GrowthNovember 15, 2017

3 ways to leverage audience A/B testing

A/B testing can help marketers understand how campaign attributes like engagement channels, creatives, and ad networks affect customer engagement.

3 ways to leverage audience A/B testing

Consumer demand is constantly evolving in the multi-screen era, and brand marketers must also evolve to meet the needs of their customers. Effective marketing relies heavily on marketers’ ability to experiment as circumstances change, and measure the results of said experiments.

Today’s customers are everywhere and you need a framework that can reach them anywhere at anytime. A foundational framework that enables a strategy for marketers to test channels, paid media partners, and creatives without a three-week engineering queue is the only way to keep up.

A/B testing

Deploying a platform that reduces marketers’ dependency on engineering and offers them the ability to perform real-time experiments tailored to your marketing initiatives is what sets brands apart.
We’re very excited to release Audience A/B testing to give marketers the tools at their fingertips to deploy flexible tactics to acquire, engage, convert, and retain users. Marketers, be liberated!

Here are three strategic examples of how to apply a framework of Audience A/B testing.

Drive healthy retention

Every marketer wants to increase retention, and there are many approaches one can take to meet your 7-, 30-, and 90-day retention goals. Re-engaging churned users can be near-impossible, so your engagement strategy must be dynamic. Common approaches are usually siloed to a single platform or database of users. A sophisticated marketer will expand the strategy and A/B test all engagement tactics to determine which performs best. We recommend building audiences and splitting them into variants that can be maintained in real time to the following channels:

  • Email
  • Push
  • Facebook/Instagram
  • Snapchat
  • Control group

Experimentation is all about learning more about your users. Our customers have discovered the right channels for their users on a per audience strategy level. Furthermore, you can start to build individual profiles of users based on their engagement receptiveness. For example, Snapchat and Instagram might be great ways to drive daily retention metrics for millennials, whereas email would be a great weekly channel to deliver emails to older generations. Lastly utilizing a control group is the only way for our customers to truly measure the impact of campaign delivery channel.

Personalize in-app experiences

Audiences shouldn’t only live in marketing engagement channels, they should also be exposed directly to your mobile app’s source code. mParticle exposes audience IDs through our native SDKs, which then enables your app to know the audiences your users are active in at runtime. We are giving marketers the power and authority to determine which audiences your users are a part of that should be exposed to in-app personalizations. Let’s take the following use case:

You have a group of users that successfully went through onboarding and did not convert during their first session, but returned for a second or third session. Let’s form a hypothesis that users are merely exploring and getting introduced to the app on their first session, before returning on their second session to purchase or convert. We can put all these users into an audience and split that audience into multiple variants. At runtime, we can identify which variant a user is a part of and offer different incentives to drive users to conversion. Experimentation also requires the ability to have a control group to truly measure the impact of your experiments.

Data science analysis

Requests get stuck in engineering queues making it difficult to scale experimentation results. There can also be a language barrier between marketers and technical analysts when it comes to communicating needs and outcomes. By providing marketers with the authority to easily define user list criteria and import users into technical tools like for analysis, marketers are able to experiment at will and measure results. Marketers can also send user lists to Amazon Kinesis Firehose, in real time, to stream, transform, and load large amounts of data to tools like S3, Redshift, and Kinesis Analytics.

Test your campaigns

To be successful, marketers need access to platforms that give them the flexibility to deliver unique customer experiences that drive loyalty. Audience A/B testing provides a powerful framework that not only allows marketers to test engagement channels, creatives, and ad networks but also bridges the gap between business intelligence and campaign impact analysis.

We have entered an era where marketers require more power to make technical decisions. Audience A/B testing puts marketers in the driver’s seat when it comes to activation of data. If you would like to refine your marketing campaign execution and create better customer experiences, get in touch!

Latest from mParticle

See all insights
A thumbnail image showing the headshot of mParticle's CPO, Jason Lynn, along with some text with the blog post headline.

Company

Jason Lynn returns to mParticle as Chief Product Officer, leading the company into a new era of innovation

A photograph of a laptop screen showing the mParticle website

Company

mParticle rebrand: Turning customer data into customer joy

mParticle 2.0

Company

Deep-dive into the new mParticle: A unified platform and updated UI