Statsig makes it really easy to get experiments up and running without a ton of overhead. I use the Power Analysis Calculator often — it’s straightforward and helps me scope tests quickly. The metrics catalog is especially useful for creating a single source of truth and reduces the need for reconciling results between sources later on. The customer support team is also quite knowledgeable and responsive to questions. Review collected by and hosted on G2.com.

There’s a bit of a learning curve when it comes to setup, especially if you're not continuing from a previous test. I’ve run into moments where it wasn’t obvious how certain events were being ingested or why the targeting logic wasn’t working as expected. Some of that could be improved with more in-product guidance or clearer documentation. I also wish debugging flows were a little more transparent — it can take some back-and-forth to figure out why a user didn’t qualify or why results aren’t showing up as expected. Review collected by and hosted on G2.com.

This is my first time using a third-party A/B testing tool. I like that Statsig has a lot of advanced functions, such as stratification, CURE, Sequential testing, etc. These features enable complex tests and also guarantee the qualify of the tests. I'm learning from Statsig about what the best practices other companies have in testing to improve my tests. The UI is easy to use and understand. The customer support is very helpful, they provide customized solutions to challenges we met. Review collected by and hosted on G2.com.

It would be great if custom query feature could be more flexible, allowing more options when customizing, automatically include new dates of allocation for scheduled queries.

PM and analytics don't always know what setting are made for the test. It would be great if Stasig can make the implementation more visible to product so that non-eng stakeholders can get a better sense about how the test is set up. Like what's the condition applied, what feature gate is used, etc. So cross functional team can help evaluate whether the test is set up in the right way as designed. Review collected by and hosted on G2.com.

Lots of features for both engineers and data scientists to make educated decisions on A/B experiments, such as filtering out outliers, custom assignment sources, winsorization, etc. Analyzing experiments from within the web interface is easy and there are many tools available such as the Explore queries where you can cut by custom dimensions, adjusting the CI percentage, adding and removing metrics. Additionally, customer support is very helpful and friendly whenever we have questions or encounter issues. Review collected by and hosted on G2.com.

Some of the "edge cases" are not clear to the user - Segments are not supposed to be more than 10K rows, but the system doesn't stop you from going over that limit - it simply becomes unstable once you do. Other things - when calling Statsig API from offline jobs (such as cron) there is a default 500 batch limit and a 60-second flush interval, so if your job has < 500 exposures and finishes in < 60 seconds, those exposures get lost unless you explicitly call Statsig.flush() before the pod exits. Review collected by and hosted on G2.com.

Statsig is easy to use and highly intuitive, even for non-technical users. The interface is clear, and it simplifies experimentation setup and analysis. It's been very helpful for running quick experiments and understanding impact with minimal effort. I especially appreciate the ability to integrate it seamlessly into our workflow and quickly validate changes. Review collected by and hosted on G2.com.

The main challenge we've faced is controlling the consumer-side experience from the creator-side configuration. When trying to roll out experiments across both roles (e.g., creator and consumer), the default 50/50 split no longer behaves as expected, and this limits our ability to test full end-to-end user flows. A clearer mechanism to manage these kinds of multi-actor experiments would make a big difference. Review collected by and hosted on G2.com.

Flexibility to allow the rules with which to rollout gates and configs. Custom fields enable many use-cases other than user focused features, and the use of schemas for dynamic configs gives us protections against user error. The fact that metrics are seamlessly integrated with rollouts is also convenient, providing an easy to access and interpret interface that translates a feature into data impacted data points, considerably speeding up data analysis. I do use it all the time Review collected by and hosted on G2.com.

No support for mapping users with different ID types, we have both string and integer type ids however we're limited on which one we can use (integer) if we want metrics. It took quite a while to provide a Rust SDK. No easy way to define "static" gates. Aka. a rule which only cares returns a boolean a given % of time, I do feel like gates are overkill for that goal. Review collected by and hosted on G2.com.

The UI is very easy to use to set up gates, experiments, configs. High level documentation is easy to follow and understand, and I found some YouTube videos that further explain concepts which is great. Slack support from the Statsig team was appreciated! Review collected by and hosted on G2.com.

Setting up the SDK had some limited documentation and we ran into a unique bug about login and value flashing that we didn't really have any reference for. Additionally there's some limitations about experiment setup and environments - you can choose different distributions for different environments but you can't choose different populations (50% of population in prod vs 100% of population in staging). Review collected by and hosted on G2.com.

Statsig is very robust and have been very useful in setting up the experiments. We have integrated Statsig with our vendors like SFMC. CDP and it has supported pretty well. Review collected by and hosted on G2.com.

We are doing POC with our vendor and since, we would like to know adhoc investigation and do lot of monitoring. We would need dashboards. We tried running dashboards in stastig but we felt it lacks maturity and can show only basic graphs. In future, this could be good opportunity to make complex dashboards as leadership have access to stastig and sometimes don't have tableau/databricks notebook access. Review collected by and hosted on G2.com.

I love that Statsig is so much more than just an A/B testing tool. Having feature flags, a parameter store, and experimentation all in one platform is incredibly versatile and has simplified our workflow immensely. On top of a great product, their support team is consistently fast and resolutive whenever we need help. Review collected by and hosted on G2.com.

While our overall experience has been very positive, one area for improvement is the implementation of user stitching. The feature itself is powerful and works well once configured, but we found the setup process to be quite complex. It requires specific technical knowledge that wasn't fully covered in the documentation, which created some initial hurdles for our team. Review collected by and hosted on G2.com.

I appreciate that I can self-serve create experiment dashboards and not having to wait on data science resources to create them to quickly monitor the health immediately after experiment launch. The ease of use of being able to deep-dive analysis by different cohorts defined by the data science team is very helpful as well. Review collected by and hosted on G2.com.

This is more of a setup rather than a downfall, but though I enjoy the self-serve aspect, it required a lot of coordination and time to get there. In order to onboard new funnel events to measure as primary and secondary metrics or even new cohorts, it requires data science resources. I also wish I can rerun past queries without having to clone it, or even save past queries but adjust the experiment id. Review collected by and hosted on G2.com.

The Statsig team is highly skilled and any person on support is able to support most of the products and features they provide. During both the support and the first time we integrated with the platform Statsig provided easy to use documentation including code examples which accelerated our rate of integration. They provide a high number of features and additionally they implemented new features we requested. Each of the new features was delivered in high quality. The tooling works as expected and each of their design decisions is well thought through. The platform has no problem handling high traffic. Review collected by and hosted on G2.com.

The rate at which they add new features is very high and sometimes a new feature is enabled before we had a chance to onboard our internal stakeholders. Review collected by and hosted on G2.com.