November 21, 2026
Share

Metrics to Track Beyond Annual Recurring Revenue

Investors and startup founders are rethinking the use of Annual Recurring Revenue as a primary metric for evaluating AI companies, as hybrid monetization models and consumption-based fees complicate traditional valuation standards.

Metrics to Track Beyond Annual Recurring Revenue

Back in October, news that OpenAI’s annualized revenue was tracking well below prior estimates pulled AI-related stocks down steeply and sparked a fresh wave of scrutiny regarding AI valuations. This also brought back a debate that investors have mulled over for some time: Does ARR still serve as the proper metric for an AI business?

These shortcomings have been recognized for a while. Originally designed for software companies relying on predictable annual contracts, ARR only captures a fraction of how most AI firms generate income. Throughout the early phases of the generative AI boom, growth was explosive enough that few questioned the metric. The OpenAI situation made it harder to overlook, partly because it highlighted how heavily public markets now rely on the numbers of a single private entity.

For startup founders, addressing this now rather than later is crucial. A new wave of AI businesses is approaching the growth-capital stage, bringing shifts to the questions they face during due diligence.

What ARR was built for

Kenza Zayani of RTP Global.
Kenza Zayani of RTP Global.

ARR emerged as a standard practice during the mid- to late-2000s, as software businesses transitioned from license fees to recurring subscription models. With 12-month customer commitments, revenue became predictable, turning annualized figures into basic arithmetic rather than forecasting.

The initial wave of post-ChatGPT companies adopted this metric without much scrutiny, even though their monetization strategies differed. Today, most operate on hybrid models combining a subscription base with consumption-based fees tied to API calls, processed tokens, or utilized compute power. An increasing number charge based on specific outcomes. Consequently, an annualized figure lumps together committed customer revenue and unpredictable usage-based earnings from the following month.

Metrics presented as ARR typically function as revenue run rates—taking recent revenue and multiplying it out to an annual figure. While this offers a decent snapshot of momentum, it serves as a less reliable indicator for the year ahead. Investors naturally begin by asking what the metric actually measures.

Firms selling through partners may report either the full value of those transactions or just their retained share, leading to vastly different figures for identical operations. Both reporting methods are valid, but they do not always offer a direct comparison.

Consider two companies each reporting a $10 million annualized run rate. For the first, the majority of revenue stems from contracted customers whose usage expanded steadily over 12 months. For the second, revenue relies on a small group of clients whose consumption surged during the most recent quarter.

These represent fundamentally different business models. The first proves that clients consistently return and increase spending, yet the headline figure conceals this distinction.

Looking beyond the headline

When a single number fails to distinguish between such businesses, alternative metrics become necessary, depending heavily on company stage. At the pre-seed and seed levels, the raw number remains important. Rapid momentum can signal strong market demand, willing buyers, and a founding team starting to master their go-to-market strategy. At that scale, run rates are rarely expected to predict performance for the entire year.

Later stages require deeper insights. Essential factors include the revenue composition behind the metric: customer concentration, the split between contracted and consumption-based revenue, predictability, and whether gross margins stay stable as volume increases.

Retention and expansion take center stage. They reveal how vital a product is and whether customers will still rely on it years down the line. Churn offers diagnostic clarity that growth figures cannot match, pointing directly to unaddressed market demands, misaligned pricing, non-recurring needs, or a market smaller than initially perceived.

What investors will expect next

Because ARR has long served as a fundamental benchmark for high-growth tech momentum, it is unlikely to vanish any time soon.

However, the amount of weight placed on that single metric is shifting. Expect investors to increasingly scrutinize what a metric actually measures, how much of it represents committed revenue, and what cohort retention looks like. Founders prepared with those answers will present a far more compelling narrative than any single number could provide.


Kenza Zayani is an early-stage investor at RTP Global, based in London. She is drawn to ambitious founders tackling the most challenging problems, with a particular interest in AI from infrastructure to applications. Zayani has built experience in venture capital across stages, and at RTP Global, she brings her multistage experience and data-driven approach to company building with founders.

Illustration: Dom Guzman

Frequently Asked Questions

01Why is ARR sometimes insufficient for evaluating AI companies?

ARR was originally designed for traditional software businesses with predictable annual contracts. Most AI companies use hybrid monetization models—combining subscriptions with consumption charges based on API calls, token usage, or compute power—making standard annualized run rates less reliable for forecasting future performance.

02What metrics should investors look at beyond ARR?

Investors look deeper into the revenue composition, including customer concentration, the ratio of contracted versus consumption-based revenue, revenue predictability, gross margin stability as volume scales, and cohort retention and expansion rates.

03What does churn reveal about an AI business?

Churn provides diagnostic insights into product-market fit. It can highlight unaddressed customer needs, pricing models that do not align with product value, non-recurring use cases, or an overall market size that is shallower than initially expected.

This content is for informational and educational purposes only and does not constitute financial, investment, or legal advice. Investing in private companies, startups, and the AI sector involves substantial risk, including the potential loss of capital. Always conduct thorough due diligence and consult with a qualified financial advisor before making any investment decisions.
Share

Leave a Reply

Your email address will not be published. Required fields are marked *