A moat you stop digging is a moat that fills back in
Competitive advantage is not an asset you acquire and then protect. It is an activity you sustain without stopping, or it disappears.
There is a comforting story most founders tell themselves once the business is working. The early years built something durable: a brand, a customer list, a proprietary data set, a network of relationships, an operating system nobody else has. The years of effort produced an asset. The asset now protects the business. You can ease off a little, because the moat is in place.
That story is wrong in a specific way, and the wrongness costs companies. A moat is not an asset sitting on a balance sheet earning rent. It is an activity that has to be performed continuously, the same way a trench dug in wet ground fills back in if nobody keeps shovelling. Stop shovelling and the water returns, slowly at first, then all at once. The competitors who looked years behind close the distance faster than you believe possible, because they were never as far behind as the comfortable story suggested. They were behind by the amount of work you were still doing. The moment you stopped, the distance began to collapse.
Every advantage you hold is a depreciating asset. Brand depreciates if you stop earning it through consistent delivery. Technical lead depreciates if you stop extending it. Customer relationships depreciate if you stop serving them at the standard that built them. Even a genuine network effect depreciates if the experience inside the network deteriorates faster than the network grows. None of these compound by sitting still. They compound through continuous work, every quarter, without end.
What a moat actually is
The most useful test for whether something qualifies as a real moat is not whether it exists. It is whether it has a mechanism that keeps widening without you actively working on it. Almost nothing passes that test. What you find when you apply it honestly is that the moats you thought you had are actually activities in disguise.
| Moat | What it looks like while working | What erodes it |
|---|---|---|
| Network effects | Each new user makes the product more valuable for every existing one | The user experience inside the network degrading faster than the network grows |
| Brand | Customers recall your name first and trust it before comparing | Inconsistent delivery, product quality drift, prolonged silence in the market |
| Switching costs | Leaving you means losing data, workflow, or integrations the customer values | Competitors building migration tools that absorb the friction on the customer's behalf |
| Proprietary data | Your product improves with use in ways competitors cannot match | Competitors acquiring enough data themselves, or new architectures requiring less of it |
| Learning rate | You ship and improve faster than anyone else in the category | Slowing down, becoming risk-averse, or losing the organisational tempo that produced the rate |
| Operational software | Internal tools that compress what would otherwise require headcount | Letting the tools decay, refusing to maintain them, treating technology as a cost rather than a lever |
The right column exists for every row. There is no entry where the moat is genuinely permanent. There is no entry where you can stop the work and the advantage continues to widen on its own. The work is the moat. The competitive position is the residue of the ongoing work.
The most dangerous moats are the ones that feel the most durable while they are performing. Brand feels durable because the recognition and trust are real. Network effects feel durable because the maths is real and the switching cost is visible. Both of these have been destroyed within a single product cycle by competitors who simply outworked the incumbent on the specific activity that was producing the effect.
The invisible moat your competitors cannot see
There is one category of moat that almost nobody copies, because they cannot see it from outside. Your competitors can study your advertising. They can replicate your landing pages. They can hire away your marketers and learn what those people knew. What they cannot copy is the layer beneath all of it: the internal metrics you track, the systems you have built up over years of running the same operations, and the small compounded gains your team has accumulated by looking at the same numbers and improving one specific thing by one percent each month.
This is the operational moat, and it is by far the most underestimated. It is underestimated because it has no name that can go on a slide. It is the version of the onboarding sequence that took eleven iterations to land on. The dashboard that took two years to evolve into something that actually drives decisions rather than reporting on them. The hiring filter that produced the particular team composition that now outperforms similar-seeming teams in every comparable role. The dozens of unglamorous operational habits that accumulate into a number nobody outside the business can reverse-engineer.
The reason this moat is so strong is precisely that it is invisible. A competitor who copies your surface assumes the surface is the business. They copy what they can see and then puzzle over why their numbers do not converge with yours. The gap they cannot close is the operational layer they never knew existed.
The same rule applies. The operational moat is also a verb. It widens because you are running the experiments, watching the numbers, improving the systems weekly. The moment that work stops, the gains stop too. Six months of not maintaining the operational layer and you discover that the team has drifted, the dashboards no longer reflect current reality, and the small habits that produced the number have quietly ceased.
When the ground shifts beneath the moat
The deepest mistake the static view produces is not slowing down inside your existing model. It is failing to notice when the model itself has changed around you.
You can run a business for twenty years and lose most of its commercial value in four. The mechanism: the world moves to a new way of doing what you do, you decide your accumulated brand and customer base will protect you during a wait-and-see period, and by the time you acknowledge that the transition is real, the people who started inside the new model have a four-year head start. The brand retains some value. The customer list retains some value. Neither is worth what it was, because the thing they used to grant access to is no longer where the commercial activity is being created.
This is what "you are only as strong as your current execution" actually means. The market does not credit what you built. It responds to what you are doing now. Customers who recommended you last year recommend whoever currently provides the best experience, and if that is no longer you, the recommendation moves without ceremony. Reputation is not a vault. It is a function of recent interactions, replenished by every competent delivery and dispersed by every prolonged absence.
The hardest discipline in a successful business is treating early success as evidence that the next round of work will pay off, rather than as a buffer that permits you to do less of it. The businesses that lose their positions almost never lose them in the first five years, when the urgency is undeniable. They lose them in years ten to fifteen, when the accumulated success has produced a felt sense of security that the competitive environment does not actually support.
Pick one moat. Then dig without stopping.
The opposite mistake from the static view is attempting to maintain every possible moat at once. You do not need them all. One moat, genuinely suited to what your business naturally does, worked on relentlessly, is enough to build a serious business. Attempting to maintain thirteen of them while giving each ten percent of your attention is how most businesses end up with none, because none of them ever reaches the threshold where it actually matters.
The selection process is honest rather than impressive. Look at the business you actually have. Ask which moat has the highest natural fit with what you already do, what your team is already skilled at, and what your customers already give you that competitors cannot access. The answer is almost always clear once you stop looking for the impressive answer and start looking for the accurate one.
A logistics business will not build better brand than a fashion company; it will build better operations. A community-led product will not outperform a research-led one on proprietary data; it will out-network it. A category creator will not beat an incumbent on switching costs; it will out-pace them on learning rate.
Choose the one moat with the clearest natural fit. Then commit to the unglamorous, multi-year work of making it deeper, every quarter, without stopping. The moat is not the choice. The moat is the next ten years of digging.
What changes once this lands
The shift, once you have genuinely accepted that moats are ongoing activities rather than accumulated assets, is in what you stop counting as protection.
You stop counting the brand as a buffer you can draw down. You start counting the activity that earns the brand, week by week. You stop counting the customer list as an asset. You start counting the relationships inside the list, and whether they are being renewed or quietly lapsing. You stop counting your technical lead as a head-start that bought you a year. You start counting the rate at which you are extending it, this month, relative to competitors doing the same.
The relief in this view, once you make peace with it, is that the work is the work. There is no end state where the moat is complete and you can stop. That sounds demanding until you notice the alternative: a slow, invisible erosion you only detect when it is too late to address, followed by the discovery that the advantage you were counting on was never as solid as it appeared. Continuous work feels heavier week by week. Over a decade, it is the lighter path by a wide margin, because the business that chose it keeps surviving.
Pick the moat that fits. Do the actual work. Never conflate what you have built with what you are still building. The day you stop digging is the day the hole begins to fill, and you will not notice until the water is already at your ankles.