3
3 Comments

**How would you break down a mailing list for testing landing pages?**

How would you break down a mailing list for testing landing pages?

If you had a mailing list of say, 10,000 people and you wanted to test different landing pages, how would you break down your list for testing?

How many trial landing pages would you try?

How many initial mailings of X recipients would you test on each page before testing the whole list?

Would you build 5 pages and send each to 200 recipients?

Would you build just 2 pages and then iterate on the better page?

What's the minimum number of emails you would send for a test?

Any insight?

Thanks!

on February 2, 2019
  1. 2

    I'd say, depends on what you're testing. My method has been to test iterations, instead of A/B test. Meaning I would build a first version, Figure out a few people to send it to, asking for advice. Then iterate, send to more people. Pretty quickly I can stop asking for advice. Sending in waves.

  2. 1

    Just one landing page.

    1. Create your best guess landing page
    2. Test email "subject" not landing page
    3. Try 3 email subjects and send 100-200 to each group
    4. If you get a clear winner on number of "opens", do the same again for email body (this time testing for "clicks")
    5. After you have the highest email open and click rate then start testing the landing page
    6. 90% of the landing page is the title, so test a few different titles
    7. Then same again for CTA and other page elements
    1. 1

      Makes sense.

      I'm probably starting at the wrong end.

      Thanks!