualitative diagnostic test.
rationale: The same technology that enables rapid flywheel-driven growth often also enables multi-homing, so the benefits are usually contingent on differential scale on one or both sides of the platform — a separate, unaddressed question.
In their words
Anecdotally, there's one exercise we did, which is really simple, but surprisingly interesting, which is, we took a flywheel of a company that's really popular, I won't name which one, but we swapped that logo with their competitor's logo in the middle of that flywheel, and it still works.
TSMC May Sacrifice Current Profit to Secure Future Customer Commitment
Source
Acquired — Platforms and Power (with Hamilton Helmer and Chenyi Shi) (April 6, 2022) · at 44:57 ▸ · at 47:47 ▸ · at 51:52 ▸
When this applies
when any capital-intensive business with predictable performance improvements must decide whether to extract maximum current margin or subsidize to lock in future volume commitments
The takeaway
In the hosts' interpretation, sacrificing current extraction for customer commitment may create power when scale differentially compounds into cost/performance advantage via large, lumpy, predictable technology investments.
Why it works
The hosts reason that TSMC's industry has unusually large, lumpy capital requirements, predictable (Moore's-Law-like) performance gains, and scarce upstream suppliers (ASML) requiring long forward commitments — these conditions could let giving up current margin help secure future customer volume, which may fund staying at the technology frontier, which in turn could yield lower cost per transistor and, they suggest, probably higher-margin, higher-performance products versus competitors.
One level up
Pricing decisions should be justified by their effect on underlying structural fundamentals (scale economies, commitment credibility), not treated as a simple tactical lever.
Where it stops holding
This stops working when the industry lacks predictable technology step-changes, lumpy capital requirements, or scarce upstream suppliers requiring long commitments — in those cases subsidizing customers doesn't translate into a compounding scale/cost advantage.
Detail
situation: TSMC could charge customers (like NVIDIA) maximally given its leading-edge fab technology, but instead prices to encourage long-term commitment.
decision: The hosts interpret that TSMC gives up current profit margin in exchange for future revenue and customer loyalty, which may enable it to commit $10B+ per fab and secure scarce upstream supplier capacity (e.g., ASML lithography machines) years in advance.
outcome: This cycle, as the hosts describe it, could let TSMC stay at the technology frontier, achieve lower cost per transistor, and produce higher-performance products that are probably higher margin, potentially extending its lead over Intel.
lesson: In industries with lumpy capital, predictable tech roadmaps, and scarce upstream capacity, subsidizing customers to encourage future revenue may be an economically rational path toward sustained power — not mere benevolence.
In their words
So that allows a scale company to get to the newer technology first. And that has profound implications for sort of continuing that cycle, but it means that they're lower cost per transistor and they also can do higher performance products, which probably are higher margin.
NVIDIA's Early Lack of Power Despite First-Mover Status
Source
Acquired — Platforms and Power (with Hamilton Helmer and Chenyi Shi) (April 6, 2022) · at 1:16:17 ▸
When this applies
when any pioneering company in a hot new category assumes being first and well-funded guarantees durability
The takeaway
First-mover status and strong backing are not sources of power on their own; power must be separately engineered even for pioneers.
Why it works
Being first, well-funded, and well-staffed creates no inherent barrier to entry if the underlying technology/market is easy for others to replicate — within months, dozens of funded competitors can emerge, reducing the first-mover's advantage to near zero.
One level up
Never conflate the prestige of being early and well-capitalized with having solved the defensibility problem.
Where it stops holding
This stops working when the pioneering company does secure a genuine structural advantage early (e.g., a hard-to-replicate technical moat), in which case first-mover status combined with that moat can compound.
Detail
situation: NVIDIA was the first computer graphics chip company in Silicon Valley, well-funded by Sequoia and Sutter Hill, with a strong team.
decision: N/A — descriptive of initial market conditions before NVIDIA had to pivot strategy.
outcome: Within roughly a year, around 90 other funded companies entered the same space, showing there was no inherent power in being first; NVIDIA had to develop power through other means later.
lesson: Early funding and first-mover status do not by themselves create a barrier to entry; power has to be separately built.
In their words
And then, you know, they wake up six, 12 months later, and there's 90 other companies that have all been funded doing the same thing. You know, there's no power there.