Big News! Split to be acquired by Harness. Read More

Split Experimentation for Azure App Configuration Now in Public Preview


Split - Blog-2160x1080-Split Experimentation for Azure

As product engineering teams push to deliver more product features to market faster, understanding the impact of every software change is becoming more critical than ever.

Experimentation tools that measure the impact of new features, including AI-powered capabilities, are now “must haves” for every development team. Microsoft recognized this early on, which is why they chose Split to be their first-to-market, feature experimentation partner. Today, after close collaboration with Microsoft, our new offering is ready to use–right inside the Azure portal!

Split Experimentation for Azure App Configuration is now available in Public Preview. This means app development teams can successfully run experiments directly in Azure using easy-to-access Split capabilities, coupling every rollout with deep experimentation data as they safely test features.

“We are pleased Azure developers can now embrace experimentation capabilities from Split into every feature release and release new intelligent apps with confidence. As the new experimentation feature enters public preview, we will continue to innovate with Microsoft over the coming months to bring the best of our experimentation capabilities to app dev teams so they can test, learn, and iterate as they go.”

Brian Bell, CEO @ Split

About the New Feature

The new experimentation feature is native to Azure App Configuration. Azure App Configuration enables Microsoft customers to centrally manage application settings and feature flags. With the addition of Split Experimentation for Azure App Configuration, Microsoft customers can leverage feature flags from Azure App Configuration, application performance monitoring data from Azure Application Insights, and experimentation capabilities from Split to measure and understand feature impact, all within the Azure portal.

Split Experimentation for Azure App Configuration gives Azure developers a more efficient way to release changes that reduce the severity and duration of incidents. Once configured, users can easily click into active feature flags, compare the performance/behavioral data of various releases, and quickly determine if their product changes are making their apps better or worse. Experimentation is essential to monitoring feature rollouts, making critical release decisions, and moving fast without breaking things.

Why It Matters

With feature flags, developers no longer have to make a trade-off between speed and safety. However, if they aren’t able to monitor the impact of every feature they release (early on in a progressive rollout), they’re missing out on critical learning opportunities. Inevitably, they’ll introduce issues without even realizing it, creating additional triage work. In a small, percentage-based rollout, traditional monitoring tools will only register feature-related issues as insignificant noise. When doing progressive delivery, teams need something more. They need feature-level observability with pinpoint precision, which becomes possible with Split Experimentation for Azure App Configuration.

By leveraging experimentation from Split, Azure developers can measure and optimize AI capabilities without slowing down, connecting each experiment to performance and behavioral data. Then, as they forge ahead with their progressive rollouts, they’ll gain quantifiable insights right away to make faster, safer product releases.

How It Benefits App Developers

Release Faster With Measurement

Why wait to slowly roll your experiments forward? If they perform well in a percentage-based approach, Split Experimentation for Azure App Configuration will give you positive performance and behavioral insights right away. These measurements are timely enough to help you ship up to 50x faster.

Get Quick Detection & Instant Triage

Quickly catch and remediate issues with feature-level release insights. You’ll know right away if your features and AI-powered experiments are creating a negative impact or unexpected issues. Then, it’s easy to quickly isolate the problem. In the average team, Split Experimentation for Azure App Configuration will help reduce Mean Time to Remediation (MTTR) to less than one minute.

Test, Learn, Iterate Without Hesitation

Learn from every feature release so you can make informed iterations. Split Experimentation for Azure App Configuration eliminates the fear of making the wrong product decision and defaulting to the highest paid person in the room on what to do next.

Give App Development Teams New Levels of Autonomy

Run statistically rigorous experiments without relying on data science experts. Split Experimentation for Azure App Configuration gives app developers the ability to independently test and interpret their own data as they feature flag.

Common Use Cases

There are many reasons to use Split Experimentation for Azure App Config. Here are a few common use cases found at organizations across all industries:

Monitored Rollouts

Releasing a feature that breaks something is a costly mistake. Scrambling to fix the situation is even worse. Split Experimentation for Azure App Configuration provides feature-level observability, so you can ensure that you do no harm with everything you release. Measure against different business outcomes, technical operation metrics, or anything else your R&D team cares about when releasing.

Optimizing Intelligent (AI) Apps

Building AI Apps can sometimes feel like a black box. It’s hard to know if you are getting the value you want with what you release. Thanks to Split Experimentation for Azure App Configuration, you can quickly test, identify issues, and take instant action with your AI-powered features. Just set the metrics, those you want to see positive and negative impacts on, to their proper significance. Then, start making AI release decisions without second-guessing.

A/B Testing

A/B testing allows you to set up variants, then identify which ones are most aligned to the outcomes you’re looking for, pushing them to the right users without delay. Split Experimentation for Azure App Configuration helps set up A/B tests and optimize successful metrics with the most positive impact.

A Real World Example

Want to see Split Experimentation for Azure App Configuration in action? Follow this step-by-step demo as it shows you how to release, learn, and iterate a sample AI chatbot in a corporate benefits application safely and swiftly.

Split Experimentation for Azure App Configuration demo

There’s No Better Time to Adopt Experimentation

The integration of artificial intelligence (AI) technologies into various industries has ushered in a paradigm shift, presenting both challenges and opportunities for product and engineering departments. Central to navigating this transformative landscape is the critical role feature management and experimentation has served in optimizing AI software release practices. Many organizations have yet to adopt practices, but the ones who do will likely be the true winners of the AI revolution.

Check out Microsoft Learn to get started with native feature experimentation on Azure. Questions? Reach out to

Let us know how it goes, and if there’s anything we can do to help you release features that matter, faster.

Want to Dive Deeper?

We have a lot to explore that can help you understand feature flags. Learn more about benefits, use cases, and real world applications that you can try.

Create Impact With Everything You Build

We’re excited to accompany you on your journey as you build faster, release safer, and launch impactful products.