Both ideas are valid but they work differently in practice. Here's how I'd approach it:
Option 1 — Let the audience do the work
If your original prospecting audience has a condition like "has NOT converted", then the moment a conversion event hits the profile and streaming segmentation evaluates it, the profile exits the audience. AEP sends the exited status to FB, DV360, and LinkedIn automatically — they remove the user from the campaign. No separate exclusion audience needed.
This only works if your audience is on streaming segmentation (not batch). Batch evaluation runs on a schedule so you could have a lag of hours before the exit signal goes out. For conversion exclusions that need to be fast, streaming is non-negotiable.
Option 2 — Dedicated suppression audience
Build a "Recently Converted" audience — anyone who had a purchase/conversion event in the last X days. Activate that audience to all three destinations and set it as an exclusion in each platform's campaign settings. FB, DV360, and LinkedIn all support exclusion audiences natively.
This is actually the more reliable approach in practice because:
- It's explicit — you can see exactly who's being suppressed
- It works even if your prospecting audience logic is complex or batch-evaluated
- You can control the suppression window (7 days, 30 days, etc.)
What I'd actually do
Both together. Keep the prospecting audience condition clean so converted profiles exit naturally, AND activate a suppression audience as a safety net. The suppression audience catches any lag and gives you a clear audit trail. The cost is minimal — it's just one more audience activation per destination.
One thing to watch — for DV360 the audience match takes time to propagate after activation. Don't expect instant suppression on the DV360 side regardless of approach. FB and LinkedIn are faster but still not truly real-time. Set expectations with the client around a 1-2 hour suppression window at minimum.