Revenue, Retention and Supernova Growth: How VCs Judge AI Start-ups

Whenever a tech start-up posts strong revenue figures for its product or service, the market generally reacts positively. However, this wasn’t the case with AI start-up Emergent. When the agentic vibe coding platform revealed it reported $100 million ARR in just eight months, it triggered an industry-wide debate.

The core of this debate was the calculation and interpretation of a metric called ARR.

Traditionally, the ARR of a company (especially in SaaS) stands for annual recurring revenue. It used to be the primary metric for investors to judge any start-up’s investability. ARR generally implies recurring, contractually committed revenue that investors can underwrite with some confidence.

In Emergent’s case, the ARR meant annualised revenue rate and was not calculated based on recurring revenue or contractual commitments. It was an annualised extrapolation of monthly figures based on recent usage.

While the debate rages on, it also opens a conversation around how investment metrics for AI start-ups are evolving. Venture Capital firms and Angel investors have started to look beyond traditional investment metrics like ARR before deciding to take a bet. Hence, an AI start-up has to signal healthy growth on multiple aspects beyond ARR to prove its investability to investors.

As Fundamentum’s Ashish Kumar puts it, “Investors are usually sophisticated. They will give some value to the ARR but it is never that the valuation will only be a multiple of ARR. It is just one metric that we would consider but there are other metrics to consider too.”

So, how have the metrics for AI start-ups changed and how do VCs look at it?

Evolved Metrics for AI Start-Ups

Beyond ARR, the metric that most investors focus on is the customer retention of the AI start-up. This metric is important because it signifies a company’s ability to sustain onboarded clients by providing desired services or products.

Customer retention is also critical for validating product-market fit and reducing customer acquisition cost (the cost to onboard new customers). Therefore, AI start-ups strive hard to keep a healthy retention rate in their business to project themselves as investible.

As per Anil Joshi of Unicorn Ventures, for a B2C AI start-up, a retention rate of “30-50% is considered healthy, which signifies high daily usage by consumers. Whereas in the case of B2B AI start-ups, a retention rate of 20% is good, as in business the usage may not be regular but need a basis.”

Fundamatum’s Kumar puts customer retention as a primary metric in his investment decisions. He said, “when investing in an AI start-up, the key metrics I focus on are retention over month one, three, six and twelve depending on the company’s stage.”

An extension of the retention metric are the DAU (or Daily Active Users) and MAU (or Monthly Active Users) of an AI start-up.

The metric measures the number of unique users interacting with a start-up's product within a day or a month. Investors pay close focus to these numbers as it is a key performance indicator for tracking growth and user engagement.

The DAU/MAU ratio, also known as "stickiness," is the percentage of monthly users active on any given day. For instance, If a start-up has 1,000 DAU and 5,000 MAU, the ratio is 20%, implying users are active about 6 days a month (20% of 30 days). Generally, a high stickiness ratio (more than 50%) signals the product is becoming part of a user's daily routine.

In Joshi’s view, the healthy ratio for a B2C AI start-up is 30-50% and around 20% for a B2B AI startup, which is equivalent to SaaS start-ups.

Chirag Gupta of 8x ventures says that while investing, he specifically spends time on what is actually the “daily active user, monthly active user and the weekly active user” of the start-up.

He said, “if you ask me today, for us, I think we spend more time on trying to figure out how much the product is being used on a daily, weekly or a monthly basis, than just purely looking at the ARR. Because if the product is not being used on a weekly or if not monthly basis, then the likelihood of that getting pushed out in the next budget (of an enterprise customer) is very high.”

Notably, for any kind of business, revenue (be it recurring or annualised) remains to be an essential metric to look at while determining its investability. But when it comes to AI start-ups, the treatment of these metrics differs slightly.

Some investors said that their expectation of revenue growth rate for an AI start-up is much more than that of a traditional SaaS company.

Squadstack CEO Apurv Agrawal explained, “Supernova growth has become almost like an expectation. Earlier for any SaaS start-up hitting a million-dollar run rate in the first 2-3 years of the journey used to be considered a good product market fit…Now I feel, it has all become 10x bigger, that in the agentic AI world you have to either grow extremely fast or die.”

But why are AI start-ups held to such high standards?

It is because for any traditional SaaS company, software and data used to be the moat (a structural barrier that protects a company's profitability). But with the advent of AI, these aspects have been democratised and the barrier to entry has been significantly lowered.

Now, this also means that for any particular use-case, there can be multiple AI players competing. Hence, to stand out an AI start-up must show at least 10x the growth.

Deepak Sharma, Co-Founder of India Accelerator explained it with the help of an example. He revealed that in early 2024, the firm had invested in an AI sales and marketing start-up, which was at a monthly revenue rate (MRR) of around ₹40- 50 lakh. The same start-up had grown to touch around ₹5 crores of MRR in October 2025, which led Sharma to participate in their next round as well.

Reaching ₹5 crores MRR from around ₹50 lakhs, in a span of around 20 months is approximately 10x growth.

However, Pranav Pai, managing partner of VC firm 3one4 Capital, argued that Annual Contracted Revenue (ACR) is a more meaningful metric when evaluating AI start-ups.

Annual Contracted Revenue (or Annual Contract Value - ACV) is a key subscription metric representing the average, normalised revenue generated from a single customer contract per year. It excludes one-time fees, focusing solely on recurring revenue, making it essential for comparing deal sizes, forecasting and measuring sales performance across different contract lengths.

According to Pai, “ACR reflects revenue that is contractually committed, based on signed customer agreements rather than extrapolations. This makes it a more reliable indicator of near-term revenue visibility and actual demand.”

ACR and ARR (recurring revenue) can be considered equivalent, since both metrics take into account only the contractually committed figures. However, in Sharma’s view the ACR of an AI start-up must be 10x the ARR of a traditional SaaS company.

Other Things That VCs Look At

Squadstack’s Agrawal stated that the AI start-up must establish a strong moat against the foundational AI companies. This means it must prove a clear differentiation to the investors, explaining why the problem it is solving cannot simply be added as a feature in the foundational models of the AI giants.

Secondly, Fundamentum’s Kumar looks at whether the product helps increase a customer’s revenue or reduce costs, since revenue-generating use cases are generally considered higher quality than cost-saving ones. Additionally, it is also important to assess whether the AI solution serves a customer’s core function or a non-core function, because in his view, solving core problems usually commands more attention from senior management and tends to be higher-quality revenue.

Lately, 3one4 ventures’ Pai noted headline metrics like ARR are frequently amplified in public discourse, especially on platforms like X, where they can contribute to hype around start-up growth. However, such figures, in his view, carry limited weight in rigorous investment decisions, where greater emphasis is placed on revenue quality, customer retention, margin structure and the sustainability of revenue streams.