Разработка ПО
Менеджмент
Для компаний
Для стартапов
Архитектура ПО
Pharmaceutical giant CVS invested $40M in H1, a startup that owns a unique database of 10 million medical specialists. The company's CEO claims: while AI copies software, the real value is in exclusive data. We examine how this works and why investors pay for information, not algorithms.
Оглавление
What happened
H1 raised $40M in investment from CVS Health — one of the largest players in American pharmaceutical retail. The startup manages a database of 10 million doctors, researchers, and medical institutions worldwide. CEO Ariel Katz emphasized the paradox of the current market in an interview: AI is excellent at copying SaaS products, but is powerless against unique data. This asset is exactly what drove the major deal.
For context: CVS spends billions annually on marketing and sales of prescription drugs. Access to doctors' profiles, their publications, specializations, and professional connections makes it possible to build a promotion strategy more precisely. Investment in H1 is not help for a startup, but a strategic acquisition of an information advantage.
How this is useful for business
The deal demonstrates a new vector for valuing technology companies. Investors and corporations are increasingly paying not for code or users, but for data that cannot be reproduced in weeks. This applies to any niche with an expert community: lawyers, engineers, scientists, teachers.
For entrepreneurs, the signal is simple: building a product around data that you collect first means creating an asset with a defensive position. AI can write text or generate a report, but it cannot instantly gain access to a database that H1 has been collecting for years.
How to make money from this
H1's monetization model is a subscription for pharmaceutical and medical companies. Clients pay for API access to doctors' profiles and for the ability to analyze their publications and connections. This is recurring revenue with high margins: operating expenses are minimal, and the data is updated constantly.
A similar structure can be adapted for any expert community. The key condition is that the data must be unique and difficult for competitors to access. Partnerships with professional associations, conferences, and publishers are all sources of information that can be legally aggregated and licensed.
Business ideas
1. A lawyer profile platform with analysis of their practice, court wins, and publications. Monetization: subscription for law firms and corporate legal departments, $500-2000/month.
2. A database of design engineers with ratings, portfolios of completed projects, and client reviews. Subscription for construction companies and developers, $300-1500/month.
3. An aggregator of scientific researchers for universities and R&D departments with analysis of citations and collaborations. Licensing to institutional clients, $10-50K/year.
4. A catalog of business school professors with data on their publications, consulting practice, and expertise. Partnerships with recruiting agencies and executive search, $200-800/month.
5. A database of marketers and advertising agencies with cases, awards, and specialization. Subscription for brands looking for contractors, $400-1200/month.
Risks and limitations
The main threat is regulation. Medical data is regulated by HIPAA in the United States, and similar laws apply in the EU. Collecting and processing information about specialists requires legal cleanliness: consent to processing, anonymization, and transparent terms of use.
The second risk is scaling. The uniqueness of data works up to a certain point. If data becomes widespread, AI companies may start aggregating it from open sources. Protection means constant updating and adding new parameters that are difficult to parse automatically.
7-day action plan
Day 1-2: Choose a niche expert community that is poorly digitized. Analyze 5-10 open sources where these specialists are represented.
Day 3: Compile a list of 50-100 potential clients — companies that work with this community. Define their pain point: who needs contacts, ratings, analytics.
Day 4: Find 2-3 partners for data access — professional associations, publishers, conference organizers.
Day 5: Assemble an MVP database of 500-1000 profiles. Estimate the labor intensity of manual and automatic collection.
Day 6: Create a landing page with a value description and an application form. Define 2-3 pricing plans.
Day 7: Conduct 10 calls with potential clients. Collect feedback on willingness to pay and key features.
Original news: TechCrunch Startups · See other news in the news section.
