Founder has idea. Founder raises money. Company grows fast. Founder raises more money. Then the growth rate softens, the next round doesn't come together, and the board starts using words like "optionality." That's the moment nobody prepared you for, and the script runs out exactly when you need it most.
The thesis here is simple: for a meaningful number of founders staring at a wind-down, selling the business is a better outcome than shutting it down, and the only reason it doesn't happen more often is that nobody tells them it's possible until the clock has nearly run out.
The venture model has a specific shape. A fund needs a few companies to return the whole fund, which means every portfolio company is quietly being measured against a 10x or 100x return threshold. A company doing $3M ARR and growing 25% year over year is, by most human standards, a genuinely good software business. By fund math, it is a problem. The growth curve doesn't bend toward the exit that makes the fund story work.
So the board gets quieter. The bridge conversations start. And the founder, who has been running hard for three or four years, suddenly finds themselves holding a business that works, a team that depends on them, and a cap table that complicates everything.
Estimates on startup failure rates vary widely depending on how you define failure and when you start the clock, but the data that matters here isn't the aggregate. It's the composition of what's actually inside the failure bucket. A meaningful share of companies that shut down aren't broken. They're businesses that ran out of time, money, or board patience inside a model that demanded something they couldn't deliver. Shutting down was the path of least resistance, not the only path.
These founders did what their investors wanted them to do. The model moved, not the founder.
A wind-down has a certain tidiness to it. You send the announcement, you help the team land well, you return whatever's left to investors, and you close the chapter. It feels like a clean ending because the process is structured. But the accounting of what disappears is rarely done honestly.
Customers who paid for something that now stops working. A team that built real expertise and now scatters. Code and infrastructure and institutional knowledge that took years to develop, all of it gone. And yes, the founder's equity: if the company had any real revenue, a sale might have returned something. A wind-down returns nothing except closure.
The human stakes are real and they run in both directions. Winding down costs people something, and so does a bad sale. The point isn't that selling is always the answer. It's that the wind-down path gets chosen by default, without the honest accounting it deserves.
When the next round isn't coming, there are really three paths. Here's what each one looks like, and whose interests it serves.
I want to be clear about our incentives here too. Curious buys software companies. We're not a neutral party. When I argue that a sale often beats a wind-down, I'm also describing the outcome we exist to make possible. That tension is worth naming, and you should weigh it accordingly.
The question founders in this position usually ask first is valuation. The honest answer is that a company with renewing revenue is worth more than its founders assume, and usually less than their best-case scenario.
In our experience reviewing deals (as of September 2026), software companies at $2M to $10M ARR with reasonable retention typically trade in a range of 2x to 5x ARR, with the spread driven by growth rate, gross margin, churn, and how much the business depends on any single customer or the founder personally. A company growing faster, with low churn and high margins, sits at the top of that range. A company flat or slightly declining, with meaningful founder dependency, sits at the bottom. Neither number is zero.
The cap table complicates the math, and that's worth working through honestly before you assume a sale doesn't pencil. Liquidation preferences mean investors get paid first, which sometimes means the founder gets nothing from a sale at a price that would otherwise feel real. The way to check: run the waterfall. Take the likely sale price, subtract the preference stack, and see what's left for common. Sometimes the answer is not much. But sometimes founders are surprised in the other direction, because the preference stack is smaller than they remembered, or because the price is higher than they expected, or because investors, facing the alternative of a zero from a wind-down, are willing to negotiate the preferences to get a deal done.
Talk to your lawyer before you sign anything. The cap table math above is directional, not legal advice, and the specifics matter enormously.
One path that often gets floated in this situation is the acquihire: a larger company hires the team, often with a nominal payment for the assets, and the company effectively shuts down as a going concern. The product usually goes dark. Customers migrate or churn. The acquiring company gets the talent.
Acquihires can be the right answer. If the product isn't viable on its own but the team is exceptional, an acquihire may recover more for investors and give the team a soft landing. But they're not the same as an acquisition, and founders sometimes accept an acquihire offer without testing whether an outright acquisition was possible. The two conversations are different, and having both is worth the time.
The distinction matters because an acquihire treats the company as a talent transaction. An acquisition treats it as a business. If there are customers paying real money to use the product, the product has a value beyond the team, and a buyer who operates software for the long term will see it differently than a strategic that just wants the engineers.
Not every company in wind-down territory is sellable. Here's what actually moves a deal forward, from the buyer's chair.
Revenue that renews. Monthly or annual subscriptions where customers come back without being re-sold are the core of what a buyer is buying. Even at a modest ARR, if customers renew, the revenue has durability.
Gross margin above 60%. Software margins are the reason software companies command a premium over other business types. A company with infrastructure costs eating into margin needs a buyer who can see a path to fixing it.
Customers who would notice if you disappeared. The simplest test of whether a product is real: would the customers complain, or would they quietly move on? A product embedded in someone's workflow, even a small number of customers, is worth something.
A business that can run without the founder. This is the hardest one for founders in this situation, because the company often depends on them in ways that are hard to document. But a buyer has to believe the company can operate after the founder leaves, or the deal doesn't close. The more documented the processes, the easier this becomes.
