We are excited to announce that today, we are launching the Split Feature Experimentation Platform. The platform offers enterprises a secure way to roll out software, target specific features to customers, and measure and analyze the impact of features on key metrics. It is the only offering that can ingest and process the full spectrum of an organization’s product metrics and tie them back to actual features within the product.
Split has been on this journey from day one, fundamentally believing that the practices of continuous delivery and agile product development needed a better guidepost. Experimentation platforms have grown over the last few years, but they have mainly targeted testing of UI changes for marketing teams. Engineering and product teams need a solution built from the ground up for continuous delivery and full-stack experimentation. Key to making this work is rooting experimentation in the fundamental building blocks of a product – actual product features.
Today’s Software Delivery and Measurement Silos
As organizations of all types build out increasingly sophisticated digital business, they face the challenge that continuous delivery and agile product development often operate in silos. Only leading software companies with enormous resources have been successful at building internal platforms that bring together constant testing of new feature releases, and consistent measurement of key product metrics.
The alternatives for most organizations are not great. Some just continue to try to get feedback from customers through traditional mechanisms, such as customer research, even while releasing code on a weekly or monthly basis. In-depth customer research will always be important, but it does not give you the immediate feedback needed in an agile development process.
Another option is to bring event data into a data warehouse and generate reports or important metrics you can track against. The problem with this approach is that it becomes an error-prone guessing game to figure out what code change actually moved the needle on a metric – or did nothing.
Finally, product teams can try to use UI experimentation tools, however, those tools are mainly focused on marketing teams. They focus on front-end changes, do not tie back to actual feature changes in a product and are not designed for developers. Even as some move into full stack experimentation, they do not fully tie experimentation to feature flags, and require developers to proactively send event data tied to visitor interactions. Ultimately, this makes it challenging to use them to measure feature releases against all the metrics product teams care about.
The result of not having a great alternative is that most product teams end up suffering from slower innovation than expected as many features are not adopted by users. Other issues arise, as every feature release has an uncontained “blast radius” and can affect every customer at once. And, product managers and senior leaders have minimal visibility as to the impact of product investments.
The Split Feature Experimentation Platform is the only platform that brings together continuous delivery and full-stack experimentation. It is designed specifically for product teams, to test actual features in a product against key product metrics.
This enables teams to release code quickly, while testing the impact constantly, enabling better product decisions and greater speed of innovation. Since Split ties to directly to feature rollout and management, it also mitigates risk with the granular targeting of features to users during roll-out. Overall, Split enables product teams to make smarter product decisions.
Key components of the Split Feature Experimentation Platform include:
- Open Source Software Development Kits (SDKs): Split is powered by custom SDKs, available for most of the popular languages in use. These SDKs install easily at the application layer and act as the engine deciding what feature version to show customers.
- Feature Flags: Split uses feature flags to control the rollout state of individual features, anywhere in the stack, and communicate with the SDK to turn the feature on or off, or to show multiple variations of the feature.
- Management Console: The Split editor is hosted in the cloud, and is an easy-to-use yet powerful way to roll out a feature to a segment of users by targeting a whitelist, a random percentage of traffic, or any set of attributes that make sense for the business.
- Intelligent Results Engine: A key component of the platform is the Split Intelligent Results Engine – a metrics-first engine that handles data ingestion from any data source, calculates metrics based on that data and correlates the data to features, automatically highlighting significant changes due to feature releases across the metrics that matter most.
- Intelligent Security Framework: Split leverages industry-standard security practices and never requires user-identifiable data to be sent to Split servers.
Split is designed to handle a variety of use cases, depending on the needs of an organization as they go through digital transformation. Engineering teams can start by using Split for feature flags and get more advanced with precise targeting during roll-outs. As product managers look to get more visibility, they can leverage the data in Split to do root cause analysis and tie application performance to code changes. Finally, organizations use Split to do full-stack experimentation and understand the complete picture of how product innovation is driving key metrics.
Getting Started with Split
Split is generally available today. The Split Intelligent Results Engine module is in public preview now. If you are interested in evaluating the module, please sign up on the Results Engine product page and we will be in touch! We ourselves use Split to gradually roll out new features to subsets of customers! If you would like to see how the complete Split Feature Experimentation Platform works, contact us today and we would be happy to work with you on a proof of concept. To get started with Split now try our 14-day free trial at www.split.io/signup.
Stay up to date
Don’t miss out! Subscribe to our digest to get the latest about feature flags, continuous delivery, experimentation, and more.
Why build or buy a robust feature flags as a service platform when you could just implement a simple toggling mechanism as a one-off and be done with it?
Data Hub combines our impressions view with new capabilities like data export and a single dashboard to view, query, and export feature flag and event data.
As we begin our work on diversity and inclusion, we’d like to introduce you to the team who will lead our taskforce, including Software Engineer Joy Ebertz.