Google Optimize Alternative: What to Use Since It Shut Down

The Google Optimize alternative you pick depends on whether you want a free visual editor, a cheap paid one, an enterprise suite or a code-based tool. Google Optimize ended on September 30, 2023 with no direct successor, so most former users moved to VWO, Convert Experiences, Optimizely, AB Tasty or something repository-based.

What happened to Google Optimize

According to BusinessMCP’s Google Optimize comparison, Google Optimize and Optimize 360 stopped working on September 30, 2023, and Google’s notice says they are no longer available. Google did not ship a replacement. It published its GA4 measurement APIs and named AB Tasty, Optimizely and VWO as integration partners. Many small sites had relied on Optimize precisely because it was free and tied to Google Analytics, so the gap felt larger than a normal vendor change.

If you are still on an old setup that references Optimize, remove leftover snippets; they no longer run experiments, and orphaned code can slow pages and confuse audits.

What you probably liked about it

Be specific about what you are replacing, because no alternative matches all of it.

• Zero cost. Nothing free now has the same scope, but see the table below for the nearest options.
• A visual editor. Click-to-edit with no code.
• Google Analytics integration. Experiment data sat next to your normal reports.
• Low setup effort. One snippet and a goal.

Key takeaway: every replacement trades something away. Decide which of “free”, “visual editor” and “inside Google Analytics” you cannot live without, and the shortlist writes itself.

Alternatives, by priority

Prices are as listed on the BusinessMCP page as of September 2026 and some are third-party estimates.

Priority	Option	Price signal	Caveat
Free, visual editor	VWO Starter	Free up to 50,000 monthly tested visitors; paid plans from around $314/mo	Some teardowns say it is being withdrawn for new accounts
Cheapest published visual editor	Convert Experiences	$299/mo annual, $399 monthly (Growth, 100k tested users)	Not free
Enterprise scale	Optimizely	Custom, roughly $40k+/yr	Priced for large programmes
Enterprise-ish	AB Tasty	Custom, roughly $15k-$150k+/yr	Merged with VWO in January 2026 per the page
Tests proposed and built for you	BusinessMCP	$199/mo Scale	No visual editor; GitHub repository required

The code-based option, and its fit

For sites that live in GitHub, BusinessMCP’s automated A/B testing is a different shape of replacement. It analyses heatmaps and session replays to propose tests, then the AI opens a pull request with the variant for human review. Analytics, replay and CRM are included in the plan.

The key differences, per the page: no visual editor, so variants are code changes reviewed in GitHub; a GitHub-hosted codebase is required, so sites on Wix, Squarespace or a Shopify theme outside git cannot run its split tests; only web A/B tests are supported, with no multivariate testing, personalization engine, server-side SDKs or feature flags; results stay in BusinessMCP rather than Google Analytics; and statistics use a two-proportion z-test with results hidden until each arm has 300 visitors and 25 conversions and the test has run 14 days.

Matching alternatives to former Optimize users

The page’s guidance maps neatly to common profiles:

• Want the free, visual-editor experience: VWO’s Starter tier if still available to you, or Convert as a low-cost paid option.
• No repository or a hosted builder site: VWO or Convert, since a code-based tool needs a pull request to run a test.
• Have a GitHub codebase and want tests proposed and built for you: BusinessMCP.
• Enterprise scale: Optimizely or AB Tasty.

If you are a solo owner of a small site, also ask the unglamorous question: does the page you want to test get enough conversions? The sample size guide explains how to check before you pay for anything, and free AB testing tools goes into free options and their limits.

Getting your data story straight

A common complaint after the shutdown was losing the tie to Google Analytics. Decide where your source of truth lives. If you want experiment results in GA4, check that the candidate tool offers a GA4 integration and how it passes variant information. If you accept results living in the testing tool, make sure you can export raw visitor and conversion counts so you can verify them. Conversions defined in two systems rarely match exactly; expect small discrepancies and document which one decides tests.

A simple migration plan

1. Remove old Optimize code from your pages.
2. Pick one page and one goal for the pilot.
3. Choose the tool using the priorities above; start with a trial or the lowest tier.
4. Compute the sample you need before launching, then run for whole weeks. How long to run an AB test shows how to turn that into dates.
5. Log every result in a document you control.

For a broader scan of the market, read our best AB testing tools overview, and confirm every price on the vendor’s own page.

Questions to ask after the shutdown

Several years on, the practical issues have changed from “what do I do now” to “did I choose well”. Check whether the tool you picked is still the right size. Teams that moved to an enterprise platform in a hurry sometimes find they use a fraction of it; teams that took a free tier may have outgrown the limits. Review the contract, usage and results annually.

Also revisit your measurement. If you relied on Optimize reading goals from Google Analytics, confirm that your replacement defines goals the same way, otherwise historical comparisons will not line up. Keep a note of when the switch happened, so later readers of your results understand the break.

Where to read the vendor’s side

BusinessMCP’s own comparison page for this decision lists the prices and differences summarised above, and its AB testing tools comparison adds Kameleoon and statistical approaches. Both are vendor documents, so confirm any price on the vendor’s own site. For a neutral framework to evaluate any of them, our AB testing tools comparison and AB testing mistakes guides are a good second read, and the overview of AI for AB testing covers automation claims that appear in newer tools.