The word “1 billion” is made from a crimped $100 bill.
getty
Validations from AI startups are rarely exciting, but the story of a billion-dollar AI company admitting that it was manually doing the work that enabled its “automation” product proves this point. According to The Verge, clients thought Fred, a chatbot, was taking meeting notes, but in reality, Fireflies.ai’s co-founders were tearing apart more than 100 meeting notes word-by-word and inputting them by hand, before writing a single line of code. Some questioned whether customers had been misled. Some questioned the startup’s ethics. But the more fundamental lesson is what many entrepreneurs experience: the unscalable, invisible effort required before reaching a milestone.
The first phase, which occurs before the funding round, valuation, and growth curve, is the most often unspoken phase of an entrepreneur’s journey. It is characterized by limited resources (tight budget) and the personal financial risks associated with being a founder. This is the phase where founders perform tasks that they hope to one day automate.
Fireflies.ai failed to gain traction because it pretended to have artificial intelligence (AI) capabilities. Rather, their product development process has been one of learning every detail of their customers’ workflows by personally performing the tasks until they are comfortable enough with automation. According to a Futurism report, the company was charging customers $100 a month for manual services, which only covered basic costs, even though it manually conducted more than 100 meetings. Manual testing methods such as Fireflies.ai are a typical way startups validate their products and did not violate any laws.
Small business owners need to assess whether they are “in love” with their product or “in love” with the problem their customers are having. One leads to tools. The other leads to companies large enough to reach $1 billion valuations.
Unscalable difficulties: What small businesses can learn
The Fireflies story is less a warning against fraud to small business owners and more an example of how customer verification works in startups. As reported by The VergeSam Udotong and Krish Ramineni initially introduced Fireflies.ai as “AI in meetings,” but the two founders actually ended up attending meetings on behalf of their product. Listen and take handwritten notes. Fireflies’ early use of a testing method known as the so-called “Wizard of Oz,” in which people behave as if a technology existed that had not yet been built, provided a way to learn the expected behavior of users before investing time in developing software that might miss the mark.
In today’s business environment, AI is frequently promoted as a shortcut to success. But the Fireflies example shows that technology cannot replace customer understanding. Fireflies.ai’s first question wasn’t “How do we create an algorithm?” Rather, they asked, “What does the perfect meeting transcription look like?” They manually crafted the answer through each call.
This process offers significant benefits to small and medium-sized businesses. This allows you to develop, test, improve, and validate new concepts without the need for extensive tools or high initial engineering costs. This allows companies to gauge the level of client interest, collect client feedback, and build services based on the experience clients have with the company. As the futurists saidthis was not done for quick revenue generation. Rather, it was a survival mechanism that provided insight into what value customers placed on products and what AI systems ultimately needed to deliver to meet those expectations.
Three Pillars of Customer-Focused AI
Small businesses can take away the fundamental lessons of Fireflies.ai without misrepresenting what they can offer their customers. The three elements of this pillar are applicable to today’s small and medium-sized businesses.
1. Manual before automation
Small business owners can implement manual operations to understand customer behavior before spending money on AI tools, software, and automation. By personally replying to customer support emails, following orders yourself, and handwriting posts to your social media accounts, you’ll know first-hand what your customers expect. Insight into customer expectations is the basis for intelligent use of automation. By using direct human interaction, rather than relying solely on chatbots or CRMs, you can establish the right prompts, workflows, and priorities.
2.Process until commercialization
The creators of Fireflies.ai didn’t just create transcriptions of meetings; they created a methodology to generate reliable and actionable notes from those transcriptions. During that process, we understood what kind of information was important, where to place it in the customer’s workflow, and how to organize follow-up actions. That methodology ultimately became the basis for AI systems. For small businesses, this means mapping customer journeys, improving onboarding processes, and defining optimal follow-up activities before automation takes place. Methodology provides the strategy and AI is the execution layer.
3. Transparency and reliability
Although Fireflies was much more implicit in its early testing methods (than what today’s customers expect), there are lessons that are very clear today. That means transparency is important. Customers responded positively when companies clearly communicated which parts of the experience would be automated and which parts would be performed by humans. An example of this type of communication is, “This is an automated response from us…If you need further assistance, our team is here to help.” Transparency creates opportunities for customers to provide feedback, ensures that AI tools do not negatively impact customer relationships, and improves overall customer satisfaction.
Table: Benefits and risks of manual vs. automated verification
IBAI: Business AI Institute
AI-powered small and medium-sized businesses in 2026
Looking to 2026 for small and medium-sized businesses that are successful with AI, it won’t be about how good their AI is, but how well they understand their customers. Fireflies.ai illustrates this concept that even the most powerful technological advancements are useless if the customer’s fundamental problems are not well understood. AI augments organizational expertise and human instincts, not replaces them.
The future of small businesses and AI has two components: the use of automation and the incorporation of human insight. Automation allows entrepreneurs to scale manual-intensive activities, streamline processes, and spend more time focused on building relationships rather than tackling mundane administrative details. AI helps organizations develop more consistent and effective operational systems. However, without a deep understanding of what your customers want from you, you cannot ultimately determine the level of success your organization achieves.
Before implementing new AI technology, small business owners should consider the question, “What is the least scalable action I can take for my customers today?” The answer to this question is usually an example of a task completed by a user (manually) that represents a possible solution for future workflows or automation.
final thoughts
The Fireflies.ai team didn’t build a company worth over $1 billion through AI. They built a company worth over $1 billion because they built it with empathy, a willingness to perform tasks manually, and a willingness to solve problems that others wouldn’t take the time to understand.
In other words, the technology used by small businesses only increases knowledge. However, it is not a substitute for the knowledge of small business owners.
