Adobe Experience Platform (AEP/CDP)
This guide will walk you through sending audiences from Adobe Experience Platform (AEP) to a Koddi-provided S3 bucket. Follow these steps to connect Adobe, select audiences, and schedule exports. This guide uses the file-based delivery model described in Sending Audiences to Koddi.
Before you begin
Get these connection details from your Koddi team:
| Detail | What you need |
|---|---|
| S3 location | Bucket name, AWS region, and assigned folder path |
| Authentication | Role details or scoped S3 credentials for the selected Adobe connection |
| Identity | The customer identifier and export column name agreed for audience matching |
| Audience selection | Audiences approved for activation |
| Export schedule | Supported cadence and expected time until updates are available for ad delivery |
You do not need to create an S3 bucket or provide Koddi with an Adobe login.
Step 1: Connect Adobe to the S3 destination
- In AEP, open Destinations > Catalog.
- Select Koddi Audiences S3 if a private destination has already been configured for your organization. Otherwise, select Amazon S3.
- Create a connection named Koddi Audiences S3.
- Enter the authentication details supplied by Koddi.
- Enter the supplied bucket name and folder path.
- Test and save the connection.
Use the authentication method supported by the selected connection. Keep credentials in Adobe's connection settings, not in exported files.
Follow Adobe's Amazon S3 connection guide for the connection fields and required permissions. A private destination must exist before you can select it; naming an Amazon S3 connection does not create a custom connector.
Step 2: Select your audiences
Activate the approved people audiences on the connection. Record each audience's name and immutable ID, plus the Adobe organization, sandbox, and audience namespace, and share these with your Koddi team.
Start with one test audience. Use audience IDs for mapping; display names can change. Read Koddi Dynamic Targeting for how audiences are used in campaigns.
Step 3: Map the customer identifier
In the mapping step, select the customer identity agreed with Koddi. Use the agreed export column name, for example:
| AEP source | Export column |
|---|---|
| Customer loyalty ID | userguid |
The identifier must match the identity used during ad delivery. Preserve it as a string, including case and leading zeros. Do not substitute an Adobe profile ID or ECID unless that is the agreed identity.
Export only the agreed identifier and any separately approved fields. For partner identities, see Using an Identity Resolution Partner, including the distinction between user.guid and user.id.
Step 4: Choose your export option
| Option | Schedule | Use |
|---|---|---|
| Full snapshots | Daily, weekly, monthly, or once, where supported | Export everyone currently in each audience |
| Initial full load, then incremental exports | Every 3, 6, 8, or 12 hours, or daily | Export an initial audience baseline, then subsequent changes |
Choose daily full snapshots for the simplest setup. Select incremental exports only after agreeing the file format, removal handling, and supported refresh timing with your Koddi team. A missing person in an incremental file does not mean they have left the audience.
Subdaily intervals apply to incremental exports. Weekly/monthly full schedules depend on the destination and audience type. Where available, use After segment evaluation to export after audience evaluation completes.
Follow Adobe's audience activation and scheduling guide. For fixed schedules, use the timezone shown in Adobe; convert business times to UTC where required.
More frequent exports do not automatically mean more frequent ad-targeting updates. Confirm the agreed availability time before choosing a cadence.
Step 5: Set the file format and naming
Use CSV with a header row, initially without compression, unless your Koddi connection instructions specify otherwise.
Example full-snapshot file:
userguid
customer-000123
customer-000456
customer-000789Include the audience ID and export date/time in file names. Preserve Adobe's split-file suffixes and use the assigned S3 path. Do not overwrite previous exports with static file names.
Use the agreed completion handoff for each export. If the handoff requires a manifest or ready marker, follow the supplied specification; Adobe does not automatically produce a custom Koddi marker. Share the successful export-run details and all file parts for the pilot.
Updating or stopping exports
Change schedules and audience selection in the connection's activation settings. Tell your Koddi team before changing identity mapping, file format, or export mode.
Notify Koddi when an audience is retired or unmapped. Stopping exports alone does not remove previously delivered membership. Use the agreed process for privacy deletions or consent withdrawals that need faster handling than the export schedule.
Supplement: AEP editions and existing connections
AEP is a cloud service. Compatibility depends on your Adobe entitlements and connection configuration rather than AEM-style version numbers.
| Your setup | What to do |
|---|---|
| Real-Time CDP Ultimate with an existing private destination | Select Koddi Audiences S3 and follow the steps above. Ultimate supports private Destination SDK authoring, but the private connector must be configured separately before activation. |
| Other Real-Time CDP editions | Use the existing Amazon S3 destination when available under your Adobe contract. Confirm access with your Adobe administrator. |
| Existing S3 connection | Create a separate test connection using the Koddi-provided bucket and access details. Verify identity mapping, file naming, and export mode before replacing the old flow. |
Updated about 2 hours ago

