Tagged As: Optimization Result

Normanrecords.com, an e-commerce records store (LPs), while working with us decided to try and copy-paste a simple Fastforward . The was backed by multiple past tests and was a solid candidate for an experiment.

The Pattern: #1 No

We typically start off conversion projects by looking through what we already known in the form of patterns (including their win probabilities and predicted effects from past tests). One such pattern that stood out ripe for testing on the Norman Records’ checkout screen was No Coupon Fields. The pattern suggests to remove coupon fields or make them less prominent as they might send people away in search for discounts on other sites (or make them feel like others are receiving discounts while they are not).

Based on numerous tests backing this pattern with a repeatability score of 2.25 we estimated a very likely chance to win (the more positive tests we have in favor of a pattern, the higher this score, and the higher we prioritize a test idea).

No Coupon Fields
Pattern #1 No Coupon Fields with a quantified repeatability from past tests

The Pattern Also Worked In A Very Similar Test On Adoramapix.com

It’s interesting (and profitable) to compare similar tests together. Before the Norman Records test was even started, we already had data from a very similar test that ran on Adoramapix.com where it generated 2.6% more sales. Both tests ran on checkouts. Both tests were e-commerce sites. Both tests had a similar conversion baseline of around 70%. We had our prediction in place.

Tested Screenshot
Pattern #1 No Coupon Fields tested on Adoramapix.com

The Test On Normanrecords.com Nicely

With positive past data backing the experiment, Normanrecords.com set up their own test within a day (patterns enable experiments quickly). Interestingly, they didn’t just remove the coupon field, but instead tweaked the pattern slightly to their own needs. In the variation, the coupon field was replaced with a “add a voucher code” link which upon clicking would show the coupon field. This made the coupon field less visible without completely removing it.

Some weeks later they called the test positive and shared the results with us. We were glad to see another positive prediction (not all tests win by the way). More so, the effects were also very similar between both comparable tests. The Adoramapix.com test had a 2.6% relative increase to sales, and Normanrecords.com detected a 2.4% increase (slight wider confidence interval). They decided to implement it from additional positive metrics aligning with this observed effect (ex: even greater revenue gains).

Tested Screenshot
Pattern #1 No Coupon Fields tested on Normanrecords.com

The Key Takeaway: Patterns Can Be Tweaked In Creative Ways

Aside of patterns helping in predicting tests, I think the key insight here is that patterns can be modified in creative ways. When looking at past test data it might feel as if there isn’t much room for other creative solutions – this is false. Creativity, interpretation and building on existing patterns is more than possible. For every AB test there is are CDEFG ideas just around the corner. Patterns can also be mixed and matched in creative ways. And replication story is just one subtle indication of this openness to vary an idea beyond what it originally was.

Next Steps: Test Patterns On Your Site

  1. See Which Conversion Pattern Are Relevant For You
    I encourage you to run a/b tests. We continuously identify conversion patterns for you to test with on GoodUI Fastforward. Many have positive data, and some have negative effects as well (members get to see all tests for given patterns so that the chances of replication increase).
  2. Get An A/B Testing Tool Such As VWO
    To test your ideas/patterns you need a good testing tool such as VWO (our recommendation if you don’t have a tool yet) that won’t hurt your wallet. Some pattern ideas are really simple and can be setup in minutes or hours.

Source link https://goodui.org/blog/how-normanrecords-com-replicated-the-no-coupon-fields-pattern-successfully/


Please enter your comment!
Please enter your name here