- Switching is a rebuild, not a migration: nothing has to be exported from the old tool except your variation code, targeting rules and old reports.
- Plain JavaScript and CSS variations can be pasted into Varify’s code fields. Code that uses helpers of the old tool needs a small adaptation.
- Your numbers follow your analytics: connect GA4, BigQuery, PostHog, Matomo or Piwik Pro, or use Varify Realtime Tracking.
- Validate with an A/A test first, finish running experiments in the old tool, then retire its snippet.
VWO and AB Tasty now operate under one brand, Wingify. If you are rethinking your testing tool, the first question is usually not price or features. It is: how painful is the move, and what happens to the tests we built?
This guide answers that for VWO, AB Tasty and Wingify alike. The steps are the same, because the work you carry over is the same: your variation code, your targeting and your goals. For the background on the rebrand, read our article on the merger. For a feature comparison, see Varify.io vs. Wingify.
How hard is it to switch from VWO or Wingify to Varify?
It is five steps, and none of them imports data into Varify. You install one snippet, connect the analytics you already use, rebuild your experiments, rebuild targeting and validate. Most of the effort sits in step 3, and that step is copying and checking code, not engineering something new.
The five steps in detail are below. Skim the table first.
| Step | What you do | What you carry over |
|---|---|---|
| 1. Snippet | Add the Varify snippet to the <head>, or through your tag manager | Nothing |
| 2. Tracking | Run the Tracking Setup Wizard and connect your analytics | Your goals, as events in your analytics |
| 3. Experiments | Rebuild each test with AI, in the visual editor or with code | JavaScript and CSS from your variations |
| 4. Targeting | Recreate page and audience rules | URL patterns, selectors, conditions |
| 5. Validate | Run an A/A test, check the event, switch over | Nothing |
What are the five steps, one by one?
1. Install the Varify snippet
Place the snippet as high in the <head> as you can, so the page does not flicker. A tag manager works too, but it carries a risk of visible flicker while a variation loads. If you cannot change the page yet, the Varify Chrome extension runs the snippet locally in your browser, so you can build and preview before anything goes live.
2. Connect your analytics
Varify separates delivering the variation from measuring the result. The Tracking Setup Wizard connects the analytics tool you already run: Google Analytics 4, BigQuery, PostHog, Matomo or Piwik Pro. Their events then show up as selectable goals. If you prefer a dedicated Varify tracking service, Varify Realtime Tracking is the alternative.
3. Rebuild your experiments
You have three routes, and you can mix them per test:
- Describe it to the AI. With Prompt-Based Experimentation you describe the change and the AI builds the variation. What your plan includes is on the plans page.
- Use the visual editor. Simple changes such as text, images and layout are quick to rebuild there.
- Paste the code. Copy the JavaScript and CSS from the variation in your current tool into Varify’s JavaScript and CSS fields.
4. Recreate targeting
Page targeting works with URL rules, regular expressions, CSS selectors or JavaScript conditions. Audience targeting uses the Segment Builder, with JavaScript for complex conditions.
5. Validate, then switch over
Start an A/A test: it assigns two variations without any visible change. Check in your analytics that the event arrives with its parameters. Then let running experiments finish in the old tool, start the rebuilt ones in Varify and remove the old snippet.
Can I reuse my JavaScript and CSS variation code?
Yes, in most cases. Varify’s code fields accept plain JavaScript and CSS, so code that selects an element and changes it moves over as it is. Three things to check before you paste:
- Helpers of the old tool. Code that calls functions the old platform injects needs a rewrite to plain JavaScript or to Varify’s helpers.
- Elements that load late.
window.varify.helpers.waitForwaits for an element with a mutation observer and runs your callback when it appears. This is the usual fix for AJAX content and single-page apps. - Single-page apps. Use the JavaScript Reset so a change does not stay visible after a page switch without a reload.
What happens to running experiments and old reports?
A running experiment does not move between tools. Let it finish in the old tool, or restart it in Varify when it makes sense. Reports and learnings stay where they were created, so export them before your contract ends. Check the notice period and do not cancel before your last test is concluded.
Will my numbers match after the switch?
Expect your results to follow your analytics, not the counts of your old tool. Varify delivers the variation, and your analytics tool, or Varify Realtime Tracking if you choose it, measures the outcome. Two different measurement systems rarely agree to the last visitor.
Compare both on one shared period before you judge a result. The A/A test from step 5 shows whether your setup counts what you think it counts. With BigQuery you can inspect the raw data yourself.
What about server-side tests, heatmaps and shop systems?
- Server-side. Varify covers client-side and server-side testing. Server-side means calling the assignment API from your server code and rendering the variation into the first HTML.
- Heatmaps and recordings. Varify focuses on experimentation. For qualitative data, connect Mouseflow, Microsoft Clarity, Hotjar or Contentsquare.
- Shop systems. Shopify is supported, including goal import for revenue, purchases and average order value. Shopware, Magento, Commercetools, WordPress and single-page apps work with the snippet.
If your current setup depends on features in the suite that Varify does not include, plan to replace those with dedicated tools before you cancel.
See your own test run in Varify.
Take one experiment you run today and rebuild it in the 15-day trial. Full feature access, no credit card.
