We haven't evaluated other products. We have an in-house product that is missing a lot of features and is very behind from making the test process easier.
Instead of evolving our in-house product with limited resources, we decided to go with Optimizely Feature Experimentation …
Overall, Optimizely Feature Experimentation is an industry leader in terms of experimentation across web and mobile. For apps I would say amplitude does slightly a better job as it is tailored to that niche.
Optimizely Feature Experimentation is better for building more complex experiments than Optimizely Web. However, Optimizely Web is much easier to kickstart your experimentation program with as the learning curve is much lower, and dedicated developer resources are not always …
Optimizely Feature Experimentation is less of a point solution than LaunchDarkly, so LD has a few extra features, but Optimizely offers a much greater solution for experimentation, personalization etc.
Feature experimentation is much more robust and allows more granular control over the decisions you want to make. While Optimizely Feature Experimentation is nice and can be delivered via Optimizely Feature Experimentation's UI, its still not ideal because its brittle and can …
Google Tag Manager was less flexible for the business and required the Google Analytics tool for analysis and metric tracking. Optimizely allows the building of use cases. Optimizely provides real-time data and metrics that are easier to use. GTM provides tracking …
Before we chose Optimizely, we looked at other options like Google Optimize. However, we decided on Optimizely because it excels at A/B testing, even compared to other A/B testing tools that only have basic capabilities. Since we were working on a controlled release project, we …
We used Firebase remote config as well to try to test different behaviorus on the same screen, using boolean flags to determine which one to show, but it was too cumbersone and difficult so we pivoted to only use optimizely as it was much better suited for that,
Highly recommend Optimizely as it’s relatively cost effective, easy to use and their customer service team is very helpful and responsive. Also has robust features which encompasses what Crazy Egg did for us before (heatmap). Also much cheaper than the Adobe suite of products.
I would use Optimizely Feature Experimentation when we would like to run basic experiments where metrics to be tracked are impressions, revenue or clicks. However, most of our experiments are tracking more complex metrics and this functionality is not enough. We still need to do work to analyse the data in our end.
Its ability to run A/B tests and multivariate experiments simultaneously allows us to identify the best-performing options quickly.
Optimizely blends into our analytics tools, giving us immediate feedback on how our experiments are performing. This tool helps us avoid interruptions. With this pairing, we can arrive at informed decisions quickly.
Additionally, feature toggles enable us to introduce new features or modifications to specific user groups, guaranteeing a smooth and controlled user experience. This tool helps us avoid interruptions.
Splitting feature flags from actual experiments is slightly clunky and can be done either as part of the same page or better still you can create a flag on the spot while starting an experiment and not always needing to start with a flag.
Recommending metrics to track based on description using AI
Optimizely Feature Experimentation is better for building more complex experiments than Optimizely Web. However, Optimizely Web is much easier to kickstart your experimentation program with as the learning curve is much lower, and dedicated developer resources are not always necessary (marketers can build experiments quickly with Optimizely Web without developers' help).
Experimentation is key to figuring out the impact of changes made on-site.
Experimentation is very helpful with pricing tests and other backend tests.
Before running an experiment, many factors need to be evaluated, such as conflicting experiments, audience, user profile service, etc. This requires a considerable amount of time.