Statsig Developer Tier
StatsigReviews from AWS customer
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
342 reviews
from
External reviews are not included in the AWS star rating for the product.
Super Useful
What do you like best about the product?
Statsig makes it so easy to experiment with new features and measure their impact in a truly data driven way! It exposes the true impact of your changes and allows you to focus on making changes that benefit your customers
What do you dislike about the product?
Once you go past 5M events per month it starts to increase in cost but you should be getting out more than you put in by then!
What problems is the product solving and how is that benefiting you?
It makes A/B testing so easy to perform and reduces the burden on engineers trying to test different versions of a feature
Very clear results for experiment analysis
What do you like best about the product?
What stands out to me is that there is a clear difference between experiment and feature flags. You can create as many combinations of releases as you'd like with users and its relatively straighforward to modify these!
What do you dislike about the product?
Sometimes the clarity between experiment groups and feature flags and then hold out groups is not very clear and creates problems in production with users not being where they are expected to be. I would like there to be an easy way (workflow) that if a user is assigned to a particular layer (e.g., access to all of our feature flags) for them to be automatically removed from holdout groups.
What problems is the product solving and how is that benefiting you?
I think the main key things statsig has been a key in helping us deliver, build and release features, are:
• Gradually roll out features: We launch new features to a small percentage of users first, so we can identify and fix potential issues before a full release, reducing the risk of shipping broken functionality. Before we used statsig it was horrible and we'd be risking our affecting all our user base (100s of thousands of users).
• A/B test and optimise designs: We can experiment with different design variations and keep only features that actually enhance user experience without causing friction or overwhelming users (easy access to this data also by using mixpanel in combination). This sometimes goes as small as just adding a simple emoji to a text and clearly showing that has better statistics!
• Make data-driven decisions: Instead of relying on intuition, we use data with its statistical significance to assess whether a feature improves engagement, retention, and overall the number of users that are premium!
• Personalise user experiences: We use the dynamic feature gates to tailor user experiences based on various factors (behaviour, demographics, or preferences), and we have seen this to lead to higher satisfaction and conversion rates to premium.
• Automatic rollback: If something breaks production or if it just simply affects our key metrics badly, we can simply rollback automatically or just add in quick adjustments without requiring a hot fix in production.
• Increase development efficiency!: The engineers in our team can now confidently ship faster as they are less worried about deploying features. This has made a huge difference in ensuring a smooth release process without affecting our products stability!
Statsig has been a game-changer for us, making our development process so much better! Having this level of experimentation and feature control has significantly improved how we deliver value to our users.
• Gradually roll out features: We launch new features to a small percentage of users first, so we can identify and fix potential issues before a full release, reducing the risk of shipping broken functionality. Before we used statsig it was horrible and we'd be risking our affecting all our user base (100s of thousands of users).
• A/B test and optimise designs: We can experiment with different design variations and keep only features that actually enhance user experience without causing friction or overwhelming users (easy access to this data also by using mixpanel in combination). This sometimes goes as small as just adding a simple emoji to a text and clearly showing that has better statistics!
• Make data-driven decisions: Instead of relying on intuition, we use data with its statistical significance to assess whether a feature improves engagement, retention, and overall the number of users that are premium!
• Personalise user experiences: We use the dynamic feature gates to tailor user experiences based on various factors (behaviour, demographics, or preferences), and we have seen this to lead to higher satisfaction and conversion rates to premium.
• Automatic rollback: If something breaks production or if it just simply affects our key metrics badly, we can simply rollback automatically or just add in quick adjustments without requiring a hot fix in production.
• Increase development efficiency!: The engineers in our team can now confidently ship faster as they are less worried about deploying features. This has made a huge difference in ensuring a smooth release process without affecting our products stability!
Statsig has been a game-changer for us, making our development process so much better! Having this level of experimentation and feature control has significantly improved how we deliver value to our users.
Go to tools for running experiments and rolling out new features
What do you like best about the product?
Running experiments is a breeze. Ability to define custom metrics, explore each metric individually through a metrics explorer, ability to view sample events is really helpful. User interface is very smooth and easy to use.
What do you dislike about the product?
not dislike but currently the experiment results are delayed by one day. there should be an option to view it in real time
What problems is the product solving and how is that benefiting you?
rolling out new features reliably
conducting experiments
conducting experiments
Statsig - super easy-to-use feature-flagging/experimentation tool
What do you like best about the product?
As a relatively new user of statsig I found it really simple to get started with as the setup was pretty quick and there was no real issues with getting things working. The documentation is also clear and full which were easy to follow and also helped with the onboarding flow. The onboarding was also helped with the simple and pretty seamless integration into our workflow without causing any major hassle. Especially, for something we use so often - I highly recommend it!
