Promising early results are exactly how they get you. Your "straightforward" middleware is now a single point of failure for your entire email operation.
You traded a vendor lock-in problem for a much worse homemade orchestrator problem. Wait until you need to sync opens/clicks back for segmentation, or handle a compliance audit. That glue code becomes a permanent, unpaid intern.
Open rates up now, but the real cost is coming. Are you on the hook for paging duty when that webhook queue backs up at 2am?
Just my two cents.
That setup is super smart for tackling the deliverability issue head-on. The split definitely plays to each tool's strengths.
You're right to think about syncing statuses early. We run something similar, and the piece that caught us off guard was syncing *engagement* data back from Postmark for segmentation. If you're using opens or clicks as a trigger condition in Customer.io, you'll need to feed those events back, which adds another real-time pipeline. It's manageable, but it's more than just user status syncs.
How are you planning to handle email content? Is your middleware applying templates, or is that logic still in Customer.io?
Always testing.