To get the most out of your web-to-app strategy, follow these tips and tricks for setting up your very own A/B variations.

1. Test one element at a time to pinpoint its impact on engagement.

It’s easy to get carried away. Simplify your A/B experiments by testing one element variation at a time. This could mean running one A/B test to compare Arial versus Helvetica fonts. Then, test top versus bottom smart banner placement in another. Focus on one difference at a time for clearer insights into what drives banner performance.  

2. Establish clear KPI success criteria.

When optimizing your web-to-app banner strategy, it’s crucial to start with clear key performance indicators (KPIs). Focus on view- and click-to-install rates as foundational conversion indicators. Also, pay attention to industry-specific metrics such as view- or click-to-custom conversion event metrics. These can help you understand which end-to-end flows to optimize beyond basic conversions like install and open.

3. Start small by assigning a blank banner to a portion of your audience.

First-time banner tests can be daunting. Ease into testing with a phased rollout by showing a subset of users a blank banner. You can adjust variation percentages at any point during your campaign, even if the banner isn’t live yet.

4. Avoid mid-test changes to your banner variations.

Whether it’s tweaking audience filtering or editing your banner creative, even small changes during a live test can compromise data. Ensure that your banner variations align with your intended campaign vision by thoroughly QAing them each time you launch a new variation.

5. Consider banner circumstances and context when evaluating engagement metrics.

It’s important to evaluate KPIs in context. For example, you might think your full-page interstitials are unsuccessful due to their high dismissal rates. In reality, when users are presented with a full-page banner, they’re required to either click on a call-to-action button or dismiss the interstitial to continue their mobile web journey. Consequently, the dismissal count will be higher compared to less-intrusive top or bottom banners.

6. Apply A/B test outcomes from larger campaigns to those with smaller audience reach.

Small campaigns may struggle to reach a significant sample size, leading to inconclusive or hard-to-interpret results. Similarly, time-sensitive campaigns like holiday promos might not allow for thorough A/B testing. To maintain a consistent user experience, carry over results from larger, similar A/B tests to smaller campaigns. Persisting treatments from larger-reach campaigns onto smaller ones streamline efforts and minimize duplication.

7. Run A/B variations for at least two to three weeks.

Branch recommends running A/B tests for two to three weeks initially. If results remain inconclusive, extend the duration or persist other experiment results onto the campaign, as mentioned above.

8. Save time on variation setup by updating existing creatives.

Cut down on time spent recreating creatives manually and avoid duplicate efforts by copying existing ones from the ‘My Creatives’ tab in your Branch Dashboard. This repository houses both drafted and live banner creatives, allowing you to duplicate and create variations from any of your existing templates.

9. Embrace testing! Find answers by putting your ideas to the test.

Journeys banners offer extensive customization options, and it may surprise you which elements impact audience engagement. From call-to-action button text to copy, title, app icon, background images, fonts, font sizes, text and background colors, banner placement and size, explore diverse variations to optimize your campaigns. With our View Editor, you can easily adjust banner aspects. For greater customization flexibility, use our CSS or HTML editors. 

Next steps

Optimizing your web-to-app smart banners is easier than ever with Branch’s A/B testing tool. By delivering multiple experiences to your audience through creative variations, you can finally gain data-backed insights into your best-performing creatives. 

Need inspiration? Learn how IMDb boosted app downloads by 443% by utilizing A/B testing in its campaigns.

If you have questions about setting up your first A/B tests, reach out to your account team. Visit the Branch Dashboard to get started!

The post Maximizing Conversions: Tips for A/B Testing Branch Journeys appeared first on Branch.

Upon completing the Journeys setup in your Branch Dashboard, you might wonder, “How can I be sure my banners look good to users and function properly once I launch them?” Good news: Branch has you covered with multiple methods to thoroughly test your banners before a full-fledged launch.

Method 1: Use your test Environment

By default, in the Branch Dashboard, you have access to both test and live environments. To switch from one to the other, you simply use the toggle at the top left of the Dashboard.

