Dedicated setup modal
A focused screen for the test — name, variants, metric, pilot %, sample size, confidence, winner strategy. No accidental audience selection, no schedule fields, no clutter.
Send campaigns · A/B testing
A/B testing that picks the winner for you
Pilot a slice of the audience, let Mailbase choose the winner by reply rate, roll the winner out to the rest in one click.
IllustrationFictional names, companies and messages. Counts come from one measured run.
01Why it works
Three things it gets right.
Most email tools either skip A/B testing or hide it behind a confusing wizard. Mailbase ships a dedicated A/B test setup modal — two to four variants, the metric you care about, the pilot %, the confidence threshold, the winner strategy. Save it as a draft; attach it to a campaign when you're ready.
Real statistical winners
Open, click, reply, or conversion as the metric. Confidence threshold defaults to 90% with a custom-set sample-size floor — Mailbase only calls a winner when the math is real.
One-click rollout
Manual override available, auto-rollout when thresholds clear. The winning variant goes out to the rest of the audience without you re-building the campaign.
02What it ships with
The pieces, one by one.
- 2–4 variants per test, subject + MJML body each
- Metrics: open, click, reply, conversion
- Confidence: 80% · 90% · 95% · 99%
- Auto-winner or manual override
- Variants saved as a draft A/B campaign — attach to a campaign whenever
03Related
Works with the rest of the loop.
ProspectingProspecting that starts from the French company registerReply inboxRead every reply, sorted the way an inbox should beDeliverabilityDeliverability that starts before the first emailAI & developer APIAI that answers where you're workingTemplatesMJML templates with a live preview that's actually live
Start today
Try it in your workspace.
Free to start. No credit card. Managed sending, or your own Amazon SES.