API Throttling

A practical example

Example: your team syncs 40,000 contacts nightly, but the vendor allows 600 requests per minute.

At one record per request, a full sync takes over an hour, so you must batch records or schedule incremental syncs instead.

What to evaluate before investing

• Ask for the exact published rate limits per endpoint, not just a marketing-level 'generous limits' claim.
• Test whether limits are per account, per user, or per app key, since shared credentials can exhaust quotas faster.
• Request documentation on throttling responses, such as HTTP 429, and whether retry-after headers are returned.

Limitations and tradeoffs

Throttling is not a defect; even well-built APIs enforce it. The tradeoff is design effort: your team must build queuing, batching, and retry logic to work within limits rather than assuming unlimited throughput.