A profile of livestream shopping app Whatnot, which added 20M accounts in 2025, facilitated the sale of ~$8B in goods, and is set for $1B+ in revenue in 2026 (Wall Street Journal)

The core change highlighted in the coverage is the rapid increase in user accounts and the associated transaction volume. Whatnot added 20 million accounts in a single year and facilitated roughly $8 billion of goods sales. This expansion directly translates to a larger number of concurrent users participating in live auctions and purchases.

Such growth imposes heightened demands on the database layer: frequent writes required storage must accommodate more user profiles, item listings, and transaction records. The real-time nature of auctions and the quick “breaks” described in the summary necessitate low-latency reads and writes to keep bidding states synchronized across participants.

Adapting to this scale entails tangible costs for engineering teams. Teams must provision additional compute and storage capacity, potentially implement sharding or partitioning schemes, and invest in robust monitoring and alerting to detect performance degradation. Operational overhead rises as the system becomes more complex to tune and maintain.

If the database infrastructure cannot keep pace with the incoming load, the user experience will suffer. Delayed updates to auction states could cause missed bids or failed “breaks,” eroding trust in the platform. Eventually, raw vertical scaling may hit limits, prompting the need for more sophisticated architectural changes or risking service instability.