Let's Test This: A Better Modal Popup - Oli Gardner's Maybe Later Pattern

Pattern Of The Month

This week Oli Gardner from Unbounce.com came up with a really promising approach to modal popups - captured as Pattern #57: Maybe Later. Now I'd like to inspire more people to test this and see if it's actually any better than a traditional / annoying modal. Hence we're running a contest. If you a/b test this pattern on your site, and share your results with us to publish for everyone to see, I'll give you 3 months access to Fastforward - it's where we publish a/b tests to see which patterns have the highest impact & repeatability.

Do You Have What It Takes? Here Is What You Need To Qualify:

  1. Run An A/B Test On Your Own
    You need to be able to conquer uncertainty and turn this pattern into an a/b test. That's the only way we can learn. Need an a/b testing tool? We like and recommend Visual Website Optimizer.

  2. Minimum Monthly Conversions
    You need to throw 1,000+ deep conversions (ex: signups or sales) per month at this experiment. I want these a/b tests to capture at least 500+ deep conversions per variant to have some reasonable statistical power.

  3. Preregister & Start The Test Before Jan 31, 2018
    You need to share the test link with me before we start. This way, if the test turns either south or north I will know. We want to publish any result (both negative or positive) to fight publication bias.

  4. Open To Publishing On Fastforward
    The test results will be published on GoodUI Fastforward right here under Pattern #57: Maybe Later (I will also enable public preview of all test data to anyone - as a bonus). It will be super fun to compare numerous results to see: how many go into the positive/negative, median effects, as well as try to understand and explain any drastic differences.

Ready? Here Is How You Should Setup The A/B Test:

  1. Testing Scope
    You can run the test anywhere on your site with any segment (it needs 1000+ conversion per month).

  2. Your Control Version
    Let's make sure that we are comparing against an existing annoying entrance popup. If you don't have a standard (Yes/No) type of popup, then add it to your control or setup an A,B,C test (A=no popup, B=standard popup, C=Oli's-maybe-later-popup).

  3. The Change
    Take Pattern #57: Maybe Later and turn it into an a/b test. Please read all of the properties carefully. If you have questions, ask me.

  4. Measure Goal 1: Shallow Metric - Instant Yes
    As a shallow goal, please measure an instant "yes" reaction from the initial popup with any of the following: views of the offer, product page or signup page.

  5. Measure Goal 2: Shallow Metric - Postponed Yes
    As another shallow goal, please also measure the postponed "yes" reaction from the persistent footer with any of the following: views of the offer, product page or signup page.

  6. Measure Goal 3: Shallow Metric - No's
    Please also measure how many people press "No".

  7. Measure Goal 4: Shallow Metric - Maybe's
    Please also measure how many people press "Maybe".

  8. Measure Goal 5: Deep Metric - Signups Or Sales
    As a deep goal, most importantly, please also measure the completed signups or sales resulting (depending on your modal message / offer with visits to some final step).

  9. Email Me Before You Hit Start
    Before you start the test, please let me know (and share your A / B screenshots).

Ok, let's do this. You have until January 31. :)


  • Jakub Linowski

    Jakub Linowski 5 years ago 30

    People try stuff in the name of science. In the name of exploration. :)

    • Ivan Burmistrov

      Ivan Burmistrov 5 years ago 00

      What to explore in this case -- that people hate popups and sticky footers? This is already well-known... :-)