Competitive Moat: How to Tell Which Advantages Will Last
A competitive moat is an advantage that gets harder to copy as a company grows and competitors respond. A head start, such as being first or shipping a better feature, fades once a funded rival decides to match it. A moat comes from structure: customers who'd pay real costs to leave, or a business whose parts reinforce each other so no single piece can be copied.
Picture two farms the morning after a summer storm. One farmer set out rain barrels, and every one is full to the brim. The neighbor spent the spring digging a well, and on that wet morning the work looks wasted.
By August the barrels are dry. The well still draws water, because something underground keeps filling it.
Most startup advantages are rain barrels. Being first or having a feature nobody else has looks full the day you launch, and it drains as competitors catch up. A plan built on it runs dry with it. A competitive moat is the well. Two questions tell them apart: does the advantage get stronger as you grow, and what would a rival have to give up to copy it?
Head starts that fade
Start with the barrels, because they're what founders pitch first. A head start is worth using while it lasts. It just won't be the reason a buyer still chooses you in three years. The common ones:
- Being first: the second company learns from your mistakes for free.
- A feature: a competent team can rebuild most features in a few months.
- A lower price: a competitor with more cash can match it the week it starts to hurt.
- More users: a bigger user count protects you only when each user makes the product better for the others.
- Better execution: good habits spread as fast as rivals can hire your people or read about how you work.
Each of these drains for the same reason. A rival can match it without giving anything up.
Seven types of competitive moat
The wells work the other way. Copying one costs the rival something it already depends on, or the advantage grows as the business does. Seven types cover nearly every moat a company can hold:
| Moat | Lasts when | Fades when |
|---|---|---|
| Network effects | Each new user makes the product better for the users already there | Users get counted, but nothing connects them |
| Switching costs | Leaving means migrating data, retraining staff or rebuilding integrations | A rival's onboarding team can move a customer in an afternoon |
| Scale economies | Unit cost keeps falling as volume grows | A mid-size rival already gets the same savings |
| Brand | Buyers pay more for the name, or skip comparing options | People know the name and still shop around |
| Cornered resource | A patent, dataset or input a rival can't legally or practically get | A good team could rebuild it in a few months |
| Regulatory position | A license or certification that's hard to earn | Every company in the category already holds it |
| Counter-positioning | Copying your model would eat into a rival's profitable business | The rival just hasn't gotten around to copying it |
Most of these take years to build. Expect an early company to hold one, maybe two.
Scoring your moats
The table only helps if you rate your own company against it honestly. Founders score the types that sound impressive and skip the evidence, so make the evidence part of the score:
- Score each of the seven types from 1 (absent) to 5 (strong), using the "Fades when" column as the test for a 1.
- Write one line of evidence beside each score: a renewal rate, a contract term, something customers actually do.
- Drop any score without evidence to 1.
- Add the seven scores. The maximum is 35.
A common way to read the total:
- Under 15: every advantage you have is a head start. Treat it that way.
- 15 to 24: one or two moats are real. Build around them and stop pitching the rest.
- 25 or more: a defensible position, and rare this early.
Take a hypothetical scheduling tool for HVAC contractors. Switching costs score a 4, because each contractor's customer history and technician calendars live inside it. Brand scores a 2, since contractors still compare it with two rivals at every renewal. Network effects score a 1: one contractor's use does nothing for another's. The other four types score low, and the total lands at 14. The founder should stop calling the dispatch feature a moat and start deepening the data a customer would lose by leaving.
When you score under 15: most early companies land there, and it isn't a verdict on the business. Win a narrow segment outright, and pick the one moat that grows with each customer you add. Just don't describe a head start as a moat to investors. The good ones will check.
Activity systems
One kind of moat hides from the scoring table, because it lives in how a company's parts fit together. Michael Porter made the case in "What Is Strategy?" (Harvard Business Review, 1996). A rival can copy any one activity, but copying a set of activities that reinforce each other means rebuilding the whole company at once. Southwest Airlines was his example: its fast gate turnarounds kept planes in the air more hours a day, and those extra hours paid for its low fares.
You can map your own system in an afternoon:
- Name the three to five themes your position rests on, such as self-serve onboarding or one industry's compliance needs.
- List the 15 to 30 things the company actually does each week to deliver them.
- Draw a line wherever one activity makes another work better.
- Circle any activity with a single line. It costs money without protecting anything.
Now ask what a rival could copy alone. An activity that delivers its full benefit without the rest is a feature, and it'll drain like a barrel. The dense knots in the web, where three or four activities only work together, are the well.
Building a moat that lasts
Your lasting advantages are the ones that grow with each customer, or that a rival can't copy without giving up something it depends on. Everything else is a head start. Use it while it's full, and put the time it buys into one moat you can deepen.
Unfair advantage covers the step before this one: deciding what you have to build on at all. A competitive analysis shows which of your advantages rivals already match, and a positioning map shows where buyers think you stand. Choosing which moat to build, and aiming the marketing at it, is part of my fractional CMO work, which starts with a free diagnostic.
Barrels fill fast. Wells keep filling.
Common questions
What's the difference between a competitive moat and a competitive advantage?
The difference between a competitive moat and a competitive advantage is how long it lasts. Any advantage helps a company win deals today. A moat keeps helping after rivals respond, so every moat is an advantage, while most advantages are head starts that fade once a funded competitor decides to match them.
Can a service business have a competitive moat?
A service business can have a competitive moat, though it rarely comes from the service itself, since a rival can hire people to deliver the same work. Service firms usually build one from switching costs, such as being wired into a client's systems, or from a specialty narrow enough that buyers default to their name.
Is AI a competitive moat for a startup?
AI on its own is rarely a competitive moat for a startup, because competitors can buy the same models. A moat can form around AI when the company owns data a rival can't get, or when the AI sits so deep in a customer's workflow that leaving would mean rebuilding it.
How often should a startup reassess its competitive moat?
A startup should reassess its competitive moat once a year and whenever a new competitor starts winning its deals. Lost deals show which advantage a rival has matched, often months before the revenue numbers do.