The Authorized Buyers Real-time Bidding system has a maximum rate of callouts,
expressed in Queries Per Second (QPS), that can be sent to a bidder's servers.
Pretargeting matches that exceed the limit are dropped. The limit serves the
following purposes:

- Protect bidders from receiving more requests than their servers can handle.
- Conserve Authorized Buyers resources when the bidder isn't interested in receiving all eligible callouts, for example, if the bidder is in test mode.

> [!NOTE]
> **Note:** Bidders use test mode to make sure the bidder endpoint responds within the deadline before it receives live traffic. The response deadline ranges from 80 to 1000 ms, depending on format and auction type; check the `BidRequest.tmax` field in the bid request for the exact value. The bidder can't win impressions in test mode, so the type of bid, such as "no bid", doesn't matter as long as a response is supplied.

## Quota limit

The limit is configured and enforced per bidder server URL, technically per
trading location-URL pair. However, bidders shouldn't use the same server URL in
multiple trading locations. The limit can be configured by a Technical Account
Manager (TAM) or by the bidder through the [Authorized Buyers API](https://developers.google.com/authorized-buyers/apis/v1.4/accounts).

When you change the limit through the API, the total quota for all the account's
URLs can't exceed the total quota configured internally by the TAM.

The maximum quota limit for your bidder also depends on your recent spend. Your
effective quota limit before error throttling is the minimum of the configured
quota and spend-based quota. You can view your spend-based quota on the RTB
Graphs dashboard.

The Callout Quota System attempts to not exceed the requested quota limit per
bidder server URL. More than 15K QPS, the system generally adheres to within 10%
of the QPS limit, and more than 45K QPS, it generally adheres to within 5% of
the limit---often 1 to 2%.

> [!IMPORTANT]
> **Key Point:** If pretargeting matches exceed the requested quota limits, excess bid requests are throttled.

Sometimes bid requests you're likely to respond to are prioritized based on your
recent bidding history. Here are some features Google uses to identify requests
you're likely to bid on:

- Publisher IDs
- Environment type
- Ad format

Programmatic Guarantee requests are always prioritized.

The Callout Quota System is less accurate at small volumes. The requested quota
limit can be exceeded if the requested limit is about 1,000 QPS or less. In such
cases, the requested quota can be doubled or occasionally more than doubled. If
you operate with small quota limits, you should consider reducing your quota
limit to account for these inaccuracies.

## Callout spillover

Cross-North America callout spillover helps maximize the number of bid requests
you receive within your QPS quota limit. This feature only applies to bidders
with server URLs in North America.

There are two Real-time Bidding (RTB)
[trading locations](https://developers.google.com/authorized-buyers/rtb/peer-guide#trading-locations) in
North America. When a callout is sent to one of these North America trading
locations and RTB detects that your bid request quota limit is already met
there, it transfers the callout to the other location, and tries to send a bid
request to your server URL there.

For example, if RTB is processing a callout in the North America East Coast
trading location and detects that your East Coast server URL's quota limit is
already met, RTB then transfers the callout to the North America West Coast
trading location and tries to send a bid request to your West Coast server URL
instead.

If your quota is full in both locations, the callout is throttled.

> [!WARNING]
> **Warning:** This feature might not work as expected for bidders with geolocation-based DNS configurations. See the [Best practices](https://developers.google.com/authorized-buyers/rtb/practices-guide#submit-static-dns) guide for recommendations on DNS configuration.

## Error throttling

The Callout Quota System uses error throttling. It sends less callouts than it
can if the bidder doesn't respond to bid requests in time or if the bidder's
responses are invalid. Note that error throttling:

- Triggers when Google deems the error rate too high for the bidder endpoints.

- Gradually lowers the outgoing QPS until the error rate or number of bid
  requests drops to an acceptable level.

- Is unrelated to the quota limit.

- Adjusts within minutes to changing error rates.