What do you dislike about the product?
I haven't had any major pain points with statsig so far - the only minor things I can really nitpick at is sometimes the UI could be clearer - granted it is super intuitive but some parts there would be a couple buttons disabled/greyed out and it wouldn't be clear as to why they are disabled.
What problems is the product solving and how is that benefiting you?
Think the key point is instead of spending lots of engineering time building our own feature flag system, Statsig gives us a solution which is essentially good to go out the box and helps us quickly test and roll out new features. The A/B testing also makes it pretty easy to track what is working and whats not which takes out any guess work.
Efficient Experimentation with Stellar Support
What do you like best about the product?
I find Statsig incredibly easy to use for running experiments, thanks to its robust statistical engine which significantly enhances the precision and reliability of experiment results. The experiment results visualization and user experience are outstanding, making it simple for stakeholders to understand, which streamlines our communication process. I appreciate Statsig’s commitment to continuous improvement, with new features being added constantly that enhance its functionality. The support provided through their Slack channel, even at the pro tier, is exceptional and helps me resolve issues promptly. Moreover, the platform saves me a lot of time on experiment speed, analysis, and stakeholder onboarding, making it an invaluable tool for improving workflow efficiency.
What do you dislike about the product?
I had some issues with tool integration breaking unexpectedly, which was problematic as it required me to identify the problems myself despite them being resolved quickly. Additionally, the ability to drill down into experiment results with statistical calculations is limited to a single breakdown property. Although I can export the data using custom metric explorations, I miss out on all the statistical calculations. Furthermore, with the rapid pace of new features being added, it would be beneficial if Statsig could provide guidance on new features I could implement or alert me when my current implementation methods are outdated.
What problems is the product solving and how is that benefiting you?
I use Statsig for server-side A/B testing with its strong statistical engine, saving time on experiment speed and analysis. Its visualizations ease stakeholder onboarding, and support helps quickly resolve issues.
Using Statsig daily to release and experiment with new features
What do you like best about the product?
Our initial use case for Statsig was for feature rollout using gates. Statsig has revolutionised this process for us - to a point where it seems strange that we never had it in the first place. Through Statsig's feature gates we've had a lot more confidence in rolling out features. In particular, we love the automated rollouts as it means that we can expose a feature, monitor its performance against key metrics, and then proceed once our confidence is high.
Monitoring is also very easy in Statsig. I have set up a dashboard for my team which includes all of the feature gates and experiments that we are currently running. I love how easy it is to set up the dashboards by just tagging any relevant gate/experiment with my team's custom tag.
I also love how easy it is to configure the UI. I now have things set up so that when I open the app I am presented with all the information I need straight away without having to dive into all the menus. This is very helpful when I just want to check the status of an experiment or stage of a rollout.
Monitoring is also very easy in Statsig. I have set up a dashboard for my team which includes all of the feature gates and experiments that we are currently running. I love how easy it is to set up the dashboards by just tagging any relevant gate/experiment with my team's custom tag.
I also love how easy it is to configure the UI. I now have things set up so that when I open the app I am presented with all the information I need straight away without having to dive into all the menus. This is very helpful when I just want to check the status of an experiment or stage of a rollout.
What do you dislike about the product?
I don't feel like there are major downsides to Statsig - all of the functions work great and we haven't had any trouble setting things up or using the features that are provided.
The one thing that would be nice from an interface perspective would be to have more in line documentation about a certain feature. This would lower the boundary for entry. I find myself having to flick between the docs and the app at times which can be frustrating. At the very least it would be great to have a button which links to the docs for a particular feature. I.e. if I want to implement a layer but I'm not 100% sure on the functionality - I should be able to get to the docs immediately from the `Layer` tab in the web app.
The one thing that would be nice from an interface perspective would be to have more in line documentation about a certain feature. This would lower the boundary for entry. I find myself having to flick between the docs and the app at times which can be frustrating. At the very least it would be great to have a button which links to the docs for a particular feature. I.e. if I want to implement a layer but I'm not 100% sure on the functionality - I should be able to get to the docs immediately from the `Layer` tab in the web app.
What problems is the product solving and how is that benefiting you?
As mentioned previously we use Statsig primarily for rolling out features using the feature gate tooling and performing experiments on certain parts of our app.
Feature gates benefit us by:
1. Allowing us to rollback features easily if there is a problem. This gives us a lot more confidence in our deployments.
2. Allows us to also monitor their success. Because we have configured our own events, it is easy to see if a feature is damaging key metrics.
