Optimizely Web Experimentation

Score8.7 out of 10

599 Reviews and Ratings

What is Optimizely Web Experimentation?

Whether launching a first test or scaling a sophisticated experimentation program, Optimizely Web Experimentation aims to deliver the insights needed to craft high-performing digital experiences that drive engagement, increase conversions, and accelerate growth.

Categories & Use Cases

Media

AI-Powered Experimentation with Opal:

- Instant Test Ideas: Generates high-quality A/B test ideas based on any goals and audience insights.
- Smarter Experimentation: The AI can suggest impactful variations, reducing guesswork and increasing test velocity.
- More Than Just Ideas: From hypothesis generation to analyzing results, Opal helps optimize every stage of the experimentation process.

Advanced Audience Targeting:

- Delivers personalized experiences by targeting users based on behaviors, attributes, and real-time conditions.
- Defines precise audience segments using first-party data, geolocation, and device type.
- Can test and optimize for different audience groups to maximize impact and engagement.

Web Experimentation Results Page:

- Data visualizations help interpret experiment performance.
- Displays which variations are winning with built-in statistical significance calculations.
- Results can be filtered by audience segments, events, and conversions to uncover key trends.

the custom code editor for JavaScript & CSS changes, used to build complex tests.

1 / 6

Heatmap tool
A tool that shows which elements of the page generate the most visitor engagement.
Category average: 8.2
Traffic allocation control
Ability to set what percentage of website traffic receives specific test variants in order to roll out code only to a subset of site visitors.
Category average: 9
a/b experiment testing
Create and test variations of a website, changing site elements such as headlines, CTAs, images, page design and layout, technical SEO changes, and new feature additions and collect statistical results of each variation’s conversion rates or other metrics.
Category average: 8.9

Dynamic experiment activation
Ability to activate an experiment after the page’s initial load based on a set of conditions (e.g. if the visitor takes certain actions).
Category average: 8
Server-side tests
Ability to run server-side tests (e.g. A/B, A/B/n, multivariate, and split URL tests) to test out more complex design changes, roll out features to specific audience segments, or split site traffic between different site versions.
Category average: 7.5
Page surveys
Create on-page surveys and select which segment of users are asked survey questions using defined audience segments (e.g. new vs. returning users, mobile users, desktop users, etc).
Category average: 7.2

Anthony Scott

Web Developer in Information Technology at Pixel Syncnet (501-1000 employees employees)

Use Cases and Deployment Scope

I'm using Optimizely web experimentation to make our websites perform measurably better rather than relying on guesswork. It helps me to prove value quickly. I often set up custom event tracking, like monitoring engagement with embeded calculators or interactive product filters so we can measure deeper behaviors, not just clicks

Pros

Setting up multivariate tests directly in the platform, wiring it up to capture drop off events through the analytics pipeline

Cons

I'm currently trying more complex setups like syncing experiment audiences with our data warehouse. It is not always plug and play, sometimes I have to write custom javascripts to track very specific behaviors.

Return on Investment

Is shortened our experiment ideation to launch time from 4 weeks to about 8 days, freeing our dev team to take on more billable client projects

Other Software Used

TeamViewer, IBM Maximo Application Suite, Figma Dev Mode

Verified User

Engineer in Information Technology (201-500 employees employees)

Use Cases and Deployment Scope

Optimizely Web experimentation is the primary tool I use to validate front-end and workflow changes. But beyond the day-to-day, I've leaned on it for more ambitious projects like an ongoing experiment with AI-driven diagnostic recommendations. I'm coordinating with our analytics team to pipe experiment data into Power BI. We've also partnered with Kin + Carta to implement a workaround that links Optimizely events with some of our older reporting systems.

Pros

Fast iteration on tools with real time reporting.
Safe rollouts on high-risk features.

Cons

I attempted to test a multi-step referral process across both desktop and mobile, but I couldn't complete it without involving Optimizely experts, as the setup requires a significant amount of custom code.

Return on Investment

Experiments that increase completion rates directly reduce support tickets saves us money in the tens of thousands.
Biggest ROI comes from avoiding sinking dev resources since we can just test first.

Other Software Used

IBM Cloud Continuous Delivery, Figma Dev Mode

Verified User

Manager in Marketing (10,001+ employees employees)

Use Cases and Deployment Scope

We do a lot of data analysis on our e-commerce journeys and one of our biggest, I guess, goals is conversion rate. So we use Optimizely Web to understand the insight that we're getting through either focus groups or other experiments, whether we should roll out a particular change or whether we should add additional information. And whether that basically enhances the journey enough to give us a better conversion uplift.

Pros

I think the best aspect of it is because I also manage a team that builds agent experiments, which are a little more complicated. They involve a lot of complex logic and conditions and really focuses in on certain audiences. So when we look at web experiments, the best benefits are getting things off the ground within a matter of minutes. Whereas agent experiments, there's a lot of background build involved with web experiments, we can have an idea, we can build it in web and it can be launched the same day. So it really helps us get to answers faster and make those decisions faster and then lead to other ideas for things we can do on other parts of the website.

Cons

I think because I work with both types of experiments, web and agent experiments, the sort of drawbacks come with web that it's the audience logic. Sometimes we have to identify specific customers that we want to target with the change. And a lot of it is probably down to our infrastructure on the site. It's not giving Optimizely the right level of data to target these customers. But I think, yeah, if we had a little bit more understanding of how we could get to that data through optimizing web, that would be useful for us.

Return on Investment

I can't probably speak to exact numbers because we're not really allowed to share, but it's in the millions we see conversion rate uplift and obviously that is our main metric. But also in terms of experiments we've done to try to increase our POO, we've seen definite success in being able to use web to validate some of our hypotheses. And like I said, the win rate is so strong that there is no, I guess question on the turn on investment.

Verified User

Employee in Engineering (201-500 employees employees)

Use Cases and Deployment Scope

In our organization, I use Optimizely Web Experimentation for A/B Testing. Optimizely Web Experimentation helps us test things prior to shipping.

Pros

A/B Testing
Helps narrow down Variants

Cons

I think there is room for improvement in Optimizely Web Experimentation with Better Native OS Apps

Return on Investment

The impact Optimizely Web Experimentation has on our organization's overall business objectives is Helping picking variants

Verified User

Manager in Marketing (201-500 employees employees)

Use Cases and Deployment Scope

We use Optimizely Web Experimentation across our websites to optimise our customer journeys, helping to drive more donors, volunteers, campaigners, service users and more. Experimentation makes it easy to create consistent experiences across multiple pages.

Pros

Robust reporting engine
Easy to use UI
Flexible experiment triggers and audience rules

Cons

Using platform can be tricky for non technical users. Particularly with selecting elements.
Limited tools for exporting reporting
Reports are retired after a certain time making it unreliable as a long-term experimention repository