What is split-testing inbox placement by segment?

You've probably run an A/B test on a subject line before. Split-testing inbox placement by segment works on the same principle, but instead of testing what you write, you're testing who you're sending to and watching how each group's placement changes as a result.

The idea is simple. Different subscriber groups have different engagement histories, and mailbox providers like Gmail and Outlook factor that into placement decisions. So if you send your campaign to everyone at once, you get one blended placement number that tells you almost nothing. Split by segment, and suddenly you can see exactly which audience groups are pulling your results up or down.

Two segments worth testing first

Engaged vs. dormant subscribers. This is the most revealing split you can run. Send the same campaign to your active openers (say, anyone who opened in the last 90 days) separately from subscribers who haven't opened in six months or more. The dormant group will almost always show worse placement. That's expected. What you're really looking for is whether their placement is bad enough to drag down your overall reputation when you send to them at all.

Acquisition source. If you've collected subscribers through different channels, test them separately. A list from a double opt-in form often lands differently than one from a single opt-in or a co-registration partner. Placement gaps between sources can tell you a lot about list quality before a single hard bounce shows up.

What you'll actually compare

For each segment, you're looking at inbox vs. spam placement rate (most seedlist testing tools break this out per mailbox provider), tab placement at Gmail (Primary, Promotions, Updates), and whether the pattern shifts over multiple sends or holds steady.

If your engaged segment lands well but your dormant one doesn't, that's a clear signal to suppress or re-engage before your next big send. If both segments land poorly, the problem is somewhere else entirely, and it's worth checking your authentication setup. You can run your domain through our free blocklist checker to rule out a reputation issue first.

One thing to keep in mind

Segment tests are only useful if your seed accounts are distributed across both groups. If all your seeds happen to be in one segment, you're not actually comparing anything. This is related to the problem of seed weighting bias, worth reading before you set up your first test.

Not sure how to structure the test in your specific setup? Our SOS hotline is free and we're happy to walk through it with you.