A/B TESTS AND EXPERIMENTS

Product experimentation

Maximize the business impact of every software feature you ship.

experimentation

Realize the promise of experimentation

LaunchDarkly removes silos and complexity from the experimentation process, allowing any team to easily set up, run, and act upon valid experiments.

Multivariate (A/B/n) experiments

Quickly test multiple feature variations with multivariate experiments.

Metric groups

Create reusable metric groups that track how experiments affect key funnel metrics.

Traffic allocation

Define your experiment audience—either a targeted segment or a random sample percentage of users. Prevent carryover bias.

Shipping experiment winners

Release winning feature variations with a few clicks by connecting experiments to the engineering process.

New

Funnel Experiments

Easily build out-of-the-box funnel optimization experiments to improve conversions and other KPIs.

Learn more

Case study

CCP Games creates self-serve experimentation.

/ /

LaunchDarkly has enabled self-serve experimentation. You don’t have to be a data scientist to run valid, actionable experiments. This is unbelievably powerful.

Nick Herring

Technical Director of Infrastructure, CCP Games

Read case study

Case study

Ritual delivers customer experiences with experimentation.

/ /

The main benefit of using LaunchDarkly experimentation is that it’s directly tied to the feature flag implementation. It’s not a separate implementation every time we want to test a feature.

Daniel Archer

VP of Engineering, Ritual

Read case study

Case study

Loom runs more experiments, increases product engagement.

/ /

Our old approach to experimentation was too complicated from an engineering perspective. It ate up developer time that would have been better spent on building actual features, not writing custom targeting rules.

Steve Milburn

Senior Software Engineer, Technical Lead, Loom

Read case study

drive growth

Feature flags + experiments = revenue growth

LaunchDarkly layers experimentation on top of feature management, making it easy to run controlled experiments and ship winning features from experiments.

Learn more

Create a flag

Wrap every new feature in a feature flag.

Create and assign metrics

Define the performance data you want to collect: conversions, clicks, page load time, etc. 

Define your sample audience

Create a targeted experiment sample or select a random percentage of users tied to a funnel stage.

Start recording

Record data for assigned metrics and tweak the experiment as needed.

Visualize results

Illustrate data stories with reliable data, giving key stakeholders a shared view of results.

Identify and ship the winner

Release the winning feature variation to the full intended audience immediately.

Server-side experimentation

Full-stack product analytics and optimization

Unlike the scores of visual editors on the market, LaunchDarkly lets engineers run back-end and infrastructure optimization experiments. For example, you can...

Easily run an experiment to measure the impact of a new API on page load times

Measure the impact of a new database implementation on system performance

Measure the impact of a new search feature in your mobile app on latency

Data export

Powerful insights at your fingertips

Export feature flag event and experiment data to destinations like Azure Event Hubs, Google Cloud PubSub, Amazon Kinesis, mParticle, and Segment. Then push it to your favorite analytics tools like Tableau, Looker, and Power BI for deeper analysis.

New

Snowflake Data Share
Export experiment data to your Snowflake data warehouse and build custom analyses on the results.