Yieldlab (SSP)

Upload tab-separated audience files (UserID + tagging attributes to add/remove) to Yieldlab’s ADITION SFTP server for programmatic tagging across the Virtual Minds ecosystem.

Prerequisites

Your ADITION Network ID (issued by Yieldlab; contact your account representative).
Yieldlab SFTP credentials (user + password). Defaults match the ADITION demo server; production customers receive per-network credentials.
Tagging keys and subkeys defined in Active Agent / ADITION before sync (manually via the UI or programmatically).

Authentication

Zeotap uses user + password authentication over SFTP. Paste the user and password into the credentials section.

Supported identifiers

Per the Yieldlab integration spec:

Yieldlab Cookie / ADITION User ID
ID5
NetID
MAID (GAID, IDFA, IDFV)

Multiple IDs can be sent per row as a comma-separated list (up to 10; the last ID must always be the stable ID).

Configuration

Field	Type	Required	Description
ADITION Network ID	Text	Yes	Numeric Yieldlab network ID (e.g. `3848`). Used as the filename prefix.
FTP Host	Text	Yes	SFTP host. Default: `vztd3.adition.com`.
Port	Number	No	Default: `22`.
Remote Directory	Text	No	Target directory on the server. Leave blank for the account default.

Target Settings

No target-specific fields — segment membership is expressed per-row via the tagging attribute columns.

Supported Operations

Sync Modes

Yieldlab is audience-only.

Audience Sync Modes

Mode	Supported
Add	Yes
Remove	Yes
Mirror	Yes
Upsert	Yes (equivalent to Add — ADITION treats add as idempotent)

Features

Field Mapping: Yes (row-level add/remove tag expressions)
Schema Introspection: No

How It Works

For each batch, Zeotap writes a TSV file and a .fin completion marker:


<ADITION-network-id>_<YYYYMMDDTHHMMSS>.tsv
<ADITION-network-id>_<YYYYMMDDTHHMMSS>.tsv.fin

The TSV carries a header row and, for each user, up to three tab-separated columns:


UserID	TaggingAttributesToAdd	TaggingAttributesToRemove

Rows without tagging attributes are allowed only when the UserID is a comma-separated list of ≥2 IDs (cookie-sync enrolment). The .fin marker is deleted by Yieldlab once the .tsv is ingested; its presence/absence signals whether processing has occurred.

Tagging attribute format


<key>.<subkey>=<value>
<key>.<subkey>

<key>: up to 32 chars, must match /^[a-zA-Z]\w+$/
<subkey>: up to 32 chars, must match /^\w+$/

Examples:


demographics.male
demographics.age_30_40
interests.productX=30

Supply these as a comma-separated list in the add_tags (or tagging_attributes_to_add) field and remove_tags (or tagging_attributes_to_remove) field on each row.

Mode-specific behaviour

Add / Upsert: only the Add column is populated.
Remove: only the Remove column is populated.
Mirror: rows flagged removed by the pipeline go into the Remove column; others into Add.

Limits

Daily user limit: 10,000,000 rows cumulatively per ADITION Network ID across all files. Rows beyond the limit are silently ignored.
Date rule: the filename date must match the current date; past/future-dated files are rejected.
Retention: TSV files remain on the server for 2 days before automatic deletion.

Troubleshooting

SSH dial / authentication failure

Verify the host (vztd3.adition.com by default) is reachable, the port (22) is open, and the credentials are current. Yieldlab rotates demo credentials periodically.

File uploaded but never processed

Check that the .fin sibling marker was uploaded successfully; Yieldlab ignores TSVs without a matching .fin. Also verify the filename date matches today (UTC).

Rows skipped (“no recognised identifier”)

Ensure your audience query projects at least one of: yieldlab_cookie, adition_user_id, id5, netid, maid, gaid, idfa, idfv, advertising_id, external_id, user_id.

Rows skipped (“no tagging attributes and only a single user ID”)

A single-ID row without any add/remove tags is invalid per Yieldlab’s spec. Either include ≥2 comma-separated IDs (cookie-sync enrolment) or provide a tagging attribute in add_tags / remove_tags.

Tagging keys/subkeys not matching

Before syncing, pre-create the matching <key> attributes and <subkey> segments in Active Agent / ADITION — uploads referencing non-existent keys are accepted but do not take effect.