Natalie Ustymenko

Head of Direct Marketing

Olha Dmytruk

Editor

How to perform an A/B test

How to perform an A/B test in Yespo CDP and evaluate the results

It's important for marketers to experiment and test ideas to find the best approaches to customers. Accurate and quick assessment of results is crucial. Successful split-testing can increase the average revenue per ecommerce site visitor by 50%. Yespo Omnichannel CDP allows for automated A/B tests and provides detailed data on their effectiveness. This article explains how to conduct experiments for email campaigns and website widgets.

A/B Testing for Email Campaigns

Rather than relying solely on best practices, forum advice, or intuition, you can test your own ideas to implement the best ones. A/B testing elements like buttons, images, and titles can benefit your email campaigns in the long run.

First, define key metrics that will help to improve your emails:

  1. Set goals: What and why you are testing.
  2. Formulate a hypothesis: An assumption or statement on why this test should work.
  3. Specify a desired, significant outcome for decision-making.

Embedded A/B Testing of Email Subject

If you can test every detail of your email, start with simple tests, like the subject. In Yespo, creating such a test takes a few minutes:

  1. Specify two subjects (or more) in the email editor.
  2. Launch the campaign.
  3. Assess results in the statistics section.

Add subject lines

In the reports after sending emails, you'll find an "Experiments" tab where you can assess the test results. If the results are roughly the same, it's better to run the experiment again.

Experiments section

Experiment with email subject lines

Experiments with Test Groups

For more comprehensive tests like checking banners, buttons, animations, etc., you need to create several emails and divide your database into segments. To do this, select the required contact group for the experiment in your account and click the "Split Segment" button.
Splitting an email list for testing

Create as many segments as needed for objective evaluation of results. For example, if you have a large database, allocate two 10% groups for the experiment, then send the winning email to the remaining contacts.

Full evaluation of test results is feasible at least a day later. Considering different conditions (time, day) for sending emails, it's better to use this testing approach for landing pages, and for newsletters, try simultaneous sending to 50% of each group.

Testing Automated Emails

In Yespo’s workflow editor, you can set up A/B testing for any automated emails: welcome, abandoned actions, reactivation, etc.

For example, this is how a test flow for abandoned cart emails looks like:

Abandoned сart email testing

The system assesses whether the customer completed a purchase and sends experimental emails in equal batches. One email may offer a list of abandoned items, another a banner with a link to the cart.

You need around six months for such tests to send enough emails for analysis. This helps understand customer behavior and determine the most effective approach. Summarize results using Yespo's campaign results and sales data from Google Analytics.

Mailing results

A/B Testing of Widgets

Widgets help businesses to convert website visitors into loyal customers. They enhance the contact base, inform about offers, and collect additional customer data. Marketers should focus on testing widgets to find the most effective for their website.

Yespo CDP allows testing all widget types available in the system:

  • Subscription forms
  • Informers
  • Launchers
  • Age gates

Marketers can test not only different parts of a single widget (design, CTA, etc.) but also various types of widgets themselves. Available for experiments are:

  • 2 widgets

Testing widgets

  • 2 widgets when one has a connection to a launcher.

Testing widgets with a launcher button

  • 2 pairs of widgets and launchers

Testing widgets with two launcher buttons
The system automatically determines the winner based on the conversions: clicks, subscriptions, and confirmations. Test results are considered reliable if they exceed 90%. 
Additionally, you can view data on widget effectiveness separately for desktop and mobile versions, and a chart of target actions over the test period.
Test results

It’s easy to test marketing ideas in Yespo. If you don’t know where to start, we have examples of how to use widgets. This article offers 9 top widget ideas for ecommerce with detailed descriptions.

Create a perfect widget for your website

Summary

Any company, regardless of database size or budget, can improve its marketing strategy with A/B testing. Such experiments help choose the best activities to convert the visitors.

Yespo Omnichannel CDP makes it easy to run A/B testing with reliable results. The system's functionality helps determine the best themes, tone of voice, CTAs for emails, types, and content of website widgets, etc.

And if you need expert assistance, contact Yespo agency specialists.

5.0 from 5 based on 1 reviews

Natalie Ustymenko

Head of Direct Marketing

Olha Dmytruk

Editor

Comments 0