# A/B testing tools: Optimizely, VWO, and how to choose

> Optimizely and VWO (now Wingify) are experimentation platforms with custom pricing. Compare what they offer, how they charge, and how to choose between them.

*Tags: ux, tool, experimentation, growth, analytics*

---


> [!info] Quick Definition
> **A/B testing tools** show variants of a page or feature to different groups of users and measure which one performs better. **Optimizely** and **VWO** are two of the best-known platforms in the sector, and both combine experiments, personalization, and feature management.


## What are these tools?

They are platforms that assign each visitor a variant, record their actions, and calculate whether the difference is statistically reliable. They go beyond classic A/B testing: they can test several pages at once, run multivariate tests, personalize content, or roll out features gradually.

According to its pricing page, Optimizely groups A/B, multipage, feature, and server-side testing under "Agentic Experimentation," together with analytics and personalization, and sells feature management separately. VWO, which now appears under the Wingify brand, offers experimentation, personalization, analytics, and feature management suites.

Some context: Google Optimize, Google's free tool, stopped being available on September 30, 2023. Google names AB Tasty, Optimizely, and VWO as providers it works with for A/B testing with Google Analytics. For simple tests you can also build your own experiment structure.

## Why does it matter to a designer?

Because a well-designed A/B test is a way to answer design questions with data. If two versions of a checkout button compete, the tool tells you which converts better. But the tool does not decide the question: you do, by framing a clear hypothesis.

It also helps to know the limits. A test shows what happens, not always why. That is why [triangulation](/en/wiki/research/triangulation/) with interviews or usability tests stays valuable when a result surprises you.

## How are they bought?

Neither Optimizely nor Wingify (VWO) publishes list prices. Optimizely's page invites you to "build a custom quote" from your goals and requirements. Wingify's page says: "No fixed plans. Just a custom proposal built around your goals and scope." It shows Growth, Pro, and Enterprise levels for each product, without figures, and explains that price depends on monthly active users (MAUs); behavior analytics uses monthly tracked users (MTUs).

Wingify also offers a 30-day trial with no credit card ("Free for 30 days. No credit card required"), which is different from a permanent free plan. Do not use figures from third-party blogs as the price: they are not official and they change. Always ask for a quote.

One more fact changes the conversation: according to the press release of September 16, 2026, VWO and AB Tasty came together under the Wingify brand, with more than 4,000 customers and a unified platform. The release does not say whether the VWO and AB Tasty names will continue as products. Before signing, confirm with the vendor which product, terms, and contract apply.

## What criteria should you use to choose?

Rather than comparing brands, answer a few questions:

| Criterion | Key question |
|---|---|
| Type of test | Are you testing web pages, a mobile app, or server-side product features? |
| Who uses it | Will designers and marketing use it, or engineering? |
| Measurement | Do you need reliable statistics or only trends? |
| Integration | Does it connect to your analytics and data warehouse? |
| Price | Is it charged per visitor, per monthly active user (MAU), or per event? |
| Velocity | How many experiments do you want running at once? |

A practical point: if your team has few visitors, most tests will take too long to reach a reliable result. In that case, improving qualitative research first is often better than investing in formal experimentation.

If you decide to try a platform, use the free trial on a real, narrow case:

1. Write the hypothesis and the metric before configuring anything.
2. Calculate how many visitors you need and how many weeks the test would take.
3. Install the tool on a single page and check that it measures the same thing as your analytics.
4. Assess how long it takes your team to launch a test without asking engineering for help.

## When is a platform the wrong choice?

A paid tool is not always necessary. For a simple test you can measure the difference with your analytics and a homemade experiment structure. A platform is justified when you need consistent user assignment, reliable statistics, several experiments in parallel, or a way to switch features on without deploying code.

Nor should you run tests out of habit. A test without a clear hypothesis produces data without a decision.

## Example

Illustrative example: an online store wants to test whether showing the final price including taxes on the product page increases purchases. It has 40,000 monthly visitors and a hypothesis: fewer surprises in the cart would reduce abandonment. Before buying a tool, the team calculates how long a test with that traffic would take to detect a two-point improvement, and finds it would take several weeks. They decide to start with a simple test in their analytics, and only evaluate a platform once they have several experiments queued.

## Common mistakes

- **Testing without a hypothesis** and declaring a winner at the first difference.
- **Stopping a test as soon as a result looks good**, which inflates false positives.
- **Buying the tool before having enough traffic** for the tests you want to run.
- **Using third-party figures as the price** without requesting a quote.
- **Forgetting qualitative research** when a result cannot be explained.
- **Running many tests at once** without coordination across teams.

If an A/B result leads you to ask why it happened, [usability testing](/en/wiki/research/usability-testing/) helps find the cause. For the theory behind experiments, the [A/B testing](/en/wiki/research/ab-testing/) page explains how to plan them. And if you combine behavioral data with observation, heatmap and session recording tools such as [[hotjar|Hotjar]], alongside general analytics such as [[Google Analytics]], make a good complement.

## Sources

- [Optimizely, pricing](https://www.optimizely.com/pricing/)
- [Wingify (VWO), pricing](https://wingify.com/pricing/)
- [Wingify (VWO), home page](https://wingify.com/)
- [AB Tasty and VWO Unite Under Wingify, press release of September 16, 2026, via CMSWire](https://www.cmswire.com/the-wire/ab-tasty-and-vwo-unite-under-wingify-launching-a-unified-platform-new-brand-identity-and-a-w/)
- [Google Optimize sunset, Google Analytics Help](https://support.google.com/analytics/answer/12979939?hl=en)


---

Source: https://www.fernandoux.com/en/wiki/tools/ab-testing-tools/
