How to Build a $1M AI Business in 90 Days

TL;DR

Start with two people: a builder who can turn an idea into a product within 24 hours, and a go-to-market partner who understands customers and social distribution. Aim to earn the first dollar by day 30 and reach $1 million in annual recurring revenue by day 90, using frequent releases, direct customer interviews, and clear measures of delivered value.

Transcript

A lot of businesses can really scale to tens of millions of dollars today profitably with AI. >> This is Alex, founder of Hicksfield, an AI company that hit 200 million annual recurring revenue in just 9 months faster than Slack or Zoom. And what he told me about starting a business today completely blew my mind. And you can copy his strategy, too.... Read More

Key Insights

• A two-person founding team can combine the essential capabilities for an AI startup: one builder who creates products within 24 hours and one go-to-market partner who understands the target audience, distribution, empathy, and social content formats.
• Daily product iteration helps an AI company discover frequent, meaningful workflows while adapting its interface. Higgsfield released product updates six days a week because users needed both ease of use and enough configuration to control creative results.
• The AI application market can reset every month because leading research labs release major capability updates. These changes may require application companies to substantially rebuild their products so customers can access the strongest possibilities offered by newly released models.
• Eight independent customer interviews identified the same unmet need for Higgsfield: creative professionals wanted control over camera effects and angles. The consistency of feedback across Hollywood-level directors and regional commercial producers gave the company a concrete feature to build.
• Higgsfield's early traction came from translating customer feedback into camera controls, then expanding the product with a library of visual effects. The company also hired four of the eight interviewees, bringing creators directly into its continuing product-development feedback loop.
• AI-native marketing agencies can produce advertisements within days and improve their margin profiles by using AI throughout production. Their growth reflected demand from brands that wanted a constant flow of social content and were interested in adopting AI-based production.
• A focused AI startup needs a nuanced customer segment because major technology companies are trying to provide AI broadly. The interview proposes serving tens of millions of potential users willing to spend a couple thousand dollars monthly by delivering sufficiently strong, specific value.
• Daily active users and average contract value are more useful to Higgsfield than monthly active users alone. Monthly activity can rise through viral effects without demonstrating frequent use, while daily engagement and annual customer spending more directly indicate recurring utility and value.

Questions & Answers

Q: How can two people start an AI business?

A two-person AI startup can divide responsibilities between a technical builder and a go-to-market partner. The builder should be capable of moving from an idea to a working product within 24 hours, using available databases, payment systems, and other infrastructure. The go-to-market partner should understand the intended customers, recognize their needs, and create social content formats that resonate with them.

Q: How can an AI startup reach $1 million ARR in 90 days?

The proposed target is to earn the first dollar by day 30 of product development and reach $1 million in annual recurring revenue by day 90. The approach begins with a two-person team, rapid product construction, and direct understanding of a specific paying audience. Frequent releases and customer feedback then help the company identify valuable, high-frequency workflows and improve the product quickly.

Q: Why should AI startups release product updates frequently?

Frequent releases allow an AI startup to test workflows, discover which use cases matter, and balance a simple interface with sufficient user control. Higgsfield iterated six days a week and released new product changes daily. Rapid iteration was also necessary because major model updates regularly expanded industry capabilities, sometimes forcing the company to substantially rebuild its application around the improved technology.

Q: How many customer interviews helped Higgsfield find its key feature?

Higgsfield interviewed eight creative professionals, and all eight independently reported the same missing capability: control over camera effects, camera angles, and related creative choices. The participants ranged from Hollywood-level movie directors to regional commercial producers. Their consistent, unfiltered feedback led the engineering team to build camera controls, which became the source of the product's initial traction.

Q: What caused Higgsfield's initial product traction?

Higgsfield gained initial traction by solving the camera-control problem identified through interviews with creative professionals. The engineering team implemented controls for camera effects and angles after hearing consistent demand from eight interviewees. The company delivered those controls in March and followed them with a visual-effects library in April, creating a product that gave creative users more control over AI-generated output.

Q: Why did Higgsfield move beyond its original mobile apps?

Higgsfield initially focused on mobile applications, but that direction was not working well because mobile-app retention was relatively low. The company changed course by repeatedly interviewing creative professionals and testing their reactions to AI-generated videos. Those conversations revealed a specific limitation, inadequate camera control, which gave the team a clearer product problem and a more relevant audience to serve.

Q: Which metrics does Higgsfield use to measure customer value?

Higgsfield emphasizes daily active users and average contract value. Daily activity reflects how frequently customers use the product, while average contract value indicates how much they are willing to pay annually. The company does not rely primarily on monthly active users because viral effects can inflate that number without proving frequent usage or demonstrating that the product delivers substantial recurring value.

Q: Why are AI-native marketing agencies an important market?

AI-native marketing agencies use AI across the production process and often try to bypass incumbent creative tools. Although early AI capabilities limited what they could produce, these agencies could improve their margin profiles and show clients that advertisements could be created within days. Their growth also matched demand from brands seeking a constant flow of social content and wanting to adopt AI.

Summary & Key Takeaways

• A compact AI startup can begin with one technical builder and one go-to-market partner. The builder should rapidly convert ideas into products, while the commercial partner identifies a specific audience and develops social content that resonates. Existing databases and payment systems make creating and selling an initial product faster than before.

• Higgsfield found its initial traction after interviewing eight independent creative professionals who consistently identified missing camera controls as their main problem. The company built those controls, added a visual-effects library, and hired four interviewees. Its team later became roughly 40% engineers and 40% creators, creating an internal product-feedback loop.

• AI application companies should pursue paying customers and cash-flow discipline from the beginning instead of optimizing primarily for broad monthly activity. Higgsfield emphasizes daily active users and average contract value because these measures better reflect usage frequency and delivered value. The proposed schedule targets first revenue by day 30 and $1 million ARR by day 90.