What is rate limiting? Definition and product impact
Rate limiting caps how many requests a client can send to a service within a given period, such as per second or per day. Once the threshold is reached, the service rejects extra requests, usually with HTTP status 429, and sometimes says when to try again. These limits protect the service from overload, abuse and runaway costs.
Definition written by Alex Dionisio, Senior Product Manager. Updated November 13, 2026.
Why it matters for a PM
A third-party API's limits set a top speed for your product: how long an import takes, how often you can sync, how many users you can serve at once. Your own product also needs limits to defend against credential stuffing or a client that sends too much traffic. Either way, these are product decisions: which quota for which plan, what message to show, what can wait in a queue.
Example
A marketing app sends text messages through a provider that caps throughput per second. During a large campaign, part of the sends come back with a 429. The team puts messages in a queue, sends them at the allowed pace, and shows an estimated completion time in the interface instead of an error.
Key points
- Status 429 means “too many requests”; a Retry-After header sometimes gives the wait time.
- Retrying with growing delays (exponential backoff) avoids making the overload worse.
- Quotas can depend on the customer's plan, which also makes them a pricing lever.
- Limiting your own sensitive endpoints, such as sign-in, slows down automated attacks.
Common mistakes
- Discovering a third-party API's limits in production, with the first large customer.
- Retrying instantly and repeatedly after a 429.
- Showing users a technical error instead of an explained wait.
Go further with Module
The courses and lessons that cover this concept:
Guide on this topic
- MCP vs API: the difference
When to expose an API, an MCP server or both, and what it changes for your product.