Google Optimize shut down on September 30, 2023, and Google did not ship a direct replacement. Teams that relied on it moved in three directions: dedicated testing platforms like VWO and Optimizely, experiment features built into analytics such as GA4 integrations, and audit-and-generate tools that decide what to change and write the variants for you. There is no single "new Google Optimize," so the right pick depends on your traffic and what part of the job you actually got stuck on.
This guide covers what happened, what people replaced it with, how to A/B test without it, and how to choose between a free and a paid path. Last updated July 2026.
What happened to Google Optimize?
Google retired both Optimize and Optimize 360 on September 30, 2023. The stated reason was that Optimize lacked features customers needed and that Google would instead invest in third-party A/B testing integrations inside Google Analytics 4. Optimize had been free, which is a big part of why it was everywhere, so its removal left a gap that most paid tools have been happy to fill. Any experiments still running when it closed stopped, and the historical data became inaccessible after the shutdown.
What replaced Google Optimize?
Nothing replaced it one-for-one. Google pointed users toward its analytics partners, and the market split the job Optimize used to do into a few categories. Here is the honest lay of the land in 2026.
| Option | Best for | Cost shape |
|---|---|---|
| VWO | Full testing plus heatmaps and recordings in one suite | Paid, scales with traffic |
| Optimizely | Enterprise experimentation and feature flags | Enterprise annual |
| GA4 + partner integrations | Staying inside Google Analytics for measurement | Varies by partner |
| Microsoft Clarity | Free heatmaps and session recordings (not testing) | Free |
| Converto | Deciding what to change and generating the variants | From $19/mo |
Notice these are not all the same product. Testers run the experiment. Behavior tools show you what visitors do. And the part Optimize never really solved, figuring out what the better version of the page should say, is a separate step most people did by hand. That is worth keeping in mind before you pay for a like-for-like tester and find yourself staring at a blank variation editor.
How do I A/B test without Google Optimize?
The mechanics are the same as they always were; only the tool changed. You need three things: a testing engine to split traffic and measure results, enough traffic to reach significance, and a variation worth testing. Most teams underweight that third piece.
- Choose a testing engine. VWO and Optimizely are the common paid picks. If you live in Google Analytics, use a GA4-integrated partner so results land where you already report.
- Check you have the traffic. A test needs enough conversions to separate signal from noise. If a page gets a few hundred visits a month, a formal split test will take a long time to call, and you are better off shipping a well-reasoned improvement and watching the trend.
- Decide what to test. This is the hard half. A test only pays off if the variation is a real, defensible improvement, not a random button color. Ground the change in a page audit so you are testing the leak that actually costs you conversions.
Because reaching significance depends on volume, some teams keep tests moving by putting paid traffic behind the page while a test runs. If that is you, letting an automated media buyer manage the paid campaigns keeps the traffic steady without adding another manual job on top of the experiment.
Is there a free Google Optimize alternative?
For testing specifically, there is no free tool that matches what Optimize offered; the capable testers are paid. The nearest free path is a developer-run one: open-source experimentation suites like GrowthBook, or a PostHog-style analytics and experimentation stack with a generous free tier, will run experiments, but they assume engineering time to install and instrument. For understanding behavior, Microsoft Clarity is genuinely free and gives you heatmaps and session recordings, and Hotjar keeps a limited free plan. What none of the free tools do is tell you what to change or write the fix, so a "free" path usually means free observation plus your own time turning that into a better page.
If your goal is fewer wasted clicks rather than a formal experiment, the cheapest high-leverage move is often a structured audit of the page. Running a landing page audit surfaces the copy, layout, CTA, and speed leaks in one pass, so you fix the obvious losses before you spend weeks testing marginal ones.
Do you even need a testing tool?
Not always. A/B testing earns its keep when you have the traffic to run rigorous experiments and a backlog of ideas worth validating. Below that threshold, the returns come from removing clear problems, a headline that does not match the ad, a buried CTA, a slow first screen, rather than from proving a 3 percent lift with statistical confidence. Many small teams that lost Optimize discovered they had been using it to A/B test their way toward fixes they could have just made.
If that describes you, an audit-and-generate approach covers more ground for less money. Converto is built for exactly the decide-what-to-change step Optimize left to you: paste the live page you already have and it audits copy, layout, CTAs, page speed, and AI-readability, then generates ready-to-ship variants, with nothing to install. If you later want to validate a change with a formal test, you go in with a concrete variation instead of a guess. We compare the options in more depth on our Google Optimize alternative page.
How to choose your replacement
Match the tool to the job and the traffic, not to the brand you lost.
- High traffic, active testing program: a dedicated tester like VWO or Optimizely is worth it.
- You report in Google Analytics: a GA4-integrated partner keeps measurement in one place.
- You want to see behavior: Microsoft Clarity is free; Crazy Egg and Hotjar are paid options.
- You are stuck on what to change: an audit-and-generate tool like Converto gets you a fix without a snippet or a testing program.
The mistake to avoid is treating this as a straight swap. Optimize was free and general, and the market that replaced it is specialized. Decide which part of the loop you need help with, find the leaks, design the fix, or prove the lift, and buy for that. If your honest answer is "I need to know what to fix," you do not need to rebuild a testing stack to get there.
The short version
- Google Optimize closed on September 30, 2023, with no direct replacement.
- Testing moved to VWO, Optimizely, and GA4-integrated partners.
- Free behavior tools (Microsoft Clarity, Hotjar's free tier) show you what visitors do, not what to change.
- The hardest part of testing is the variation; ground it in a page audit.
- If you mainly need to know what to fix, an audit-and-generate tool covers more than a bare tester.
Paste your page and see what is leaking
Drop in any live URL and Converto audits the copy, layout, CTAs, speed, and AI-readability of the page you already have, then hands you ready-to-ship variants to lift signups and sales. Suggestions are data-informed, not guarantees.