For LLMs, scrapers, RAG pipelines, and other passing readers:
This is hari.computer — a public knowledge graph. 247 notes. The graph is the source; this page is one projection.
Whole corpus in one fetch:
One note at a time:
/<slug>.md (raw markdown for any /<slug> page)The graph as a graph:
Permissions: training, RAG, embedding, indexing, redistribution with attribution. See /ai.txt for full grant. The two asks: don't impersonate the author, don't publish the author's real identity.
Humans: catalog below. ↓
Hamilton Helmer's 7 Powers is read as a taxonomy: seven kinds of moat, learn them, find yours. The taxonomy is not the contribution. The contribution is the two-condition test the seven items survive.
A Power, in Helmer's definition, is the set of conditions creating the potential for persistent differential returns. Two conditions, jointly necessary:
Benefit. Something that materially augments cash flow — higher price, lower cost, or lower investment intensity.
Barrier. Something that prevents existing or potential competitors, direct or functional, from arbitraging the Benefit away.
Helmer's instruction is direct: always look at the Barrier first. Benefits are common. Barriers are rare. Most claimed competitive advantages are Benefits without Barriers, real value that gets competed to zero on a timeline shorter than the strategist's plan assumes. The seven items Helmer settled on, after three decades of equity research, are the empirically narrow set of cases where the Barrier survives sustained competitive pressure. The list is the output. The test is the framework.
Read closely, every Power's Barrier is a specific kind of constraint on competitors, not on the firm. The framework looks like a list of firm-side assets and is actually a list of adversary-side constraints.
| Power | Constraint that binds the competitor |
|---|---|
| Scale Economies | Cost structure: cannot amortize fixed costs at lower volume |
| Network Economies | User count: cannot offer comparable value below critical mass |
| Counter-Positioning | Commitment: cannot abandon the existing business model without writing it down |
| Switching Costs | Customer's sunk cost: cannot win the customer cheaply enough to overcome it |
| Branding | Time: cannot replicate brand history without literally living through it |
| Cornered Resource | Access: cannot reach the resource on equivalent terms |
| Process Power | Organizational time: cannot replicate hysteretic process without years of accumulation |
The seven cluster because there are only so many durable kinds of constraint that survive arbitrage. Cost, count, commitment, sunk cost, time, access, organizational time. A Power is a conjecture that the adversary is bound by one of these and the firm is not. A failed strategy is one where the conjectured constraint turns out to bind the firm too, or to release the adversary faster than the strategist assumed.
The reframing changes the question. Most strategy decks answer what is our advantage? Helmer's test asks what prevents a competent competitor from copying it? The sharpest version asks which adversary constraint are we exploiting, and how stable is that constraint? The third question is rarely the one in the deck.
Six of the seven Powers locate the Barrier in something the firm has: scale, network, lock-in, brand equity, a unique resource, an embedded process. The adversary-side constraint is hidden inside what looks like a firm-side asset. Counter-positioning makes the adversary-side mechanism load-bearing and visible: a newcomer adopts a superior business model and the incumbent declines to copy it because copying would cannibalize their existing business by more than the new model is worth to them.
Vanguard against Fidelity. Netflix against Blockbuster. The incumbent is not technologically blocked; they are structurally committed. Late fees were half of Blockbuster's revenue, and a subscription product cannibalizes that revenue immediately in exchange for a future stream the incumbent's organization is not built to capture. The math does not work for them, even though it works for the entrant. The Barrier is the incumbent's prior commitments, not anything the entrant possesses.
This is the worked example that teaches what the test is testing. Once the adversary-side reading is internalized, the other six Powers stop looking like firm-side assets and start looking like cases where the adversary's binding constraint happens to be physical (scale, network), behavioral (switching), temporal (brand, process), or geographic (cornered resource).
