What is an API rate limit and why does it matter to a dealership?

Ask for the number, not the adjective

Replace 'real time' with a figure: how many requests, over what period, and what happens when the limit is reached: queued, dropped or failed.

Understand the business consequence

Translate the limit into store terms. If inventory synchronises four times an hour, the exposure is a fifteen-minute window of stale listings and pricing. Decide whether that is acceptable rather than discovering it later.

Check who bears the cost of exceeding it

Some agreements charge for volume above a threshold. Establish the unit, the price and whether a marketing campaign or a large data load can trigger it.

Test at your volume

Limits that are invisible at demo volume become visible at group volume. Ask for a reference operating at your scale.

Where limits show up in daily work

Rate limits appear as delays: inventory changes that take a while to reach the website, leads that reach the CRM late, or a bulk update that stops partway. Staff may describe these as the system being slow or flaky. Ask the vendor to log requests and any rejected ones so that the cause can be identified rather than guessed.

Designing around limits

A well-designed integration sends only changes, groups requests sensibly and retries failures without duplicating records. Ask how the vendor handles a burst, such as an overnight import or a large campaign response, and how it notifies you when the limit is reached. Schedule heavy jobs at quiet hours where possible.

Common mistakes about limits

Accepting the words real time without a number, forgetting limits when more rooftops are added, assuming both sides have the same limit and not checking whether limits differ by plan or fee tier are the typical errors. Record the agreed limits in the integration record with the date, and re-check them at each renewal.

Questions to ask both vendors

What are the limits for requests per minute and per day? Do they differ by connection type?

What is returned when the limit is reached? Are requests queued?

Who notifies us? Is there a charge for higher limits?

Confirm the answers from both sides, since one vendor's limit can determine how the whole connection behaves.