To get started, make sure you are in the test environment. Then, navigate to the Journeys tab to set up the banners you want to test. Note that to carry over Journeys from the test environment to the live environment, you need to manually recreate them.

Next, verify that your web engineers have integrated the Branch Web SDK into your webpage test environment using the Branch API test key from the Dashboard.

Before launching your test banners, ensure that the filtering criteria are simple and easy for the tester to match during testing (e.g., ‘user has completed one web session’), then adjust as needed when moving the Journey to the live environment.

You’re now ready to launch and validate test banners on your internal test site! Remember to replace the Branch API test key with your live key when you’re ready to launch banners on your website from the live Branch environment.

Method 2: Use the Branch URL testing tool

When you reach the final step of configuring Journeys banners in the Branch live environment, you’ll see a preview of your Journeys and an input field. Here, enter your website’s URL to generate a test link.

This URL can be any page on your web map where the web SDK is currently active. If possible, we advise using the web URL of a page that will display the banner once it’s live. This will help you visualize how the banner will look and feel in production.

After clicking on the “Get Test Link” CTA, you’ll receive your inputted URL with an additional query parameter indicating it’s a test link. Only individuals with this test link can view the banner test rendering on the live page. Don’t worry — the banner isn’t live on this web page, so your end users won’t see any changes while you’re testing with this URL!

This tool allows you to test the look and feel of banners on your web pages but not the actual deep linking functionality. Additionally, since this unique URL is a single page, if you navigate around to other pages, the banner will not persist. Note that filtering criteria are not considered in this scenario, so we don’t recommend using this method to test end-to-end user flows.

At the top of the page, you can use the Preview dropdown button to switch the creative variation you’re previewing. Don’t forget to click “Update Link” when you switch between variations.

Method 3: Test live on a low-traffic URL

This final technique allows you to test your Journeys on a live, in-production page, so you can test the entire end-to-end user experience in a small, controlled simulation. When you’re ready to launch a banner in your Branch live environment, identify your least-visited pages. Then, adjust the Journeys filtering criteria to target one or a few of these low-traffic pages.

To minimize the likelihood of your end users encountering the banner during testing, we recommend conducting tests during your website’s lowest traffic hours.

Now, you’re ready to launch your banner test! With this method, you’ll be able to thoroughly validate in your live environment that your banner renders correctly, deep links users as expected, and responds appropriately to user navigation between pages — persisting on pages that match filtering criteria and not rendering on those that don’t.

Once you confirm that your banner behaves as intended, you can fine-tune the filtering criteria to your desired configurations and validate once more, confirming that all end-to-end user flows function.

Best practices to remember

Follow these best practices to ensure a smooth and successful launch:

1. Use Branch’s A/B tool to facilitate a phased rollout. By leveraging the A/B tool in Step 2 of your Journeys setup process, you gain complete control over how many users see a given Journeys’ campaign. Use the tool to gradually launch your campaign by incrementally increasing the audience reach from 0% to 100%.
2. Launch invisible banners with your desired filtering criteria. If you’re hesitant about launching a full-fledged Journey campaign to a certain audience, consider starting with blank banners. This won’t disrupt your current user experience, yet with the Web SDK’s ability to track pageviews out of the box, you can gauge the true impact of your audience filtering concept.
3. Regression test all new and existing banners with every launch. Validate end-to-end flows by mimicking complex user journeys with every new Journey you launch.
4. Validate all variations separately. If you’re running A/B variations, set one of your variations to target 100% of your audience upon launch. Confirm this variation functions as expected, then repeat for all of your remaining variations. Once all of your variations have been validated, assign the desired audience percentages for this campaign.

If you have questions about configuring and testing your Branch Journeys, reach out to your account team. Go to the Branch Dashboard to get started!

The post Testing Branch Journeys: Best Practices and Techniques appeared first on Branch.