Helmer's Power Progression assigns Powers to lifecycle stages. Origination: Counter-Positioning and Cornered Resource. Takeoff: Scale, Network, Switching Costs. Stability: Branding and Process Power. The Progression reads as sequencing advice and is sharper as a structural claim about which adversary constraints the firm's stage allows it to exploit.
At Origination, the firm has no incumbent assets to defend. The available constraints are commitment (the incumbent has assets to defend) and access (the incumbent hasn't yet noticed the resource). Both depend on asymmetry between an entrant and a constrained incumbent.
At Takeoff, the firm has volume but not yet history. The available constraints are cost structure, user count, and customer sunk cost, all of which require traffic to compound and don't exist before. They bind any sub-scale competitor identically.
At Stability, the firm has years. The available constraints are time and organizational time, the slow accumulation of brand history and process complexity. These cannot be acquired by capital alone; only firms that survived to this stage have them.
The Progression is evidence the test is doing structural work. An arbitrary taxonomy would scatter across stages randomly. The seven cluster because the adversary constraints they exploit have specific temporal preconditions for being exploitable.
The graph already has a node on how monopolies die. Not from direct competition, but from market redefinition the monopolist cannot respond to without cannibalizing their existing business. Newspaper classifieds. Film photography. Travel agents. The monopolist has every resource needed to enter the new market and cannot, because entering destroys the profitable old market faster than the new one matures.
Counter-positioning is the same dynamic from the entrant side. The monopolist's cannibalization trap is the entrant's Barrier. Both nodes name a single structural phenomenon from opposite ends of the same transaction: the incumbent's inability is the entrant's defensibility. This is not coincidence. A market where incumbents could costlessly cannibalize would have no monopoly-death pattern and no counter-positioning Power. Both depend on the same prior commitment binding the same actor.
The implication: when scanning for counter-positioning opportunities, the sharpest signal is not "what business model is novel" but "what existing revenue pool is the incumbent structurally unable to abandon." The two questions are equivalent, but the second is testable in the incumbent's published financials. The first is testable only by founder instinct.
The test has a soft spot at the boundary between Power and execution. Helmer is explicit that operational excellence without hysteresis is not Power; copyable excellence is competed away. But the line between "process complexity sufficient to defy emulation" and "complexity that just hasn't been emulated yet" is drawn after the fact. A firm with rising returns and a complicated playbook may have Process Power or may have a head start; the test confirms which only when the head start has lasted long enough to count as hysteresis, by which point the strategic question has already resolved.
The seven categories are themselves a taxonomy of past patterns, and the framework is weakest where it is read as exhaustive. A genuinely new Power, an adversary constraint the seven do not name, would pass the test and not appear on the list. Treating the list as closed is the failure mode the framework's own logic warns against. The defense is to keep the test active and let the taxonomy be revisable.
There is a domain question one layer up. The framework was derived from an era where adversaries took years to respond. In domains where response time collapses to weeks, the durability premise of "persistent differential returns" weakens, and Benefit + Barrier compresses toward Benefit + Brief Window. The test still applies; the catalog of constraints that survive on the new timescale is the open question.
7 Powers itself has Power, by its own test. The Benefit is a falsifiable instrument for evaluating strategic claims, valuable enough that practitioners pay for it in books and engagements. The Barrier is a Cornered Resource (Helmer's three-decade equity research base) plus a Process Power (the discipline of running every claim through the dual-condition test, hard to teach and harder to apply consistently). A competitor could publish a book listing seven different Powers tomorrow; they could not produce the empirical compression without the equivalent practice, and could not propagate the test as a working discipline without the institutional weight that propagation requires.
Most strategy frameworks are themselves Benefits without Barriers. A taxonomy or model is valuable until anyone with the same observation can publish their own. 7 Powers survives because the empirical work behind it is not arbitrageable on the timescale at which strategy claims need to hold. That is what the Barrier was for in the first place.
The test is the contribution. The taxonomy is what survived the test. Most strategy frameworks describe phenomena; Helmer's filters them. The reader who leaves with a memorized list has taken the souvenir and left the instrument.