a single source, nothing corroborates it

Name the conversion rate you are judging

Decide before you look which rates matter and for which visitors, then track only those. Otherwise analytics becomes a search for the flattering combination of page, source and morning. Published benchmarks give the bar: under one percent of visitors reaching a trial when you take a card upfront, above five when you do not.

The method

1. 01

   Decide in advance which conversion rates matter and for which groups of visitors, then track only those instead of hunting for the flattering combination afterwards.

2. 02

   Judge your homepage by whether it converts steadily into trials, and use these targets: under one percent of visitors when you take a card upfront, above five percent when you do not.

What it returned

• 0.75

  Justin's benchmarks: 0.75-1% visitor-to-trial with credit card upfront and 5%+ without.

• He contrasts this with digging through analytics until some pairing of a landing page and an ad shows a high conversion rate on a particular morning.

Sources

• The vague headlineJustin Jackson · justinjackson.ca · The vague headline · undated
• The Pattern-Seeking FallacyJason Cohen · longform.asmartbear.com · The solution: Form a theory, test the theory · 2010-06-14