How to Assess a Startup's Defensibility (Scorecard)
Assess a startup's defensibility by scoring your idea against Hamilton Helmer's seven powers — scale economies, network effects, counter-positioning, switching costs, branding, cornered resource, and process power — and rating each weak, medium, or strong on real evidence. Before product-market fit, most rows score weak, and that is normal.
Quick Answer: There is no passing number. A defensibility scorecard rates your idea against Helmer's seven powers (7 Powers), scoring each weak, medium, or strong on evidence — not ambition. Most pre-PMF startups score mostly weak; the useful output is an honest thesis about which one power you can credibly build toward, not a claim that a moat already exists.
The defensibility scorecard: seven moat sources to rate
Defensibility isn't one thing you have or don't — it's seven distinct sources you rate one at a time. The scorecard below walks each of Hamilton Helmer's seven powers, from his book 7 Powers, and pairs it with the single question that tells you whether the barrier is real. These are the same types of startup moats every durable business traces back to.
Score each row weak, medium, or strong — and read the barrier, not the benefit. Helmer's core insight is that every power has two halves: a benefit that lifts your economics and a barrier that stops a competitor from copying it away. A benefit without a barrier is a feature. The question column below targets the barrier on purpose.
| Moat source (Helmer's 7 Powers) | The question to ask | Weak signal | Strong signal |
|---|---|---|---|
| Scale economies | Do our unit costs fall with volume in a way a smaller rival can't match? | Costs are flat or rising as we grow | Structural per-unit cost decline rivals can't reach without equal volume |
| Network effects | Does each new user make the product more valuable to existing users? | Growth is linear; users don't benefit from one another | Value compounds per user; late entrants can't match without an equivalent base |
| Switching costs | Would leaving cost our customers data, retraining, or integration rework? | A customer could export and switch in an afternoon | Accumulated data, workflows, and integrations make leaving painful |
| Counter-positioning | Is there an incumbent who structurally won't copy our model? | Incumbents could copy us tomorrow with no downside | Copying would cannibalize the incumbent's core profits, so they won't |
| Cornered resource | Do we hold an asset rivals can't obtain on equal terms? | "Great team" or generic tech anyone can hire or build | Issued patent, exclusive license, or genuinely scarce, contracted asset |
| Branding | Would customers pay more for our identical offering because of who we are? | The name means nothing yet; buyers compare on price | Durable trust that commands a premium for the same object |
| Process power | Do our operations beat rivals' even when they know how we do it? | No proprietary process; anyone could run the same playbook | Embedded operating capability built over years, hard to copy or even see |
Takeaway: If you can answer the benefit half fluently but stall on the barrier, that row is weak — and for a pre-PMF startup, most rows should be. A scorecard of mostly-weak scores isn't a failing grade; it's an accurate one.
How to rate each moat source honestly (weak, medium, strong)
Rate on evidence you can point to, not the moat you hope to build — and when in doubt, score weak. The whole value of the exercise collapses the moment you grade generously. Here is what each tier actually means.
- Weak (the honest default): No barrier exists yet. At most you have a benefit, or a seed you intend to plant. Nearly every power scores weak before product-market fit — scale, networks, brand, and process all need traction you don't have yet.
- Medium: A barrier is visibly forming and you have early, measurable evidence — a retention curve bending up, switching costs starting to bite — but a well-funded rival could still clear it.
- Strong: A structural barrier a competitor can't clear without disproportionate time or cost, backed by evidence, that gets taller as you scale.
Before you score any row above weak, run it through a five-question honesty gate:
- Can I name the barrier in one sentence — not just the benefit?
- Do I have evidence, or am I describing an aspiration?
- Would this survive a well-capitalized competitor copying us tomorrow?
- Does it strengthen with scale, or erode with imitation?
- Is this a moat now, or a seed that only becomes one after PMF?
Only two powers are realistically scorable above weak this early. Counter-positioning is a business-model decision you can make on day one, and a cornered resource — a patent, an exclusive dataset, a locked-in license — is sometimes in hand from the start. The other five are seeds. Helmer himself notes that some powers can only be established at particular stages of a company's life; scale and process power simply aren't available to a three-month-old startup.
