Landing page personalization with Oracle Maxymiser

Introduction

This page details how to use Oracle Maxymiser to A/B test and personalise your website using audiences and behavioral targeting.

Objectives

You will address one of the simplest yet powerful use cases of A/B testing and personalization, using audience/behavioral targeting. To this end, you will:

  1. Create a simple A/B test campaign on your site.
  2. Target specific audience(s) with the right offer.
  3. In the process, you will be exposed to concepts and methods around:

Prerequisites

Step 1: Choose the Right “Real Estate” on Your Website and Audience to Target

Using the Oracle Maxymiser platform, you can create a campaign on any area of your site. One of the first decisions to make is which site page(s) are you going to use - and what is the most suitable real estate on these pages in order to achieve your goal?

To illustrate the key steps, we are going to use a demo site that is structured like a typical clothing retailer:

https://sale.maxymised.com/

Suppose that our objective is to increase click-through to the Dresses category for new female customers. One way to do this is to use a welcome offer for Dresses in the Women’s landing page (LP) for new visitors only. Therefore, our thinking here (and the test hypothesis) is the following:

Step 2: Pick an Alternative Banner with the Offer That Matches Your Audience

Now there is a clear hypothesis in place: We will be testing whether the presence of a hero banner with a welcome offer on the Women’s LP leads to more Category Page Views for the Dresses category. Assuming we have such an offer in place (for example, with claim code DRESS21), we can create a suitable banner that show the claim code. For example, the banner below would do the job.

Oracle Maxymiser - Sample Banner

Oracle Maxymiser - Sample Banner

We believe this banner is suitable because it follows the look and feel of our site, it shows the right offer for the new season dresses, and also because it’s the right type of content to show to new visitors coming to the Women’s LP. However, there is only one way to find out if this is indeed the right banner by testing it.

Step 3: Set Up an A/B Test Campaign to Understand How Your Audience Responds

Now we come to the crucial step of creating a new campaign. Now we come to the crucial step of creating a new campaign.

3.1 Create a New Campaign in Oracle Maxymiser

In the Oracle Maxymiser UI, select New Campaign > Campaign Designer to open the visual editing tool. Give your campaign a suitable name and, if you wish to, describe what it’s about - what is the test hypothesis and which stakeholders are involved? What is the main KPI you are looking to improve with this campaign?

Oracle Maxymiser - Campaign Designer

Oracle Maxymiser - Campaign Designer

Once the campaign has been created, the Campaign Overview page opens. From there, you can do various things. We will start by adding URLs (that is, choosing where the campaign is going to be served). Click Add URLs, as shown below, and then use the visual tool to navigate to the page where you want to put the banner (Women’s LP). Then simply click on the plus button in the top right corner area, to add the page(s) you wish to include in your campaign.

Oracle Maxymiser - Add your Campaign URLs

Oracle Maxymiser - Add your Campaign URLs

3.2 Edit the Variant and Insert Hero Banner

You will now insert the new banner in the Women’s LP. At the top (next to the Code Editor option), you have the Edit and Browse modes. Make sure that you are in the Edit mode. Next, select an element on the page, (for example, the text box), so that you can insert the banner right above that (or below, or replace it altogether, if that’s more relevant).

From the right-hand side menu, you can now either add a specific modal (have a look at the templates contained in the Objects option, from the + button) or you can add your own piece of content. We will add our own content in this example, through the Library option (the folder icon), which we use to upload the banner and then choose the Insert Above option through the relevant icon.

Now that the banner is in the right place, we need to do two key things: (1) add a link to the banner, so that visitors who click on it go to the right place; and, (2) tell the system which Action to track, that is, how to measure success.

Tracking Actions

In the above steps, we have added a single (primary) action, which tracks clicks on this banner. More actions can be added, so that more user behaviors can be tracked and reported on as well, but there can only be one Primary Action at a time - the one used to identify the winner in a test campaign.

3.3 Preview the New Banner

The next step is Preview (in the following tab at the top). You should use this now to have a look at the two variants: your default (Control) variant and your new, alternative variant. By selecting each of them in turn, you see the different experiences that your visitors are going to be served once the campaign is published. This gives you a first opportunity to make sure that your alternative experience is as it should be (that is, the variant looks the way you want it to look, the URL link works as intended, and the like) If you find any discrepancies, or decide you need to make any changes, go back to the Content tab in the previous step and adjust accordingly.

3.4 Target New Visitors Only

In this crucial step you can determine two things: (1) who is allowed into your campaign; and, (2) who may see which variant. Formally, these two sets of rules are called (1) Campaign Rules and (2) Content Rules. The first of these is used to target the campaign as a whole to specific audience(s), while the second set of rules can be used to target variants. In this example, we are going to target the campaign as a whole to New Visitors.

In other words, we only want New Visitors to qualify into the campaign. So what we do is this:

Standard & Custom Attributes

As we saw, we make use of the left-hand side menu with the Standard Attributes. These are visitor attributes that are collected by the platform automatically. There are many of them, and include technology attributes, geolocation, weather, and many more. You can see the entire list in our documentation. But another crucial aspect to this, one which is essential for a personalization strategy that is meaningful for your particular business needs, is the feature of Custom Attributes.

Custom Attributes allow you to capture (or import) additional visitor data from the page or other data sources. Here are some suggestions on what you may want to capture, based on your specific industry. We have summarized these after working with hundreds of clients, and seeing what they find most useful for their personalization campaigns.

3.5 Use the QA Tool and Publish the Campaign

This step may take only a few minutes (assuming you have already installed the QA tool), or much longer, depending on the level of QA that needs to be done. For example, if there are multiple targeting rules using Custom Attributes then all of these rules need to be quality assured as well before publishing the campaign. After you QA the campaign (and any targeting rules), mark the campaign as “Ready to Publish” and action this through the Publish Center.

Please see Oracle Maxymiser Help Center - Preview, QA & Publish for details on this final step.

Frequently Asked Questions

How Long Should You Wait for Statistically Significant Results?

Publishing a campaign is just the start of the experimentation process. One of the key things to consider concerns campaign conclusion: When is the test ready to conclude? Our unique statistical methodology, using a group sequential approach, removes the need to wait until a predetermined sample size is reached before concluding a campaign. Tests can be concluded at any point in time and a statistically significant winner can be declared as soon as it is reported as such. Campaign conclusion messages are displayed at the top of the campaign performance report telling you when a winner is declared or giving an estimate of when your campaign will be ready to conclude.

Please see Oracle Maxymiser Help Center - Campaign Conclusion for more details on Campaign Conclusion.

Calculating Test Duration and Analyzing Campaign Performance

You may use our Test Duration Calculator to get a good sense of how long you should expect to wait for test results.

Please see Oracle Maxymiser Help Center - Campaign Performance Report on how to analyze your campaign performance.

More Learning Resources

Explore other labs on docs.oracle.com/learn or access more free learning content on the Oracle Learning YouTube channel. Additionally, visit education.oracle.com/learning-explorer to become an Oracle Learning Explorer.

For product documentation, visit Oracle Help Center.