Turn any feature into an A/B test, free and out-of-the-box. Let Statsig take care of running clean experiments and give you back results really really fast, using best-practices and data science techniques that power Facebook, Uber, and Airbnb.
Features come and go, but your business metrics are here to stay. Pulse helps you identify the overall impact of any feature to your core business metrics before you launch.
Take control of your business. With Holdouts, instantly discover causes for any trend. Double down only on what matters.
Pulse automatically computes how every new product improvement impacts your business metrics. This empowers you to release new features and functionality with conviction and uncover any unexpected impact to better understand your customers.
Pulse enables every engineer, data scientist, product manager, designer, and busines leader to view the same source of truth and make better decisions at all levels of the organization.
A/B testing is the scientific gold standard for understanding causality. Also known as split or bucket tests, A/B tests pinpoint what changes cause which effects.
Once you set up Feature Gates, Statsig automatically calculates statistically significant test and control groups with randomly-selected users, labeled A & B. The test group sees a new feature while the control group sees the existing experience.
Statsig's out-of-the-box A/B testing results put the power of industry leading experimentation platforms used at Facebook, Airbnb and Uber at your fingertips.
Holdouts show the aggregate impact of a set of features. With Holdouts, you can now quantifiably monitor how much progress you have made towards your strategic goals over a given period.
You can set up a holdout by creating a holdout user group, the control group, and selecting the features that you want to compare. Pulse automatically computes the aggregate impact of all selected features compared to the holdout user group.
Statsig brings enterprise grade reliability and high availability to your tools. The service is spread across multiple regions, with three-way redundancy across regions and availability zones on AWS and Azure to ensure that it never goes down. You can rest easy with your product relying on rock-solid infra.
Our SDKs are resilient against last-mile network issues and perform most of the heavy-lifting on the client-side, ensuring decisions are made instantaneously, without a server round-trip.
Statsig’s Experiments+ tests multiple hypotheses (or treatments or A/B/n tests) in a single experiment. With Experiments+, you can also run multiple independent experiments in parallel with exclusive user groups to prevent collisions.
You can set up a Statsig experiment in three simple steps within minutes. For every experiment, Pulse automatically computes the estimated lift for the primary metrics you expected to change with the statistical significance of these estimated lifts.
Experiments+ empowers you to use the most trusted scientific analysis and results to develop deep insights about your customers at no extra cost.
Autotune automatically computes the optimal setting for a parameter in order to optimize for a key metric. Autotune is useful in several optimization scenarios, from tuning ad copy and catalog artwork to user interface design and website optimization.
You can run an Autotune optimization by simply selecting the variants you want to test and the key metric that you want to optimize.
The Users tool provides breakdowns at the user level, including any attributes sent with a Statsig SDK and those inferred from the browser or device used to access your service, in addition to any feature gate or dynamic config checks.
Ultrasound lets you view how a metric is affected by your features. It can identify the cause of sudden or long term drops or rises in your metrics.