Inflated ARR in AI startups: how to truly evaluate them

The context: why AI startup ARR metrics are under scrutiny

In May 2026, TechCrunch published an in-depth analysis on a practice that is now widespread in the AI ecosystem. Some startups inflate their Annual Recurring Revenue metrics to build convincing growth narratives. Furthermore, the venture capitalists funding them are often aware of this distortion. However, they continue to disseminate those numbers publicly.

The phenomenon isn't new in the tech world. However, in the age of artificial intelligence, it takes on particular contours. In fact, AI startups' business models are often hybrid: they combine software licenses, professional services, API credits, and pilot contracts. Consequently, the very definition of "recurring revenue" becomes elastic and manipulable.

For Italian SMEs operating in the B2B or retail sectors, this scenario has direct implications. In particular, it concerns anyone considering adopting AI tools, choosing technology partners, or integrating third-party platforms into their workflows.

The numbers that matter: how ARR is distorted in the AI sector

ARR, or Annual Recurring Revenue, measures annualized revenue from recurring contracts. It is a standard metric in traditional SaaS. However, in the AI context, it undergoes various forms of manipulation.

The first technique is aggressive annualization . A startup with a monthly contract of 10,000 euros declares an ARR of 120,000 euros, even if the contract only lasts three months. Therefore, the number appears solid, but it does not reflect any multi-year commitment from the client.

The second technique involves the inclusion of professional services . Consulting, implementation, and training are counted as recurring revenue. By definition, they are not: they end with the project.

The third distortion involves the pilot contracts . Some startups even count proof-of-concept projects as ARR, which often do not convert into definitive contracts. Therefore, the actual conversion rate remains hidden from external observers.

According to an analysis by Gartner on AI startup metrics , over 40% of companies in the sector use non-standard definitions of ARR in public communications. Furthermore, less than 20% clearly specify which revenue components they include in the calculation.

Strategic reading: what it means for those who need to choose an AI supplier

B2B SMEs often find themselves evaluating AI solution providers without the analytical tools of venture capital funds. However, there are readable signals even without access to certified financial statements.

The first red flag is about customer base breakdown . A startup with 500 low-ticket paying customers is structurally more solid than one with 5 enterprise clients on a pilot contract, even if the declared ARR is identical. Therefore, it is appropriate to ask how many active customers pay monthly, not just the total number.

The second element to check is the Net Revenue Retention (NRR). This metric measures how much existing customers spend over time. An NRR above 100% indicates organic expansion. Conversely, an NRR below 90% signals high churn, regardless of the declared ARR.

The third criteria is the contract structure . Prepaid annual contracts show real commitment from customers. On the other hand, monthly agreements that can be cancelled at any time significantly reduce the strength of recurring revenue.

In this sense, Harvard Business Review dedicated an analysis to assessing the stability of AI vendors. The article highlights how operational due diligence is now more important than the supplier's media reputation.

The hidden risk for SMEs: dependency on fragile platforms

Adopting an AI tool from a startup with inflated metrics exposes SMEs to real risks. In particular, the main risk is service disruption . If the startup runs out of funds or is acquired, existing contracts can be terminated with minimal notice.

We at SHM Studio we have seen this scenario happen in at least three cases over the past eighteen months. Client companies had integrated third-party AI tools into their processes of Digital marketing and SEO . Afterwards, the suppliers shut down or pivoted their product. As a result, an urgent migration became necessary, bringing unexpected costs and downtime.

Beyond this, there is a risk of tech lock-in Some AI platforms build deep dependencies into business workflows. Therefore, switching vendors becomes costly even when the service deteriorates. Thus, the initial vendor assessment must also include the ease of exiting the contract.

, there is reputational risk. Publicly associating with a startup that later turns out to have inflated its metrics can damage the credibility of the client company, especially in B2B contexts where trust is a critical asset.

Operational implications: a checklist for evaluating AI vendors

Based on these elements, it is possible to build a structured evaluation process. Below are the main criteria we suggest applying before signing any contract with an AI provider.

• Request the ARR breakdown : how much comes from prepaid annual contracts, how much from monthly ones, and how much from non-recurring professional services.
• Verify the number of active customers : not just the historical total, but how many are paying today and how frequently.
• Ask for the Net Revenue Retention : a figure above 100% is a positive sign of product solidity.
• Analyze the financial runway : how many months of operations current liquidity guarantees, regardless of reported revenue.
• Evaluate exit clauses : notice periods, data portability, API availability in case of service closure.
• Check customer base concentration : if 50% of ARR comes from one or two clients, the volatility risk is high.

These criteria apply both to choosing tools for the AI management within the company , and to the selection of platforms for google ads campaigns or LinkedIn campaigns with smart automation components.

: what nobody says openly: the role of VCs in building the myth

TechCrunch's investigation raises a question that goes beyond creative accounting. Venture capitalists are not unwitting victims of these practices. On the contrary, they often actively encourage them.

The mechanism is simple. A fund that has invested in a startup is interested in maximizing the valuation at the time of exit. Therefore, spreading growth narratives based on generous ARR serves to prepare the ground for subsequent rounds or acquisitions at high multiples. Thus, the secondary market absorbs valuations that would not hold up to rigorous analysis.

According to McKinsey in its State of AI 2025 , the pressure to show rapid growth has led many organizations to prioritize vanity metrics over indicators of real value. Furthermore, the competition for talent and media visibility amplifies this effect in the AI sector more than in any other tech vertical.

For SMEs, this means operating in a market where public information about vendors is structurally distorted. Therefore, independent due diligence is not an accessory option: it is an operational necessity.

The SHM Studio perspective: how we guide clients' technology choices

SHM Studio supports Italian SMEs in selecting and integrating digital technologies, including AI solutions applied to web development , SEO copywriting and marketing automation. In this context, critical vendor evaluation is an integral part of our consulting process.

In particular, when we evaluate tools to integrate into customer workflows, we systematically apply the criteria described in this article. Furthermore, we monitor the financial stability of suppliers over time, not just at the time of onboarding. This way, we reduce the risk of having to manage urgent migrations or unplanned service interruptions.

For companies considering the adoption of AI tools or wanting to review their tech stack, the SHM Studio blog offers regular updates on these types of dynamics. Furthermore, the team is available for a personalized evaluation through the page contacts .

Finally, for those who manage activities of Digital marketing with AI components, we suggest periodically reviewing contracts with automated platform suppliers. Therefore, it is useful to set internal alerts when a vendor changes pricing, modifies APIs, or announces funding rounds with valuations that are anomalous compared to declared revenues. These signals, read together, offer a more reliable picture of an AI startup's real health than any press release.