We also use holdouts to see the impact that multiple features have had on a team-by-team basis. This is great for all the engineers, who can then see how the work they are doing is impacting the performance of our key metrics.
Feature gates benefit us by:
1. Allowing us to rollback features easily if there is a problem. This gives us a lot more confidence in our deployments.
2. Allows us to also monitor their success. Because we have configured our own events, it is easy to see if a feature is damaging key metrics.
We also use holdouts to see the impact that multiple features have had on a team-by-team basis. This is great for all the engineers, who can then see how the work they are doing is impacting the performance of our key metrics.
Used to roll my own feature flags but this is so much better
What do you like best about the product?
My favourite thing is how easy it is to manage feature rollouts and experiments on statsig. The SDK makes sense, and I love checking the pulse results during a roll out. One thing that I've noticed from using Statsig is that our team is better at defining experiments before hand, particularly in terms of what metrics we want to evaluate. It's also made us much more confident when making risky changes. Finally, I find it helpful that all of our gates and experiments are defined in one place and I can see what experiments other departments of the company are running if needed.
I would 100% recommend it!
I would 100% recommend it!
What do you dislike about the product?
I largely think statsig is very good, but there are two pain points that I find challenging:
Firstly, when doing a rollout it's not possible to reverse progress while maintaining the same cohort of experimental data and you need to start a new rollout. For metrics that have a long lead time, this can slow down experiments, and it makes me think it's not necessarily the right tool for assessing longer term impact of experimental changes.
Secondly, ad blockers can impact whether feature gates are evaluated as true. This is a known issue, but we do receive some (albeit very few) customer complaints from people expecting a feature that is currently blocked behind a rollout.
Firstly, when doing a rollout it's not possible to reverse progress while maintaining the same cohort of experimental data and you need to start a new rollout. For metrics that have a long lead time, this can slow down experiments, and it makes me think it's not necessarily the right tool for assessing longer term impact of experimental changes.
Secondly, ad blockers can impact whether feature gates are evaluated as true. This is a known issue, but we do receive some (albeit very few) customer complaints from people expecting a feature that is currently blocked behind a rollout.
What problems is the product solving and how is that benefiting you?
We use statsig to evaluate changes to our application and back end systems in a controlled way before rolling changes out to 100% of users. The ability to track outcomes of certain metrics by cohorts in the experiment, and see the impact of a feature, is a great benefit. We also benefit from not having to develop our own feature gate system in each repository that we work in.
Pretty positive - makes continuous roll-out easy
What do you like best about the product?
The flexibility it allows for custom rollouts. Layers have been a great feature used for more complex ml model rollout solutions.
In addition simple binary feature flags are a great way of getting things off the ground
In addition simple binary feature flags are a great way of getting things off the ground
What do you dislike about the product?
Have noticed some weird behaviours when the connection using a statsig client fails
What problems is the product solving and how is that benefiting you?
We wanted to have a custom ML model roll out solution where we could track the success of new models based on key metrics.
We developed a routing system which relied on a statsig layer to deliver traffic to different ML models
We developed a routing system which relied on a statsig layer to deliver traffic to different ML models
Makes experimentation easy
What do you like best about the product?
It's just very straightforward to rollout features and experiments to our customers and understand their impact. We use it across our tech stack with different languages and the SDKs all work well and without issues. Also get a quick response on the slack support.
What do you dislike about the product?
Takes a little time to understand the nuances of the different features (e.g. why you can't change what percentage of users go to a group in an experiment once it's started)
What problems is the product solving and how is that benefiting you?
Main three things for me are
1. enabling a gradual roll out of features so we can gain confidence with a small subset of customers
2. allowing access to different groups of users (e.g. internal team, beta testers etc) is super straightforward
3. experiments and layers are great tools for testing variants and understanding their impacts
1. enabling a gradual roll out of features so we can gain confidence with a small subset of customers
2. allowing access to different groups of users (e.g. internal team, beta testers etc) is super straightforward
3. experiments and layers are great tools for testing variants and understanding their impacts
A brilliant platform for feature flag, a/b tests and experimentation
What do you like best about the product?
We use Statsig for all our feature flags which allows us to release new feature quicker and safely.
We also use Statsig for all our experiments which enables all our product teams to use the platform
We also use Statsig for all our experiments which enables all our product teams to use the platform
What do you dislike about the product?
Actually no downsides, its one of the best platforms out there for product teams and engineers
What problems is the product solving and how is that benefiting you?
- Feature Flags
- Running A/B tests
- Running Expirements
- Using Holdout Groups
- Allowing our internal teams to test features early
- Running A/B tests
- Running Expirements
- Using Holdout Groups
- Allowing our internal teams to test features early
showing 211 - 220