This article is an opinion piece written by Maor Farid, founder and CEO of Leo AI, about the criteria that may govern the financing of AI companies in the coming period. His core thesis is that reaching $1 million in annual recurring revenue quickly is no longer sufficient evidence of a company’s strength, after previous experiences showed that some products can attract customers and then fail to retain them.
From Rapid Growth to Sustainable Value
Farid recalls the experience of 2021, when including AI in a presentation was sometimes enough to attract funding, while small and medium-sized companies rushed to purchase new products. But he observed that many products offered traditional functions with a superficial AI addition, or provided limited improvement in noncritical tasks. In his view, a 20% improvement in a task on which the business does not depend may produce a good demonstration, but it does not guarantee subscription renewal.
By contrast, he believes the strongest opportunities emerge when technology compresses a core process from weeks to minutes or delivers substantial financial savings. To develop Leo AI, Farid says that the company’s founder and his partner conducted interviews with more than 900 mechanical engineers, from entry-level employees to vice presidents, to understand the tasks that consume their time and for which they might pay to reclaim that time.
Three Elements for Building a More Defensible Company
- Impact that justifies the investment: The improvement must be significant enough to change how the organization operates, not merely accelerate a rarely used task.
- Industry expertise: Advanced models have become infrastructure available to almost any company through APIs, so Farid does not consider access to a good language model a sufficient commercial advantage. In his view, the advantage lies in understanding the domain and solving problems that general-purpose tools do not address.
- Customer-owned context: Industrial companies possess decades of engineering knowledge contained in old drawings or held by experienced employees, and this knowledge is not necessarily included in the foundational models’ training data. He believes that a product capable of incorporating this knowledge into the workflow becomes harder to replace.
What Will Investors Examine?
Farid expects upcoming AI company listings to impose greater scrutiny on business models. According to his argument, public-market investors will care about net revenue retention, gross margins, and whether the cost of supporting deployments declines as the company scales.
They will also focus on the expansion of customer spending after the initial implementation. Farid considers this expansion a closer indicator that the product has changed how the organization operates, because an increased budget means that the customer has tested the value and decided to allocate additional spending to it. He says that this type of question is already emerging in his financing conversations, after reaching $1 million in annual recurring revenue within a year became possible for some products without guaranteeing their continuation in the following year.
certi.news Analysis
The shift described in the article is not a new accounting standard, but a reordering of priorities: from measuring the speed of revenue acquisition to testing whether the product becomes embedded in the customer’s operations and generates growing spending with sustainable margins. This matters to startup founders, particularly those building thin layers on top of third-party models, because an insufficient customer list or inadequate retention metrics may not compensate for the absence of clear operational value.
However, this conclusion remains the view of an author associated with a startup in the field, rather than an independent market study. The source also provides no comparative data or specific names of upcoming listings, so its expectations regarding the next twelve to twenty-four months still require practical testing. According to his argument, capital will increasingly flow to companies that combine industry expertise, major improvements in critical business processes, and access to customers’ proprietary knowledge, even if their growth requires slower manual deployment than a self-service model.