Countdown timer A/B test: find the design your readers act on

The test URL doesn't render a timer itself. When an email client asks for it, flashtimer picks variant A or B and redirects to that timer's image. The pick comes from a hash of the test and the subscriber ID your platform fills in, so the same person lands on the same variant on every open and every device. The assignment is also stored for 90 days, so a later change can't move someone who has already been counted.

On a platform without a reliable per-subscriber merge tag, the test page says so: each open may then get a random pick, weighted by your split, and the test's tracked link isn't offered, because a click couldn't be tied to the variant that reader saw. Link the image in your email platform instead, or choose a platform with a merge tag.

Pause a test and every reader gets A until you resume it, so sent emails keep showing a working timer. Views while it's paused aren't counted, and clicks are kept apart as unattributed, so a pause never tilts the result toward A. A/B tests are part of the Business and Enterprise plans.

For each variant the test page shows views, clicks and CTR. Views count only images served through this test's URL, not every load of the two timers elsewhere, and leave out Apple Mail Privacy Protection prefetches, renders after the timer expired, and rate-limited requests, which are mostly automated. Clicks are clicks through the test's own tracked link, each credited to the variant that reader was shown, using the merge tag in the link. A click with no subscriber ID can't be credited, so it is counted separately as unattributed rather than guessed. CTR is clicks divided by views.

The page also shows how sure you can be that the two click rates really differ, as a confidence percentage from a standard two-sided two-proportion test, capped at 99%. Until each variant has 100 views and the two have enough clicks between them to compare, there is no figure. A test reaches Needs your decision at 95% confidence, whichever variant leads. Small lists rarely get there: a few hundred opens can't separate a 2.0% from a 2.3% click rate, and that is the test telling you the truth.

Test one change at a time, so a winner tells you why it won. These are the variables a timer test can isolate:

• Layout: boxes versus circles, or the compact layout versus the rings layout.
• Contrast and color: brand colors versus a high-contrast red and black.
• Days or hours: show the days unit, or hide it so a 3-day sale reads as 71 hours.
• Size: large digits versus medium in a busy hero section.
• Window length: two evergreen timers, one 24 hours and one 48 hours, in the same abandoned-cart email.
• Expiry message: "This deal has ended" versus a message pointing to what is next.

Look at CTR, not raw views. Views tell you the split is working; the difference in clicks per view is the result. If both variants still sit at about the same CTR after most of your list has opened, the honest conclusion is that the design isn't what moves your readers, which is useful too.

Don't change the split mid-test. It skews the result, so pause and start a new test instead. And treat a winning design as a winner for that audience and offer: re-test it on your next big send rather than assuming it carries over.

To test a timer against no timer at all, use your email platform's own split test with two email versions. That answers a different question, and our email countdown timer best practices cover how to run it.

Emails you have already sent point at the test URL, so ending a test never breaks it. Make A the winner or Make B the winner ends the test for good: from then on the URL shows the winner to every reader, in every email that carries it, sent or not, and its tracked link keeps redirecting. There is nothing to swap in your templates or flows. The numbers the decision was made on are frozen and kept in History on the A/B Tests page, with the winner and the date the test ended.

Pause is the reversible option: every reader gets A until you resume. Only Delete test retires the URL: older emails then show a “This timer has been deactivated” notice instead of a timer, and the test's tracked links stop working, so keep deleting for tests that were never sent. That makes it safe to decide mid-campaign, even for a flash sale still being opened or an always-on flow.