116 results found
-
Title: Improving Revenue Experiment Accuracy for High-Variance Transactions
Optimizely currently defines revenue outliers using a fixed threshold of mean + 3 standard deviations. This approach works well for classic e-commerce use cases with stable pricing, but it becomes less reliable in scenarios where transaction values are inherently unpredictable.
In donation-based fundraising—and in other models with variable or open-ended transaction values (e.g. donations, contributions, pay-what-you-want, grants, or memberships) - there is no fixed price catalogue. As a result, a small number of very high transactions can disproportionately influence revenue and average gift metrics.
After reviewing this with our analysts and experimentation team, we found that applying a mean +…
1 vote -
allow navigator.sendBeacon() calls for events in web experimentation
As an experiment builder
I would like to utilize the navigator.sendBeacon API to send events to Optimizely
So I can ensure all events are accounted for and my metrics are more complete and likely to match internal measures of same.
https://developer.mozilla.org/en-US/docs/Web/API/Navigator/sendBeacon
1 vote -
Improved clarity around scheduled experiments
As an experimentation user, I want the UI around scheduled experiments to more clear about the experiment status.
Currently, the experiment status is either "Not Started" or "Paused". It would be great if the experiment status instead was simply "Scheduled", or alternatively "Paused (Scheduled)" / "Not Started (Scheduled)".
Additionally, when changing a paused scheduled experiment, there is a warning "Publish your experiment if you want your draft changes to go live on the scheduled date.". However, when clicking Publish, the preferred option "Publish Only" is not highlighted and instead "Publish and Start" is highlighted. This might cause someone to accidentally…
1 vote -
Ability to track MAU consumption per project when using Custom Snippets
Currently in admin centre the reporting available for MAU's per project does not provide data if the customers projects were set up using custom snippets. As custom snippets are a popular way to set this up, it would be great to be able to able to report on this.
1 vote -
variation editor option to publish directly
As an experiment builder that generally uses the code editors and who uses the cookie and querystring model for qa and works with intricate code that is load dependent and requires more than a few updates to get working properly within the actual site code
I would like to have the ability to publish the variation update on the variation screen (perhaps a 'publish' button)
So I don't have to save the code changes, then go to main experiment screen and publish and then navigate back to the variation and wait for the page (which will fail because it is…
1 vote -
variation screen editor page load option
As a builder of web experiments that always uses the code editors or targets locations inaccessible to the editor
I would like to have a setting that stops/blocks the page load every single time you pull up the variation code. It would be nice to be a persistent setting that could either stick to the variation or to my login.
So I don't have to click through error messages that are not relevant every time I go to edit.
1 voteThis is planned, Q3 at latest.
-
Deployed experiments (winner rollout) should be active in Preview mode for other experiments
Deployed experiments (winner rollout) should be active in Preview mode for other experiments.
Currently, when an experiment is concluded, and the winning variation is deployed, this deployed variation is active for all visitors on the website. However, when Previewing another/new experiment, that deployed experiment is not active.
Since the deployed variation should be considered the "is-state" of the website, and it is publicly visible to regular visitors, it should also be visible in the Preview mode.
1 vote -
program reporting data
As an experimentation program manager in an experimentation COE with no access permissions to build a system to assess API data but with access to data tools that I can use to build out queries and dashboards.
I want to be able to access meta information on experiments (the same data set accessible via the API) in my s3 (bigQuery, etc.) feed.
So I can build queries and dashboards to assess and report on the health of the program, create alerts, etc.
1 vote -
event / metric tagging
As an experimentation program manager in an experimentation COE
I want to be able to 'tag' events/metrics with properties like 'funnel', 'engagement', 'conversion'. This information should be available via export in some manner - s3 type of export preferably.
So I can analyze the metrics use in experiments to better understand what our teams have been / are focusing on and understand the health of the program better.
1 vote -
Show which audience a visitor matched when an Experiment has multiple audiences.
Show which audience a visitor matched when an Experiment has multiple audiences.
Both the Audience builder and the Experiments builder support OR clauses. However, there is no way to dissect the data to know which specific Audience triggered the experiment for the visitor. Our organization will run experiments where we may use a Javascript based audience and a Real-time segments audience to trigger a single experiment. The Experiment report will show the visitor to variations counts but there is no way to tell which audience the visitor matched.
The suggestion is to have a breakdown of the visitor counts, by…1 vote -
Conditional Activation: Select Multiple Conditions
Suggestion:
Ability to add multiple conditions.
At the moment we can only select 1 condition, but selecting multiple (2 or more) conditions would be a great add.
Use Case - Ecommerce example:
if a new user clicks over the jackets category and has an exit intent, we show them a pop up with a promotion
1 vote -
Conditional Activation: Scroll To % option
Topic: Conditional Activation
Use case: run a global experiment with a pop up, but not all pages have the same elements so we need to be able to scroll to % on the page.
Solution: Scroll To option, add the possibility to scroll to % (not just element).
1 vote -
Include ability to target html elements based on their css specificity.
Currentlh when using the Visual editor and an element is selected such as an H1, the css class or ID attributes are not accounted for in the editor.
Sometime is is necessary to be more specific, such as
H1 vs H1.classname
The idea is to include an option to use specificity and have the editor use the ID or classname based on a setting.
Such as...
Use HTML element only (H2)
Use HTML ID (#id-name)
Use HTML element + CSS Class name (H2.classname)1 vote -
Support for rearrange tests on Dynamic websites
Optimizely Web Experimentation does not currently support rearrange tests (tests that swap the positions of nodes in a document on dynamic websites but may extend support in the future.
Adding support will mean elements in Dynamic pages can be rearranged and the objects will be re-applied like the other support for changes.1 vote -
Shortcut to turn off interactive mode
I often have customers that struggle to change fly-out menu's. They have to select the interactive mode to open the flyout, but then they can't select an element as they have to turn off the interactive mode. By moving out of the flyout to turn off interactive mode this closes the flyout again however, meaning they can never select an element in these flyouts.
If there was/is a shortcut to quickly turn off interactive mode that would solve this I believe.1 vote -
Shared code - double scroll bar
On the Shared Code section, when the code is too long, you get a double scroll bar on the right. It is not particularly problematic, but it bothers the UX a bit.
1 vote -
Result reset warning should include the name of the test
To avoid accidentally resetting the wrong test, it would be helpful to include the name of the test whose results are being reset in the warning message.
1 vote -
Metric Selector
The new metric selector has made the process of selecting metrics take a long time.
1 vote -
List Experiments endpoint - filter by experiment status
Please add the option to filter results by status of the experiment.
1 vote -
Display run dates for archived experiements
For archived experiments, it would be helpful to quickly see the experiment's run dates (start and end without having to dig through event history.
1 vote
- Don't see your idea?