A/B TESTING

Stop guessing which destination converts

Split-test link destinations from a single short link. Set two or more variants with your own traffic weights, let visitors be assigned consistently, and read the results with Bayesian statistics. When a winner is clear, one click routes all traffic to it.

2+ variants with custom weights Consistent per-visitor assignment Bayesian & sequential statistics One-click winner promotion
A/B test — acme.link/launch
Variant A · /launchWEIGHT 50%

3.2% CTR

Variant B · /launch-v2WEIGHT 50%

4.8% CTR

Variant B leadingRunning
2+
variants per test
Bayesian
statistics
2–20
interim analyses
1 click
winner promotion
HOW IT WORKS

A/B Testing in three steps.

01

Define your variants

Create a test with two or more destination URLs and set the share of traffic each variant receives. A single short link serves the whole test.

02

Run it

Move the test from draft to running; pause or stop it whenever you like. Each visitor is assigned consistently — they keep seeing the same variant on repeat visits.

03

Promote the winner

Results show click-through and conversion rates per variant, evaluated with Bayesian statistics. One click promotes the winner and routes all traffic to it.

CAPABILITIES

A closer look at a/b testing.

Split Traffic from One Link

No duplicate links, no manual traffic juggling. One short link splits its visitors across your variants according to the weights you set.

  • 2 or more variants per test
  • Per-variant traffic weights
  • A single short link serves the whole test

Statistics That Hold Up

Results are evaluated with Bayesian statistics. If you want to check results before the test ends, optional sequential testing keeps those peeks honest.

  • Bayesian statistics on every test
  • Optional sequential testing with O'Brien-Fleming or Pocock spending functions
  • 2–20 interim analyses per sequential test

Clean Assignment

A visitor who saw variant A keeps seeing variant A. Consistent assignment keeps your results free of cross-contamination.

  • Per-visitor consistent assignment
  • Repeat visits land on the same variant

Lifecycle & Winner Promotion

Tests move through a clear lifecycle, and finishing one takes a single click.

  • Draft → running → paused or stopped
  • CTR and conversion results per variant
  • One-click winner promotion routes all traffic to the winning variant
USE CASES

Where teams put it to work.

  • Landing page comparison
  • Call-to-action testing
  • Offer comparison
  • Funnel optimization
  • Message testing
  • Pricing page experiments

Let the data decide

Split traffic between destinations, read the results with Bayesian statistics, and promote the winner with one click.

Bayesian
statistics
1 click
winner promotion
FAQ

Common questions.

KEEP EXPLORING

Related features.

Put a/b testing to work on your next campaign.

Everything is free while we're in beta. No credit card required.