PubMatic

Drive measurable advertising performance with PubMatic's AI-powered, agentic platform — built for buyers, publishers, and commerce media across CTV and mobile.

PubMatic syncs upload the users in an audience segment to PubMatic over SFTP so you can target them in advertising campaigns. Each sync registers your segment with PubMatic automatically, then uploads the matching users, so there's no separate taxonomy sync to schedule ahead of time. If your model is missing the identifiers PubMatic matches on, you can add emails, mobile ad IDs, and IP addresses with Match Booster.

Supported syncing

Both sync types send only changes: each run uploads the users added to or removed from your query result since the last run, as separate addseg and removeseg files. For more information about sync modes, refer to the sync modes docs.

Connect to PubMatic

To connect PubMatic, you need an SFTP account and a Data Partner ID (DPID). Your PubMatic Account Manager provides both during onboarding, in an email that also includes the SFTP hostname and the directories to upload to.

Enter these credentials when you create the destination:

• SFTP Host: the regional PubMatic SFTP hostname from your onboarding email. Use the region closest to your user base for fastest ingestion.
• SFTP Port: defaults to 22.
• Username and Password: your PubMatic SFTP credentials.
• Data Partner ID (DPID): your PubMatic-assigned data partner ID.
• Data File Path: the directory where audience data files should land, typically Reports.
• Taxonomy File Path: the directory where segment registration files should land, typically taxonomy. It must be different from the Data File Path, because PubMatic separates the two file types by folder.
• HEM Data File Path: the directory where hashed email files should land, typically hem. PubMatic requires hashed emails in their own directory. Fill this in if you send hashed emails with either sync type.

Sync configuration

Pick the sync type by the identifiers you send. Use Audience (multiple identifiers) for mobile ad IDs, IP addresses, partner IDs, or several identifier types in one sync, and to use Match Booster. Use Hashed email (HEM) audience when you send only hashed emails and some are MD5 hashes, because Audience (multiple identifiers) sends SHA-256 hashes only.

Both sync types register your segment with PubMatic, then upload the matching users. You configure the segment on the sync itself, so each sync targets a single segment.

Register a segment

PubMatic ingests user data only for segments that already exist in its taxonomy. Both sync types register their segment from the fields below before uploading any users, so a new segment always exists before its data arrives.

Hightouch re-registers the segment on the first sync run, on a full resync, and whenever you change any of these fields. That way edits such as a rename, a new price, or a new segment ID reach PubMatic on the next run without a separate step.

Audience (multiple identifiers)

Audience (multiple identifiers) syncs upload one segment's users keyed on any mix of identifiers. Each run uploads one gzipped TSV file per identifier type: hashed emails go to your HEM Data File Path, and every other type goes to your Data File Path with its PubMatic UID type in the last column. This sync type replaces the Audience data sync type, which sent one identifier type per sync and is no longer offered for new syncs.

Record matching

Map each identifier column to one of the identifier types below. Hightouch checks and normalizes every value before it uploads the value:

Partner IDs keep their exact value. The only exception is a value containing a tab or line break, which would split a row of the file, so Hightouch drops it. A row with no identifier in a usable format is rejected.

Hashed emails require a HEM Data File Path on the destination. A sync that maps an email without one fails before it uploads anything. PubMatic supports hashed emails in the U.S. region only.

Match Booster can boost the email, mobile ad ID, and IPv4 address identifier types. Partner IDs can't be boosted.

Configure a custom partner ID

Map Custom partner ID when PubMatic assigned your identity partner a UID type that isn't in the list above. Enter that number in Custom UID type, for example 52. A sync that maps a custom partner ID without a numeric Custom UID type fails before it uploads anything.

Each sync sends one custom UID type. To send IDs from a second custom partner, create another sync.

Choose target identifiers

Use Target UID type to tell PubMatic which identifiers to resolve hashed emails into: Cookie, MAID (IDFA/AAID), or Cookie and MAID. It applies only to hashed emails. Leave it empty to use the default target configured with your PubMatic Account Manager at onboarding.

Field mappings

PubMatic needs a country on every row. Map a column with the user's country to Country, or map a static value such as US. A sync without a Country mapping fails before it uploads anything.

There's no separate IP address field. An IPv4 address you map as an identifier fills the IP column of its own file.

Hashed email (HEM) audience

Hashed email audience syncs upload users keyed on a hashed email address instead of an advertising identifier. PubMatic resolves the hashed emails into the identifiers you target. Each run uploads a gzipped TSV file to your HEM Data File Path.

Configure email hashing

Use the hashing options to tell Hightouch whether your mapped Email column is already hashed:

• If the column holds raw email addresses, Hightouch normalizes each address and hashes it for you.
• If the column is already hashed, Hightouch sends the values as is.

Then set Which algorithm are your emails hashed with? to SHA-256 or MD5. This sets the algorithm Hightouch uses when it hashes raw emails, and tells PubMatic which hash type you're sending.

Choose target identifiers

Use Target UID type to tell PubMatic which identifiers to resolve the emails into: Cookie, MAID (IDFA/AAID), or Cookie and MAID. This field is optional. Leave it empty to use the default target configured with your PubMatic Account Manager at onboarding.

Field mappings

Map your source columns to PubMatic fields:

Tips and troubleshooting

Behavior and limitations

Hightouch doesn't remove a boosted identifier that drops out of a user's Match Booster enrichment while the user stays in the segment. When a user stays in the segment, Hightouch removes only the identifiers that leave your model's mapped columns. Ask your PubMatic Account Manager how long PubMatic keeps segment membership from SFTP uploads.

Validate your setup

After your first sync completes, verify the upload in PubMatic:

• Check that the files landed in the configured directories using your SFTP client. Hashed emails land in the HEM Data File Path, every other identifier type in the Data File Path, and the segment registration file in the Taxonomy File Path. Audience (multiple identifiers) syncs write one file per identifier type, with the PubMatic UID type in the last column of each row.
• Hightouch uploads files with a .UPLOADING suffix and renames them once the transfer completes, so only fully renamed files should appear.
• Ask your PubMatic Account Manager to confirm that PubMatic ingested the files. Data files reference a segment by its segid, so PubMatic drops data for any segment it hasn't finished registering.

Common errors

To date, our customers haven't experienced any errors while using this destination. If you run into any issues, please don't hesitate to reach out. We're here to help.

Live debugger

Hightouch provides complete visibility into the API calls made during each of your sync runs. We recommend reading our article on debugging tips and tricks to learn more.

Sync alerts

Hightouch can alert you of sync issues via Slack, PagerDuty, SMS, or email. For details, please visit our article on alerting.