116 results found
-
Multi-device Experiments
Most experiments need different images for desktop, mobile, and tablet. Without a shared library, the customer has to either create a separate experience per device — splitting traffic — or write CSS manually. Images are also
locked inside each variation and cannot be reused across experiments. A Media Library exists, but it is a paid extra. A basic shared repository should be included as standard.1 vote -
Change approval workflow in Web Experimentation
I'd like to have a change approval workflow for launching and updating experiments in Web Experimentation. Today I need to use workarounds like applying a QA audience until the experiment is "approved" by the right individuals.
1 vote -
UX Change: Run Time Calculator
Masorange has requested the run time calculator to be usable without having to create an experiment first. Once an experiment is set up, the calculator can be accessed, but it was flagged as difficult to locate and sometimes they want to use the calculator ahead of set-up. The ask is to move it to a more prominent and easily discoverable location.
4 votes -
Multi-Armed Bandit Results
Can we add a traffic split visualization to the results dashboard? This would let the team verify that the bandit is working as intended and see exactly how traffic is being rerouted to the top-performing variations in real-time.
1 vote -
Word Wrap Option in Visual Editor
When editing HTML - allow the text/code in the HTML box to wrap. This makes it easier to see and edit.
1 vote -
Experiment Organization Folders
It would be great to be able to create folders in the experiments list to group items together. (Not the sort feature). Often, I have multiple experiments (separate mobile and desktop project that are the same content) - it would be useful to be able to organize these into groups - like in a file folder system. (See mockup)
7 votesWe are planning to redo some parts of the dashboard and the setup of certain things. As a lot of creation things will be moving into the new VE, we will be taking this into account and I will be reaching out when we get to the dashboard part.
-
experimentation forced re-evaluation each decision - override UPS / sticky assignment
As an experiment builder
I would like to be able to force re-evaluation of flag/campaign audiences if I want my experiment to float by user properties and am working in web or have a UPS set in feature
So that I can have transactional experiments that will match the user up properly to their current properties and not persist them into something they no longer belong in.
1 vote -
View experimentation sections from editor
It would be nice to access all the experimentation/personalization modules from the editor. In the old editor, we always had the option to click any module from th left, like metrics, shared code, targeting, etc., but now, the editor opens on top of that dashboard, and the only way to go back is by clicking 'return to setup,' which makes it a longer process to go through. It is also very slow, so I take too long checking all the sections of the experiment/personalization. I get that the editor now takes up the whole screen, but adding a drop-down to…
1 vote -
NEW Responsive Mode
I'm glad to see that the Emulator has been improved. But there are some "WTH" issues in the UX. In no particular order...
DON'T blur the content when resizing -- I frequently want to see how objects flow as the browser width changes, this exposes flaws in CSS or coding and I can't see what's happening when It's blurry.
Disabled the editor in the Emulation mode (Add one of the fingers at the top bar to disable it, next to the rotate button) - I need to engage with the site and without it disabled it think I want to…
1 vote -
reusable/transportable variants in web experimentation
As a user building variants in web experimentation,
I would like to be able to save and reuse variants across campaigns, pages and projects while deciding at the campaign/page level whether to allow the selected version to stand apart or be kept in sync (copy local versus linked). Localized from linked save could require a new name but retain the link to the original and allow for a button to copy the parent for resync/update needs.
So that I can more easily have a dev, qa, prod environment setup in web experimentation.
This would likely require that variants live outside…
2 votes -
Experiment Activation and associated changes
AS
an experiment builder in webI WANT TO
be able to make changes to pages associated with the experiment and not have the visitor activate the experiment when those pages loadSO THAT
I can keep changes related to an experiment together while limiting activation of the experiment to a specific page or actionUSE CASE
I have a change for a specific product display page that I want to evaluate. There are associated changes that I will need to make to the cart and every checkout page. In order to make changes to those pages to support this,…1 vote -
Developer experience
I want Optimizely’s new developer experience to support a true IDE-first, GitHub-native workflow. I should be able to do all coding in my own IDE, manage changes through GitHub, and publish through automation and CI/CD workflows without depending on the Optimizely editor for core development tasks. The editor should support preview and collaboration, not replace standard developer tooling and deployment workflows.
1 vote -
I want Optimizely’s new developer experience to support a true IDE-first, GitHub-native workflow. I should be able to do all coding in my ow
Software Engineer
1 vote -
o Adding pagination to events to limit the number of events loaded at once.
Currently, loading the full list of events puts a significant load on the page.
The Events view becomes very slow, and in many cases it takes several minutes before the events finish loading.
Pagination (or lazy loading / load more) is needed to reduce the amount of data rendered at once and to improve usability.1 vote -
Opal UI - is Awful!!
Opal is very difficult to see -- there's no contrast between it and a page in reporting (or anywhere Opal appears)
Recommend giving her some color! See screenshots.
1 vote -
Bug Fix
Bug Fixes on text editor (mainly shared code)
Native find on page doesn't work reliably (sometimes shows more occurrences than there actually are, doesn't jump to the place in the code when you hit enter)
- Would be great to implement that properly
Starting typing at all makes the current line jump to the bottom of the screen, which is really frustrating and quite confusing.
- Ideally just keep the scroll position what it is when you start typing.
(I'm using Chrome on Mac OS)
1 vote -
Version control
A slightly better interface of what changes have been made before committing. I know you can view history, but it's not that easy to understand. Being able to see lines of code that have changed in each of the shared js, shared css and variations would be very helpful, rather than just the number of lines changed.
1 vote -
experiment publish note
AS an experiment builder and program manager,
I WOULD LIKE to be able to add a custom note each time I publish an experiment in both web and feature experimentation and have that note show up in history as part of the list view and details
SO I can more easily find changes in history by seeing the note displayed in the history list and associate changes to ticketing systems when appropriate.
2 votes -
results page versions
AS An
experiment analystI WANT TO
save the settings of the results page including start/end time, segments, etc... and have it accessible via share link and also as a link displayed on left sideSO THAT
i can share detailed and specific findings with others easily and return to the ones i've created easily.1 vote -
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
.png){kind=link}
- Don't see your idea?