The pay-per-record model with optional enrichments creates predictable costs, but potential users cannot evaluate affordability without understanding their likely usage patterns. An interactive calculator on the landing page that lets users input their needs (platform, query frequency, whether they need transcripts/comments) and see real-time cost estimates would remove a major conversion barrier.

The business context indicates the company serves AI labs with custom datasets like "50k cooking videos featuring hand-object interactions," but this human-in-the-loop pipeline capability isn't visible in the API product. Creating pre-built, ready-to-download datasets for common training scenarios (action recognition, scene understanding, audio-visual learning) would attract a new user segment while showcasing the labeling pipeline's value.

The API offers rich metadata and optional enrichments, but users unfamiliar with SQL query optimization or the specific schema structure may over-fetch data and incur unnecessary costs. Platform-specific guides showing how to write efficient queries (using selective field retrieval, appropriate filtering, avoiding redundant enrichments) would improve user satisfaction and reduce cost-driven churn.

The API serves technical users who can write SQL queries, but many potential customers (marketers, agencies, brand managers) need social media insights without the technical overhead. A dashboard product that provides pre-built visualizations, trend detection, and alerts would expand the addressable market while demonstrating the API's capabilities to technical buyers.

The flat $0.0005/record pricing is simple and transparent, but it doesn't provide incentives for users to expand usage or reward loyal high-volume customers. Implementing tiered volume discounts (e.g., 10% off after 1M records/month, 20% off after 10M) would increase revenue from existing customers while maintaining the low-friction entry point for new users.