There are deal-killers too, and I'd rather name them than let a founder waste months on a process that won't close. Single-customer concentration above 30% of revenue makes buyers nervous about what happens if that customer churns after close. Unresolved IP ownership questions, especially around code developed by contractors without clear assignment agreements, can stop diligence cold. And a preference stack that leaves nothing for common equity can make investors reluctant to approve a sale at the price a buyer will actually pay.
Founders who haven't run a sale process before sometimes imagine it as a lengthy formal auction with bankers and data rooms and competing bids. That process exists, but it's not the only one, and it's often not the right one at this stage. A company at $3M ARR in a distressed position doesn't need a nine-month process. It needs a buyer who can move quickly and close.
The practical version looks like this: identify two or three buyers who buy software companies in your ARR range, reach out directly, have a conversation about fit, share financials under NDA, negotiate an LOI, and run diligence. We close within 60 days and pay cash. That's not a tagline in this context; it's the practical answer to the problem a founder in this position actually has, which is that time is the enemy.
A broker can help, especially if you don't know who the buyers are or want someone to run the process while you run the company. The cost is real: a success fee of 10% to 15% on a $3M exit is $300K to $450K. Whether that fee is worth it depends on whether the broker finds a buyer you wouldn't have found yourself, or negotiates a price high enough to cover the fee. Sometimes yes, sometimes no. We've written about when a banker makes sense and when it doesn't.
Selling is not always the right answer. Some companies really have run their course, the product has lost its relevance, the customers are churning for reasons a new owner can't fix, and a wind-down is the honest call. I don't want to oversell the sale option because we're buyers; that would be exactly the kind of motivated reasoning founders should ignore.
But the default toward wind-down, chosen without testing the alternative, costs founders something they don't get back. Two weeks of honest inquiry, a few conversations with potential buyers, a quick look at whether the cap table math works: that's the cost of finding out. The downside of looking is low. The downside of not looking, and later learning there was a buyer who would have paid a real number, is something else entirely.
Curious is a long-term holding company that buys and grows software companies with empathy. We operate what we buy for decades, the team stays together or hands off cleanly, and we're not scared of situations that aren't perfectly tidy. Backed by operator LPs with cash on our balance sheet, so there's no financing risk. No ghost equity or complicated structures, just cash for your business.
If you're trying to figure out whether your situation fits, the best version of that conversation happens early, before the bank account sets the deadline. You can reach us at hello@curious.vc.
And if you haven't read A Bridge Too Far, that's where this thinking started. The bridge rounds were the inevitable extension of hope. Most of what this essay argues follows from that one sentence.
If your company has renewing revenue, customers who rely on the product, and gross margins above 60%, it is likely sellable, and a sale will return more than a wind-down. The right question is whether you have tested the sale option before defaulting to dissolution. Two weeks of conversations with buyers who work in your ARR range costs almost nothing and can answer the question definitively. A wind-down that happens without testing the alternative is a decision by default, not by choice.
Startup failure rate estimates vary by study, stage, and definition of failure, but the more useful framing is what is inside the failure bucket. A meaningful share of companies that wind down are not broken products; they are businesses that ran out of time or patience inside a fund model that needed faster growth. That distinction matters because a company with real revenue and real customers is often sellable even when it is not fundable. The aggregate failure rate does not tell you whether your specific company is one of those.
An acquihire is a talent transaction: a larger company hires your team, often with a nominal payment for assets, and the product typically shuts down. An acquisition treats the company as a going concern: the product continues, customers keep their accounts, and the buyer pays for the business as a business. Founders in a distressed position sometimes accept an acquihire offer without testing whether an outright acquisition was possible. The two conversations are different, and having both is worth the time.
A sale in this position does not require a formal auction or an investment bank. The practical version is: identify two or three buyers who buy software companies in your ARR range, reach out directly, share financials under NDA, negotiate a letter of intent, and run diligence. A buyer like Curious closes within 60 days and pays cash from its balance sheet, which matters when time is the constraint. A broker can help if you do not know which buyers are active in your size range, but the success fee of 10% to 15% is real money on a smaller deal, so weigh whether the broker adds enough to cover it.
Liquidation preferences mean investors get paid before founders and employees see any proceeds. On a low-price sale, the preference stack can absorb most or all of the purchase price, leaving common equity with nothing. The way to check is to run the waterfall: take the likely sale price, subtract the preference stack in order, and see what remains. Sometimes the number is discouraging. But investors facing a zero from a wind-down are often willing to negotiate preferences to get a deal done, and founders are sometimes surprised by how the math actually lands. Talk to your lawyer before you sign anything.
The core factors buyers look at are renewing revenue, gross margin above 60%, customers who would notice if the product disappeared, and a business that can operate without the founder day-to-day. Deal-killers include single-customer concentration above 30% of revenue, unresolved IP ownership from contractor-developed code, and a preference stack that leaves no room for common equity at a realistic sale price. A company does not need to be growing fast to be sellable; it needs to have real customers paying real money for something they depend on.
Yes. Curious is a long-term holding company that buys and grows software companies with empathy. We are vertical agnostic and not scared of messy situations. If a company has $2M to $10M in ARR and renewing revenue, we will have the conversation regardless of whether the growth rate fits a venture model. We are backed by operator LPs with cash on our balance sheet, so there is no financing risk and we close within 60 days.
Curious is a long-term holding company that buys and grows software companies with empathy.
Join the Curious Dispatch to receive our latest thinking, news, and opportunities within our portfolio.
