A/B Testing

1. Do's

1.1. Know how long to run a test before giving up

1.2. Show repeat visitors the same variations

1.3. A/B test should be consistent across the whole website

1.4. Do many A/B tests

2. Don'ts

2.1. Always text both versions simultaneously

2.2. Don't conclude too early

2.3. Don't surprise regular visitors

2.4. Don't let your gut feeling overrule test results

3. Tools for A/B testing

3.1. Google Website Optimizer

3.2. A/Bingo and Vanity

3.3. Visual Website Optimizer

3.4. Unbounce and Performable

3.5. Vertster, SiteSpect, Webtrends Optimize and Omniture's Test&Target

4. Pros

4.1. Fast

4.2. Test reality, not theory

4.3. Quantifiable

4.4. Accurate

5. Cons

5.1. Can hurt website results

5.2. Missing the "Why"

5.3. Not predictive

5.4. Needs traffic

6. What is A/B testing?

6.1. Element A

6.2. Element B

7. What to test?

7.1. Depending on goals

7.1.1. Set metrics

8. Create your first A/B test

8.1. Replace the element to be tested before the page loads

8.2. Redirect to another page