For CTV and Mobile devices
Frequency capping is a best-effort feature intended to help the automatic optimization of your campaigns.
Frequency Capping may not work properly for several reasons:
Lack of identifiers
Due to consent management, privacy options, or simply because an app or a site does not send a device identifier frequency capping will not work. This can be avoided by targeting only traffic that includes identifiers.
Data centers
When the same user comes through different data centers, even when we have their user ID.
Suppose a user sees an impression in Paris, that comes from SSP 1 with traffic coming from a US-based data center and another impression with SSP2 from the EU. In that case, we cannot do frequency capping because the information on impressions per user, which is managed in memory, is in different data centers.
Learn more about how our data centers are distributed.
Delay between purchase and information (caché)
There is often a delay between when we place the bid and when we win it and know the impression has been served. In many cases, this delay can be as high as one hour.
Within this period, the frequency capping may not be respected because the system keeps buying and only notices the "extra" impressions an hour later.
Monthly limit
Campaign level frequency capping has a limit in time in terms of days. If your campaign lasts for longer than a month only the last 30 days of the campaign will be considered toward frequency capping for the full campaign.
Mobile web
On mobile web, when the IDs sent in the bid requests are not unique per user. This can happen when the user comes through different publishers or publishers from different exchanges. If we cannot univocally identify the user, we cannot cap.