Opticon Stockholm is on Tuesday September 10th, hope to see you there!
AI OnAI Off
Feature Experimentation
In order to provide a fully customizable, proven digital experience for visitors, consider optimizing your applications. In other words, let’s not only test the website, let’s test everything! While Optimizely’s Web Experimentation product helps to test and optimize your web experiences, the Feature Experimentation product will help you go further by optimizing your development applications. Feature Experimentation (previously known as Full Stack) is a feature flagging and experimentation platform for websites, mobile apps, chatbots, APIs, smart devices, and anything else with a network connection. Some example applications include search results, promotions, payment options, recommendations, and much more.
How it works
Optimizely Feature Experimentation is a feature flagging and experimentation platform that enables server-side testing with experiences decided in the backend that are sent to the user instantly. Users can turn on pre-programmed feature flags and decide on the percentage of users to which they would like to gradually roll out those features. This data is sent to the Project Datafile (JSON) which sends the request to the server SDKs, Client SDKs, and Optimizely Agent (Microservice). The Client SDKs and Optimizely Agent tracks the test and sends the results to Optimizely's experimentation platform.
Feature flags and tests can be set up in the browser with a user-friendly interface and are distributed via the Content Delivery Network (CDN). Decisions are made in memory through the open-source SDKs.
SDK Programming Compatibility
You can deploy code behind feature flags, experiment with A/B tests, and roll out features via Optimizely’s easy-to-use, open-source SDKs. Learn how to initialize and install any of the supported Feature Experimentation SDK languages.
Explore nowFeatures
Optimizely's Web Experimentation has a plethera of new features! Choose your level of personalization from the Essential, Enhanced, Advanced, or the Ultimate packages. Check out the top features for each package level.
Essential
- Unlimited concurrent experiments
- Feature flags, variables, and rollout
- IDE plugins
- Stats engine and accelerator
- Bot filtering
- REST API
- Support options (community, academy, online tickets, etc.)
- And more
Enhanced
- All Essential features
- Multi-Armed Bandits (Allocate traffic to variations that are performing well with machine learning)
- Experiment Calendar and Scheduler
- Rule scheduling
- AI variation generator
- Single sign-on (SSO)
- And more
Advanced
- All Enhanced features
- Mutually exclusive experiments (exclusive visitors)
- MS 365 connector
- Cross-project events
- IP address filtering
- Snowflake export
- And more
Ultimate
- All Advanced features
- BigQuery data export
- Optimizely reporting for experimentation
- And more
Ready to take your digital experiences to the next level?
Optimizely is trusted by over 9,000 leading companies for digital experiences, content management, experimentation and commerce.
FAQs
What are the skillset requirements?
Check out the skillsets page for a full list of skills needed to implement Optimizely’s Feature Experimentation.
What is the size of the datafile?
The empty datafile is around 0.5KB and grows with experiments and feature flags added. However, it is merely a JSON representation of the state of the project, and it does not grow significantly through that. Since this file will not be downloaded for every user, this should not impact performance significantly in most cases. Learn more on this topic here.
How long should I expect implementation to take?
This will depend upon the complexity of the stack and scope of experimentation within that stack. A simple implementation on an iOS or Android App with no integrations and basic event tracking can be done wtihin a day. An enterprise level implementation within a stack of multiple microservices, applications, and multiple clients will take longer and has a lot more factors to consider.
What happens if the Experimentation dashboard becomes unavailable?
From the perspective of Feature Experimentation, if the Experimentation dashboard becomes unavailable you will no longer be able to edit experiments or feature flags using that dashboard. Your experiments, feature flags, event tracking etc. will continue to function as they used to as the datafile is independent of our dashboard. However, you still can make changes via the API in that case.
Learn more about Optimizely’s Feature Experimentation
Last updated: Apr 05, 2022
