Funding & VC

When ARR Stops Working: Investors Rewire How They Value AI

October's OpenAI revenue shock sent AI stocks lower. RTP Global's Kenza Zayani explains why ARR no longer captures how AI companies earn — and what replaces it.

By Amara Osei

3 min read

What To Measure When ARR Isn’t Enough
What To Measure When ARR Isn’t EnoughBuffik / pixabay

What's News

  • October reports that OpenAI's annualized revenue ran below estimates sent AI-linked stocks sharply lower.
  • ARR became standard in the mid- to late-2000s as software moved from license fees to subscriptions.
  • Two companies can both report $10 million in annualized run rate while resting on entirely different revenue quality.
  • Most post-ChatGPT AI companies run hybrid models: subscriptions plus consumption charges on API calls, tokens or compute.
  • Zayani expects investors to demand cohort-level retention and committed-revenue breakdowns going forward.

Reports in October that OpenAI's annualized revenue was running well below earlier estimates sent AI-linked stocks sharply lower — and put the industry's favorite metric, ARR, back on trial. Kenza Zayani, an early-stage investor at RTP Global in London, argues the episode reopened a question investors have circled for a while: is ARR still the right way to describe an AI business?

The limitations aren't new. ARR was built for software companies selling predictable annual contracts, which covers only part of how most AI companies actually make money. Through the early generative AI boom, growth was steep enough that few pressed the point. The OpenAI episode made that harder to ignore, partly because it showed how much weight public markets now place on a single private company's numbers.

For founders, the timing matters. A cohort of AI companies is reaching the growth-capital stage, and the questions they face in diligence are changing fast.

What was ARR built for?

ARR became standard in the mid- to late-2000s, when software shifted from license fees to recurring subscriptions. A customer signed for 12 months, revenue became predictable, and annualizing it was arithmetic rather than forecasting.

The first generation of post-ChatGPT companies inherited the metric without much examination — despite earning money very differently. Most now run hybrid models: a subscription base with consumption charges on top, billed against API calls, tokens processed or compute used. A growing number charge per outcome. The annualized figure blends revenue a customer has committed to with revenue that depends on what they use next month.

What gets presented as ARR is usually a revenue run rate: recent revenue multiplied out to an annual figure. That's a reasonable way to describe momentum, and a less reliable guide to the year ahead. The first question an investor tends to ask is what the figure actually counts.

Partner sales muddy it further. Companies selling partly through partners can report either the full value of those sales or only the share they keep — and the two conventions produce very different numbers for the same business. Both are legitimate, Zayani notes, but they're not necessarily comparable.

Why two $10 million startups aren't the same

Consider two companies both reporting a $10 million annualized run rate. In the first, most revenue is contracted, from customers whose usage has expanded steadily over 12 months. In the second, it rests on a handful of customers whose consumption spiked in the last quarter.

These are not similar businesses. The first has demonstrated that customers keep coming back and spending more. Nothing in the headline figure tells you that.

What should founders measure instead?

The answer depends on stage. At pre-seed and seed, the number still matters: a steep trajectory can point to market demand, to customers willing to pay, and to a team beginning to crack go-to-market. A run rate isn't generally expected to predict the year at that size.

Later, investors need more. What matters is the revenue behind the number:

  • Customer concentration
  • How much revenue is contracted rather than consumed
  • How predictable that revenue is
  • Whether gross margins hold as volume rises

Retention and expansion are the centerpiece, Zayani writes. They show how essential a product has become and whether it will still be in use in a few years. Churn is diagnostic in a way a growth figure never is: it can point to demand the product isn't answering, pricing that doesn't match the value, a need that was never really recurring, or a market shallower than it first looked.

What will investors ask next?

ARR has been a staple of how high-growth tech companies demonstrate momentum for so long that it's unlikely to disappear anytime soon, Zayani writes.

What's changing is how much weight that single number can carry on its own. She expects more investors to ask what a figure actually counts, how much of it is committed revenue, and what retention looks like by cohort.

Her bottom line: "The founders who have those answers ready will tell a much more convincing story than a single number ever could."

Original: crunchbase.com

Share this article:

More from Amara Osei

Amara Osei

Show full bio

Senior reporter covering companies, deals and strategy in entrepreneurship, startups and venture funding at Business Bearings.

2,248 articles

Related articles

« Previous articleNext article »