A/B Testing with Redirects

How A/B Testing Works

RedirectFlow's A/B testing feature allows you to split traffic between multiple destinations based on configurable percentages. This enables you to:

• Test different landing page designs
• Compare conversion rates across variants
• Gradually roll out new pages
• Run multivariate tests with multiple variants

Setting Up an A/B Test

Source: /landing

Variant A (Control):
  Destination: /landing-v1
  Traffic: 50%

Variant B (Test):
  Destination: /landing-v2
  Traffic: 50%

API Configuration

You can also configure A/B tests via the API:

POST /api/v1/redirects

{
  "source": "/landing",
  "type": "ab_test",
  "sticky_session": true,
  "variants": [
    {
      "name": "control",
      "destination": "/landing-v1",
      "weight": 50
    },
    {
      "name": "variant-b",
      "destination": "/landing-v2",
      "weight": 50
    }
  ]
}

Analyzing Results

RedirectFlow tracks key metrics for each variant:

For conversion tracking, integrate with your analytics platform using our webhook events or the analytics API.

Best Practices

• Run tests long enough: Ensure statistical significance before declaring a winner (typically 1-2 weeks minimum).
• Test one variable at a time: Isolate changes to understand what drives results.
• Use sticky sessions: Prevent user confusion from seeing different variants.
• Document your tests: Use redirect names and notes to track test hypotheses.
• Gradually increase traffic: Start with small percentages for new variants to limit risk.