Failure modes the defensibility scorecard catches
The scorecard earns its keep by catching the specific ways founders flatter themselves. Each failure mode below is a weak row dressed up as a strong one. Naming them is what keeps the exercise honest.
- Scoring a benefit as a barrier. The most common error by far: you describe why you make money, not why you keep it. Those are two different sentences — if you only have the first, the row is weak.
- Logging "great team" as a cornered resource. Talent is hireable. It's cornered only if it's genuinely irreplaceable and contractually locked in.
- Logging "first-mover" as a moat. A head start is not a barrier. Fast followers overtake pioneers routinely; being early counts only if you spend the lead building an actual power.
- Logging "better UX" as switching costs. A superior product is copyable and a preference, not a lock-in. Switching costs live in the data and workflows a customer would lose, not in how much they like you.
- Logging "we have data" as a network effect. A static dataset is a cornered-resource candidate at best. It's a network effect only if it compounds and improves the product for every user as more people join.
- Grading your favorite idea generously. The pull to score the moat you wish you had is strong. Default to weak and make the evidence argue you upward.
The thread running through every one is the benefit-versus-barrier confusion — the exact line drawn in moat vs. competitive advantage. An advantage helps you win now; only a barrier keeps a rival from taking the win back.
Turning your scores into a defensibility thesis
A scorecard of mostly-weak scores is the starting input, not the verdict. No investor expects a seed-stage company to have a finished moat. What they — and you — actually need is a thesis: a credible account of which single power you are building toward and why it will form. The output of the scorecard is that thesis, not a claim that a moat already exists.
Build the thesis from the row with the most credible path, and answer four questions:
- Which power? Name one — usually your least-weak row, often counter-positioning or a cornered resource this early. One power stated with conviction beats five gestured at.
- Why now? What shift in technology, regulation, or behavior makes this the moment the barrier can finally be built?
- What's the leading indicator? The measurable, small-scale signal that the barrier is forming — deepening retention, accumulating proprietary data, an incumbent visibly declining to respond.
- Why does it compound? The mechanism by which each new customer, dataset, or unit of scale makes the barrier taller rather than flatter.
Treat the thesis as a falsifiable assumption, not a slide to assert. "Customers will be locked in by their data" has a truth value you confirm by talking to customers who tried to leave — the same evidence-first discipline you'd apply when running an idea through the idea validation scorecard. Instrumenting for those leading indicators now, at small scale, is exactly the validation work Edmired is built to support.
Key Takeaways
- Defensibility is seven sources, rated one at a time — scale economies, network effects, switching costs, counter-positioning, cornered resource, branding, and process power, from Helmer's 7 Powers.
- Every power is a benefit plus a barrier — score the barrier; a benefit alone is a feature, not a moat.
- Weak is the honest default before PMF — most rows should score weak, and a mostly-weak scorecard is accurate, not a failure.
- Only counter-positioning and (sometimes) a cornered resource are scorable above weak early — the other five are seeds that need traction first.
- Run every above-weak score through the honesty gate — name the barrier, show evidence, and confirm it strengthens with scale.
- Most failure modes are one confusion — logging a benefit, a head start, or a preference as a barrier.
- The output is a defensibility thesis, not a moat — one power, why now, a leading indicator, and why it compounds.
Frequently Asked Questions
What's a good defensibility score for a pre-PMF startup?
There is no passing number, because the scoring is qualitative, not additive. A pre-product-market-fit startup that honestly scores weak on five or six of the seven powers is normal, not doomed. What sophisticated investors evaluate is the thesis — which one power you can credibly build toward — far more than the current state of the scorecard.
Which moat can an early-stage startup realistically score strong on?
Usually just counter-positioning, and sometimes a cornered resource. Counter-positioning is a business-model choice you can make on day one — adopting a model an incumbent won't copy because it would cannibalize their profits. A cornered resource, like a patent or exclusive license, is occasionally in hand early. The remaining powers require scale, a network, or years of trust.
How often should I redo the defensibility scorecard?
Rescore at every major milestone — a new raise, a pricing change, a jump in retention — because defensibility is a moving picture, not a one-time grade. The signal you're watching for is rows migrating from weak to medium as real evidence accumulates. If they stay weak across several rounds of traction, your moat thesis needs rethinking.