28 results found
-
Edge Delivery with Fastly CDN
I am the Product Manager at Hearst Magazines.
Our current Web Experimentation implementation is set up with the snippet in the <body> tag, which causes significant flicker for experiments involving changes above the fold.
Your new Edge on CDN feature could be a game changer for us. However, the blocker is that we currently use Fastly CDN, which is not supported by your Edge product.12 votes -
Multi-Select Metrics using Checkboxes
To shorten the set-up time, rather than clicking each metric separately, Michael proposes an option to multi-select metrics in the dropdown with checkboxes.
5 votes -
warning when changing audience
As an experimenter, I should see a warning/alert when changing an Audience in either Web or FX that is being used in an active experiment to prevent unwanted impacts on live experiments.
3 votes -
Allow unsafe-eval
When I use the Preview button within Web Experimentation, I see the following error in my developer console:
Uncaught EvalError: Refused to evaluate a string as JavaScript because 'unsafe-eval' is not an allowed source of script in the following Content Security Policy directive
When I reached out to Optimizely support, they confirmed that this is a known issue; Preview Mode uses the eval() function which is only possible if your CSP allows unsafe-eval.
As allowing unsafe-eval is not recommended due to the security risks and would essentially nullify the protections we gain from CSP, we are unable to use Web…
2 votes -
Sequential Metrics for User Journey Analysis
Dear Team,
We would like to propose an extension to the recently released ratio metrics functionality through the introduction of sequential metrics capability.
This feature would enable the combination of two metrics based on their occurrence sequence for individual users. The system should accommodate both flexible ordering (where metrics may occur in any sequence) and prescribed ordering (where a specific sequence is required).
One might aptly term this functionality "sequential metrics."
This capability would prove invaluable for analysing the impact of experimental changes on subsequent user actions. For instance, when testing a new button design, we could examine users who…
2 votes -
Ratio Metrics for Experiment-Specific Metrics
Dear Team,
We would like to request the extension of ratio metric functionality to include experiment-specific metrics, rather than limiting this capability to project-wide metrics alone.
Currently, ratio metrics are only available when combining project-wide metrics, which restricts our analytical capabilities at the experiment level. This limitation becomes particularly evident when examining conversion funnels and user behaviour patterns within individual experiments.
To illustrate this requirement: consider a microcopy test conducted on a button element. We can readily track the click-through rate using an experiment-specific metric, whilst simultaneously monitoring checkout initiations through a separate project-wide metric. From an experimenter's perspective, understanding…
2 votes -
Preview Experiment Results in CMP via Collaboration
I am the product owner of the web experimentation platform. We also use the CMP. I connected the two platforms so that we could collaborate on variations, view metrics & run time within the CMP, however, results are not viewable within the CMP.
This defeats a large purpose of the connection between the two tools, because I want users within the CMP to see the results within our projects, without having to go into the Experimentation platform. This will save me time manually reporting on experiments, and would enable anyone who is part of that CMP Hypothesis to see the…
3 votes -
search for
Feedback from client: I want to be able to search for an old test on the results dashabord just once, rather than have to search for it within each status category.
2 votes -
Tool should flag when a test is low power (ie not likely to reach stat sig)
State exactly why it was flagged; would likely require customer to input some data
3 votes -
Bring the Sample Rate Calculator into Test Build
Its inconvenient navigating back and forth
3 votes -
Campaign Export Capability
I am working as analytics engineer in western union , we are having trouble to export the campaigns for our audit activity.
We have 100s of running campaigns in our prod project and its not feasible for all our teams to use the API end points to list more than 200 campaigns and filter Running campaigns manually from the JSON.
And our non tech team cannot do this activity and we don't have bandwidth to build application using the Optimizely api end point to get a UI for this.
We used other AB & Personalization tool and has this basic…
2 votes -
1 vote
-
Do not remove code when changing activation targets.
Frequently I need to switch my coding / set up from a staging to a production environment. Optimizely warns me and then REMOVES all my code.
This is fine if you are a basic non-technical user.
It would be better if this was an optional change.
Possibly provide a check box and a message in the button (see screenshot) - which would allow me to change the targeting.
1 vote -
Page Properties for Enhanced Targeting and Metrics
Dear Team,
We would like to request the extension of the recently released properties functionality for custom events to include page-level implementation. This enhancement would significantly improve both targeting capabilities and metric definition for page-based analytics.
Currently, when managing numerous product detail pages (PDPs) with varying attributes such as product categories, we face limitations in creating meaningful metrics. Whilst page attributes exist, they cannot be utilised within metric definitions.
The ability to define page properties would enable us to create metrics such as "product category PDP visits" without requiring dedicated category pages or specific experimental targeting for each variant.
This…
1 vote -
Multiple audience preview
I would like to be able to test multiple audiences on the same page, to see how the activities associated with those audiences would appear to a user in multiple audiences.
The Optimizely preview helper would be a great way to do this. In the override, it currently only shows the audiences associated with the campaign being previewed. If this allowed for the override to allow for any audiences to be selected, and hence allow to test for how it might look for a user who is in multiple audiences, it would be extremely helpful.
2 votes -
Export code based on conclusion
Once a test is concluded, export the changes as a set of code in HTML+CSS+JS, or React, etc. Has come up several times in the past month. Suggest investigating whether competitors do this.
2 votes -
Guided Solutioning for Common WYSIWYG problems
There are common WYSIWYG editor problems, based on site architecture or test type. Create self-service solutioning guides for customers to diagnose and solve problem.
2 votes -
Improved QA plug-in
Customers use the Chrome Plug-in often for debugging, QA and forcing themselves into variations. It does not always work, and is pretty limited.
2 votes -
Custom Labels (Owners, Next Steps, Jira Link, Figma Link, etc.)
Give customers the ability to create own labels to attach to experiments, and make these columns you can add in the overview, as well as making it filterable and sortable. This feature would help create ownership of experiments
2 votes -
URL conditions for saved pages
I create experiments for a major retailer, and current URL substring match capabilities force me to use JS conditions or callbacks for all campaigns on my PDPs and category pages. This is tedious and makes it difficult for non-technical teammates to run experiments.
For example, many retailers use structures like retailer.com/p/ to indicate PDPs and retailer.com/c/ for category pages. However, using those in the substring match fields will include all substrings with a "p" or "c" since the backslashes are not accounted for, opening up the test for many, many more visitors than we would anticipate. For example, the visitor…
1 vote
- Don't see your idea?