Skip to main content
Level 2
August 26, 2026
Solved

Using batch audiences in a journey triggered by streaming audience qualification

  • August 26, 2026
  • 2 replies
  • 36 views

Hi there, 

Is it possible to use a combination of batch and streaming audiences in a single journey (noting - both use active on edge evaluation method). 

I have created a journey with the following:

  1. audience qualification based on a streaming audience using events
  2. Node checks if the member is part of a batch audience
  3. One hour wait 
  4. Node checks if member is part of a streaming audience
  5. Email
  6. Five day wait
  7. Node checks if the member is part of a batch audience. 

Will this journey reliably work with the combination of streaming and batch audiences? 

Best answer by Pulkit_Jain_

@serena_stevens I think it will work structurally, but there's a timing nuance to flag: Node 2 checks batch audience membership immediately after the streaming trigger, before any Wait activity, so it reads from the batch profile projection which can lag up to 2 hours behind actual ingestion meaning it may not reflect the most current membership right out of the gate. To fix the gap at Node 2- insert a short wait before it.

Nodes 4 and 7, however, both occur after a Wait activity, so they actually read from the streaming (unitary) projection (freshness within ~15 minutes) regardless of whether the audience itself is batch or streaming — making them reliable.

 

2 replies

Pulkit_Jain_
Adobe Employee
Pulkit_Jain_Adobe EmployeeAccepted solution
Adobe Employee
August 26, 2026

@serena_stevens I think it will work structurally, but there's a timing nuance to flag: Node 2 checks batch audience membership immediately after the streaming trigger, before any Wait activity, so it reads from the batch profile projection which can lag up to 2 hours behind actual ingestion meaning it may not reflect the most current membership right out of the gate. To fix the gap at Node 2- insert a short wait before it.

Nodes 4 and 7, however, both occur after a Wait activity, so they actually read from the streaming (unitary) projection (freshness within ~15 minutes) regardless of whether the audience itself is batch or streaming — making them reliable.

 

Level 2
August 26, 2026

Thank you so much, that’s very helpful information. I’ll update the first check to a streaming audience!