Below I’m cross-posting an article I wrote for the launch of Topline Media, the media spin-out from Pavilion, a popular community for go-to-market (GTM) leaders. This article was originally published by Topline on 10/9/25.

Since this was written for the launch of a new publication, I made it somewhat more sensationalist than usual. It’s also shorter and tighter, without the usual deep-drives and asides.

The reception was not without controversy, in part because I touched the third rail by mentioning 996. Some hastily took that to mean, “some VC is telling portfolio companies to grind 996.” That I’m not a VC and never told everyone to blindly grind 996 seemed beside the point.

What I said was: if you’re in a winner-take-all market, then you need to win. Grinding 996 might be a part of that. But the point isn’t to work hard, it’s to win. You can grind yourself to death at a strategically doomed company and it won’t change much. (I’ve tried that. AMA.)

Here’s the article. Thanks to Sam Jacobs and Asad Zaman for inviting me to write it:

AI has created an era of haves and have-nots. AI-native companies with spectacular growth rates are grabbing all the attention, talent and money. Is this insanity? How long will it last? If you’re not among the ranks of the AI-native high flyers, how do you avoid becoming seen as a zombie, a living-dead SaaS company with uninteresting growth, little profit, and no future?

First, it’s important to understand the external environment. While the world may seem insane, it’s not. We are at the start of a major disruptive cycle on the order of client/server computing or the Internet. Such cycles come maybe every 20 years in my experience, just long enough for us to have forgotten what they feel like.

These technology disruptions create opportunities to build enormously valuable companies that will lead their markets for a generation. This is the system at work. It’s chaotic. It’s inefficient. It feels crazy when you’re in it. But always remember that from the wreckage of Webvan, Pets.com, and a hundred other dot-coms, sprung Amazon, Google, and Salesforce. Nobody said creative destruction came without casualties.

These cycles reflect the nature of venture capital. While fixer-upper private equity (PE) has always been about driving modest growth with ever-increasing EBITDA margins, venture capital (VC) has always been a hits business. I remember nearly a decade ago reading the prospectus of a top-tier fund which said that the internal rate of return (IRR) of their previous fund was 36%, but that dropped to 12% with the top two investments omitted. Most of what makes VC a great investment, worth the 10-12 year illiquidity, comes from a handful of fund-returning companies. While consistent base hits are the PE business model, the VC model is not just about home runs, but grand slams.

Viewed in this light, today’s ARR multiples seem much less insane. After all, if a company is going to be worth $20B at exit, it doesn’t matter much if you bought at a valuation of $50M or $80M. This is what drives the valuation insensitivity and fear-of-missing-out (FOMO) that we see today in AI. Moreover, if you remember that in greenfield platform markets, first place ends up worth 10-100x second place, and second 10-100x third, you should be willing to pay almost anything to get into the leader. And if you’re currently in second place, you should be willing to spend almost anything to get into first. Second prize really is a set of steak knives.

While some will question the durability of high-growth AI revenue, to many investors it’s surprisingly unimportant. Yes, a lot of the $100M in revenues (that a company hit in 18 months) may not recur, but 70% of something is worth a lot more than 100% of nothing. Thus, we are seeing a surprising lack of interest in traditional SaaS metrics and the very notion of ARR — particularly the recurring part — is starting to lose meaning. Increasingly, companies are just talking about revenue or product revenue because today’s pricing models (e.g., consumption, outcomes) no longer align to subscriptions and traditional SaaS metrics.

While we can’t help wondering how long this will last, that’s the wrong question. It will last until it doesn’t. Shorting bubbles is a dangerous business because the market can stay irrational longer than you can stay liquid. Eventually, some trigger will start an unwind cycle. And once again, we will learn that this time wasn’t different from all the times before.

If you’re an AI-native growth company, the strategy is simple: win. Take no prisoners. Grind 996. Grow faster than your competitors, blunt all attempts to overtake you. In the words of Larry Ellison, it’s not enough that you win, all others must lose. Hire people who are so aggressive they make you uncomfortable. Think: “you want me on the wall, you need me on that wall.”

But what if you’re not? Per Jason Lemkin et al., you probably can’t raise new money. Even T2D3 (triple, triple, double, double, double) — a growth trajectory that takes you from $0 to $100M in seven to nine years — is no longer interesting to 80% of VCs. Instead of T2D3, we hear of Q2T3 (quadruple, quadruple, triple, triple, triple). We now measure time to $100M in ARR in months, not years. And, by the way, do it with a tiny team, driving ARR/head of $1M+.

That the bar has been raised so high is a mixed blessing because now there’s no kidding yourself. There’s no pitching a cloud story while still selling on-premises. The bluff factor has been eliminated. If you want to raise money at an AI valuation, then you don’t just need an AI story, you need an AI growth rate to match it. Clear and simple, but far from easy.

If 80% of VCs aren’t interested in talking to you, how might you win over the other 20%?

Hunkering down is not good enough. Particularly if hunker means something like 10% growth and 5% EBITDA at $30M in ARR. Financially, that business might be worth 10-20x FCF, so $15M to $30M. That’s not bad if you’re bootstrapped and you’re a founder who owns 100% of the company. But, even then, that works only if there is confidence that the $1.5M in annual EBITDA will continue. That is, that you won’t be disrupted by AI natives who vibe code your replacement app over the weekend. However, if the same business raised $50M in VC then it’s effectively worthless, because the entire business is worth less than preference stack.

So how do you create value? One word: growth. Growth is what takes you from an EBITDA-based multiple to a revenue-based multiple. Mathematically, a point of growth is worth about 2.3x a point of profit. One way or another you have to figure out growth.

But how?

1. Make growth at positive FCF the top financial goal. Note that this is not a strategy, but a constraint.
2. Build an AI story. Do an inside round, raise debt, or even cut traditional R&D if you need to, but you have to find money to build an AI product and story. If you get it right, it will not only enable current sales but increase your value at exit.
3. Be relentless in sales model optimization. You are fighting for your corporate life. This isn’t about arguing with the board about how much to invest in growth. You are highly constrained, but let those constraints drive creativity. Do market research. Do win/loss analysis. Get good at listening. Figure out what you can do to improve sales productivity. Often, that will be doubling down on a key segment. Or stopping in an unproductive segment. Or changing key assumptions in your sales model (e.g., SC to AE ratio, AE hiring/cost profile) that might have been heretical to consider in the past.
4. Consolidate the space. Investors who have “no money” for operational experiments often do have money for new strategies. If you’re competing with the usual suspects in every deal and everyone is struggling, then consolidate the space. It should increase both win rates and prices.
5. Fresh eyes. You might think you’ve tried everything already over the past few years. But have you? And if you tried something and it didn’t work, was that because it was a bad idea or because you didn’t execute it well? Beware false knowledge that blinds you to solutions. Or bring in fresh eyes to challenge your assumptions. Yes, it’s not going to be easy, and yes you’ve tried a lot already, but you need to look at things with fresh eyes to find fresh solutions.

In a world of haves and have-nots, you want to be a have. And the key to doing that, no matter how many times you’ve tried before, is to figure out growth.