The moat you are about to buy, or sell, may be evaporating.

TL;DR: The same force that makes building a replacement cheap makes holding a software-only business dangerous. If your moat is mostly code, it is a melting ice cube, valuable now and smaller every quarter, and buyers know it. They run a replaceability test before the LOI and push the risk that your moat fails into an earnout you have to survive. For a soft moat, waiting is no longer a virtue, it is the risk. Sell into strength, or change what the moat is made of, before the market reprices it for you.

On Wednesday I told you the cheapest way to take a category is no longer to buy it. You build the replacement for a fraction of the capital, and let the incumbent's customers come to you. That was the offense.

This is the defense, and it is the one that should keep you up at night.

The same force that makes building the replacement cheap makes holding the original dangerous. If you own a business whose moat is mostly software, or you are about to buy one, that moat may be a melting ice cube. Valuable today. Smaller every quarter. And nobody at the closing table wants to be the one holding it when it turns to water.

You can already see it in the numbers. Public software multiples compressed about forty percent this year, from roughly 5.6 times revenue to 3.7, and more than two trillion dollars of value came off the sector. But here is the detail that matters more than the headline: the repricing did not fall evenly.

The durable platforms kept their premium, and some of them got more expensive. The replaceable ones got cut in half. The market is no longer paying for "software." It is paying for defensible software and discounting the rest, and it is sorting one from the other in real time, on your business, whether you choose to participate or not.

Two seats feel this, and both are exposed. The buyer can overpay for a platform multiple on an asset that is quietly becoming a feature, and find out a year after close. The seller can be sitting on the ice cube without knowing it, still modeling next year off last year's multiple, still telling themselves the slow quarters are a sales problem rather than the first sign of the melt.

Here is what the sharp buyers now do before they sign an LOI (Letter of Intent). They run what amounts to a replaceability test. They stopped asking "does this company have software?" and started asking "is this software a moat, or a feature an AI could clone in a quarter?" Some have made it a formal line item in diligence.

They put an engineering team on one question: can we rebuild eighty percent of this platform's core workflow with AI orchestration inside ninety days for under a hundred thousand dollars? If the answer is yes, you get filed under soft-moat point solution, and the deal you are offered reflects it.

And it does not just shave the price. It changes the shape of the whole deal. A purchase that two years ago would have been eighty-five percent cash at close now comes back at fifty percent cash, twenty-five in rolled equity, and twenty-five tied to an earnout. The earnout is no longer about growth. It is pinned to keeping the customers you already have, measured on net revenue retention and seat stability, and if they churn out to an internal tool or an AI-native competitor, that last quarter of your price simply evaporates. The buyer has handed you the risk that your moat fails, and made you sign for it.

The rest of the test is short and brutal, and you should run it on yourself before anyone runs it on you.

Those four questions separate a fortress from an ice cube, and most owners have never honestly answered them about their own business.

And here is what most founders miss, the whole point of this letter. For a soft-moat business, waiting is no longer a virtue. It is the risk.

Most owners treat time as an ally. Hold a little longer, grow a little more, sell at a higher multiple next year. That instinct is correct for a durable asset and exactly backward for an appifiable one. AI does not lift everyone equally.

It lifts your hungry challenger faster than it lifts you, so every quarter you hold, the gap they have to close gets smaller, on a clock you do not control and cannot see. The multiple you can get today is very likely the highest you will ever see for that business again. A year spent "growing into a better number" can easily turn out to be a year of watching the number come down to meet you.

And if the buyer suspects the ice cube is melting, they do not just discount the price. They hand the risk back to you as a retention earnout, so that if the moat fails after close, you are the one who pays for it. You sold the business and kept the risk. That is the worst trade in the deal, and soft-moat sellers walk into it constantly.

At Scalable we run the replaceability test from both sides of the table. Before we buy a company, we pressure-test whether the moat we are paying for is real or rentable, so we are not the ones left holding the ice cube. Before a founder sells, we run the same test on their business, so they walk into the room knowing exactly what the buyer already knows, instead of finding it out in the markup of their own LOI.

If the moat is real, the move is to document it and quantify it, the proprietary data, the regulatory lock-in, the genuine switching cost, and then defend the premium with evidence instead of hope. The durable platforms are still getting paid well, and a moat you can prove is worth more in a market that has learned to be suspicious of every one. If the moat is soft, you do the harder, smarter thing. You sell sooner, cleaner, and with less earnout exposure, before the erosion is visible to everyone bidding against you.

And if you are not ready to sell, you do not just wait and hope the melt holds off. You change what you are selling. You wrap the software inside a service and start charging for the outcome instead of the seat, so the thing a competitor can copy is now hidden behind execution they cannot. And you start hoarding the one asset a model cannot reproduce, your own proprietary workflow data, until the product that was easy to clone is sitting on top of something that is not. You cannot always stop the ice cube from melting. You can change what it is made of.

The worst outcome is not selling for less than you hoped. It is finding out at the LOI, from the buyer, that the thing you spent a decade believing was a fortress was an ice cube the whole time, and that they figured it out before you did.

If you are inside the next twenty-four months of a possible sale, run the test before the buyer runs it on you. The fastest place to start is the Exit Ready Score, a three-minute assessment that shows you how a buyer will size up your business and where you are most exposed, before you find it out in the markup of your own LOI.

Score yours here: https://getexitreadyscore.com/

The founders who win the next few years in software-defined categories will not be the ones who held the longest. They will be the ones who knew which kind of asset they were holding, and moved while it was still cold.

— Roland

Most owners think the biggest problem before a sale is their financials.

It usually isn’t. The issues that destroy value tend to be hiding in plain sight.

A business that depends on one person, one customer, or one channel may look healthy on paper… right up until a buyer starts asking questions. That’s when valuation discounts show up.

The best exits aren’t built in the months before a sale. They’re built years earlier by systematically removing the risks buyers care about most.

Want to see what buyers look for before they write a check? Watch the video below.