18 results found
-
Setting significance on experiment level
We are supporting over 25 teams running experiments, running close to 500 experiments per year. The requirements regarding significance level differ strongly from team to team - sometimes even from experiment to experiment.
It would help us a lot if we could set significance level for each experiment explicitely.
We also think this is a low-hanging fruit since we can see from API calls that significance level is actually a variable that already exists per experiment. Also - as results are calculated every time I view the results page, this appears to be possible to implement without too much difficulty?
2 votes -
Results to Experiment Details Page Navigation Improvement
When you click on the Results button of a specific experiment, you want to be able to navigate from the results screen to the Experiment details page. Instead, currently, you have to go back to the main list of experiments, then find your specific experiment, to click into the details.
1 vote -
Versioning within an Experiment
A versioning capability within an experiment, so that when you are building, if someone makes unintended changes, they can revert back to the old version. There is interest in this feature both in the visual editor and for general changes in the experiment around page, audience, etc. These could be tackled separately.
1 vote -
IP Filtering
IP filtering lets you exclude certain IP ranges from showing up in your experiment results. This is also how you can exclude yourself or your company from experiment results. - https://support.optimizely.com/hc/en-us/articles/4410283982989-IP-Filtering-Exclude-IP-addresses-or-ranges-from-your-results
This is currently available in Web Experimentation but not in Feature Experimentation.
Internal stakeholders and engineers are regularly forcing themselves into experiments to demo and debug, and this will be impacting our results. We would like to be able to exclude these
1 vote -
Provide SRM monitoring on the results page
Hey!
As Team Lead Web Analytics I often see teams struggling to solve SRM issues.
It would be immensely helpful if the results interface would provide a graph depicting the user distribution between variants. This seems like a low-hanging fruit for Optimizely and would facilitate debugging SRMs strongly.
Thanks!
1 vote -
1 vote
-
"Status Last Changed" Field
Add a "status changed" date to the FX Flags overview page. Because currently, there is a "created" and "modified" column. However, "Modified" refers to any changes, and there is no easy way to see when a status was changed, so when a rule was changed to its current state (started/paused/concluded)... it would be valuable information for the customer just within the FX UI.
Additional context: https://jira.sso.episerver.net/browse/DHK-2914?focusedId=2087943&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-2087943
4 votes -
Remove unused experiments from user profile service
Currently, user profile service (UPS) maintains a map of user IDs to the experiment IDs they've previously been exposed to and the variation ID they received. The SDKs continuously append to the UPS without any cleanup, even if an experiment has been concluded and is no longer relevant.
Customers have in the past implemented diff logic to compare the live experiment IDs in the datafile vs UPS and remove experiment IDs from the UPS that are no longer in the datafile.
2 votes -
Display bucketing ranges in the GUI (when changing traffic allocation)
I know that Optimizely tries to keep bucketing consistent when traffic allocation and or distribution changes.
It would be great to be able to see the bucketing allocation (eg. 1-5000 for A, 5001-10000 for B) in the interface.
In this example it is simple, but when ramping up (and possibly at the same time changing the distribution of traffic) it would be great to be able to verify via the interface which buckets are being set.
I am aware of Customer Profile Service but I see this as independent from that.1 vote -
Display bucketing ranges in the GUI (when changing traffic allocation)
I know that Optimizely tries to keep bucketing consistent when traffic allocation and or distribution changes.
It would be great to be able to see the bucketing allocation (eg. 1-5000 for A, 5001-10000 for B) in the interface.
In this example it is simple, but when ramping up (and possibly at the same time changing the distribution of traffic) it would be great to be able to verify via the interface which buckets are being set.
I am aware of Customer Profile Service but I see this as independent from that.1 vote -
Disable losing variant(s) during experiment
As of now, in Feature Experimentation, if a variant performs (very) bad, there is no way to deactivate it or set its behavior back to baseline without needing to create a new rule.
This of course slows down experimentation speed.
I do understand that the results of that variant are not usable after setting a variant's behavior back to baseline behavior. However, that is not the issue. The goal is to simply be able to continue letting the test run while disabling a bad-performing variant.1 vote -
Audience integrations
To enable FX personalization, be able to pull audiences from other tools.
3 votes -
Datafile Relay Proxy
If Optimizely's CDN goes down or is inaccessible, the SDKs don't have a default fallback mechanism to evaluate feature flags without access to the datafile hosted on the CDN.
It's possible to initialize the SDK with a cached datafile, but that requires custom logic. Ideally, Optimizely could provide a default mechanism to provide a fallback datafile (e.g., a "relay proxy" service that caches the datafile, or a default mechanism within the SDKs).
1 vote -
Flag Triggers/Kill Switches
Make change to flag (e.g., toggle on/off) based on event from external system (e.g., APM alert).
2 votes -
Dependent flags
Be able to trigger flags if certain actions happen, or make flags dependent on other flags. Both stateless and stateful approaches.
2 votes -
Environment Filter for Webooks
Filter activity notifications via webhook to only production environment.
1 vote -
OpenFeature SDK Support
Make SDKs compatible with OpenFeature standard.
1 vote -
Introduce concept of Feature Flag Types
Almost none of my customers implement long-term flags. Every time they run a new experiment, they create a new flag, and remove it after. Even if they could have reused the flag for a new experiment a month later. This creates a lot of additional work.
1 vote
- Don't see your idea?