Growth alone stopped being enough

Raising on top-line growth alone was a 2021 thing. For most of the last two years it just hasn't worked, and honestly, that's healthy. The companies raising well now are the ones whose unit economics stack up. We've always cared about this, so from where we sit it's less a shift than a return to sense.

One caveat up front: the frontier AI infrastructure players raising hundreds of millions are a different animal, and we'll come back to them. For everyone else, the fundamentals are doing the work.

The public markets set the tone

What happens in listed companies tends to trickle into private ones. The S&P 500 ran a 12.8% net profit margin in Q2 2025, ahead of both last year and the five-year average, with double-digit earnings growth three quarters running. Investors are paying for efficiency and durable margins. Profitable growth is the benchmark now, not a line founders trot out in a pitch.

Venture money, meanwhile, is pooling at the top. Mega-rounds of $100m and up have taken 60 to 80% of AI venture dollars in recent quarters, which leaves early-stage founders raising in a far more disciplined room. You have to show capital efficiency and a real path to profit. For most, the Rule of 40 is still the most useful single gauge.

What we actually look at

A few things tell us most of the story: the mix of recurring versus one-off revenue and how concentrated the customers are; CAC and how long it takes to earn back, alongside where gross margin is heading; how cash really moves through the business; and efficiency, ARR per head and the burn multiple.

Rough marks of a healthy business: CAC paid back inside a year, gross margins holding or improving, net revenue retention around 100% or better, a customer base that isn't one logo away from trouble, and ARR per head climbing (usually $150k to $250k at growth stage, the best above $300k) with a burn multiple under 1.0 once things settle.

The AI exception

None of this reads the same for the frontier AI model and infrastructure crowd burning hundreds of millions on compute, training and data centres. Their economics look more like heavy industry than software. Important outliers, but outliers, not the world most scale-ups live in.

Companies that get their unit economics right hold up better when things get hard, raise more easily, and look better to an acquirer. The founders who build their dashboard around CAC payback, burn multiple and ARR per head keep hold of the wheel. Our job is to help design for growth that lasts, which is sometimes slower now and almost always stronger later.