Most "best AB testing tools" lists are written for buyers in the US or Europe. If you're running experiments from Brazil, the ranking changes. Sometimes dramatically.

A tool that is excellent on paper can be painful in practice when the invoice arrives in dollars, support answers at 3 a.m. Brasília time, and nobody on your team has ever seen a case study involving boleto, PIX, or 12x installment checkout.

This guide evaluates the platforms Brazilian teams actually use, scored on four things that matter locally: pricing and currency, support (language and timezone), local customer base, and community.

Something happened in 2025

First, some 2025-2026 context you need before you sign anything.

The experimentation market consolidated hard over the last 18 months, and it directly affects what you're buying:

1. VWO and AB Tasty merged

   On 20 January 2026, the two vendors announced a merger, creating a platform with over $100M in annual revenue and more than 4,000 customers, operating from 11 offices across North America, LATAM, Europe, and APAC. Around 90% of the combined revenue comes from the US and Europe.

2. Statsig changed hands twice

   OpenAI acquired it in September 2025 for roughly $1.1B. In May 2026, Amplitude took over the Statsig brand and customer base, while the original engineering team stayed at OpenAI.

3. Datadog acquired Eppo; Webflow acquired Intellimize

   The independent mid-market layer is thinning out.

4. Before that

   Google Optimize has been gone since 2023, and VTEX deprecated its native AB Tester the same year, which is why so many Brazilian e-commerce teams have been re-tooling.

The practical takeaway: pricing and roadmaps are drifting upmarket. If you're a mid-sized Brazilian company, you have fewer cheap, stable, self-serve options than you did two years ago.

The Brazil filter: what actually differs

Easy, right?

• Currency

  Almost every global vendor prices in USD. On top of the exchange rate, a Brazilian company paying by international card absorbs 3.5% IOF, and remitting payment abroad for an imported service adds further withholding and contribution taxes plus paperwork. A USD 500/month tool is rarely R$ 2,700/month by the time finance is done with it. Vendors that contract in BRL and issue a nota fiscal remove an entire category of friction, making the expense predictable when the real moves.

• Support

  Brasília is UTC−3. US vendors overlap comfortably. European vendors give you a half day. India- and APAC-based support means your production incident sits in a queue overnight. Layer language on top: English-only support is workable for engineers and brutal for the marketing analyst who actually configures most tests.

• Local customer base

  Reference cases matter more than feature lists. Does the vendor understand VTEX, Shopify Brazil, Nuvemshop, RD Station, PIX and installment checkout? Can they show you a Brazilian retailer, bank or marketplace, or only European fashion brands?

• Community

  Tooling is a small part of experimentation maturity. In Brazil, the ecosystem now includes the Comunidade CRO Brasil, specialist agencies such as Supersonic, Conversion and Looptimize, and mature in-house programs. A platform with local partners and local practitioners is easier to hire for.

1. Croct: the Brazilian option

The only serious experimentation platform built in Brazil. It combines AB testing with real-time personalization, segmentation, feature flags, and a component-based CMS aimed at headless stacks, meaning content, tests, and personalization live in one place instead of three tools.

Pros

• Pricing: free forever up to 10k monthly active users (3 content slots, 1 experiment), Growth plan starts at $600/month billed annually for 20k MAU. Plans can be contracted in reais, which removes IOF and FX exposure. It also has startup and agency programs.
• Support: local team in Brazil, in Portuguese, English, and Spanish, in your timezone. Support via email, Slack, and Google Chat, with an account manager and onboarding check-ins regardless of plan size.
• Local customers: genuinely Brazilian references (see list here), plus published guidance on running tests on VTEX after the native tester was deprecated.
• Community: open Slack community available even on the free plan, plus active Portuguese-language content and playbooks.

Cons

• Smallest vendor on this list, which makes them less battle-tested at extreme scale than Optimizely or the VWO/AB Tasty entity.
• Reviewers on Capterra praise the low developer dependency but flag limits on the number of variants per test and want richer metrics in the dashboard.
• Strongest fit for modern JS and PHP frameworks (heavier legacy stacks require more integration work).
• Mobile app testing is not available yet.

Best for: Brazilian product and growth teams on React and Vue ecosystems who want personalization and testing together, in reais, with quick support that speaks their language.

2. VWO: now VWO + AB Tasty

The most commonly recognized name among Brazilian marketing teams, and the one with the deepest all-in-one suite: testing, heatmaps, session recordings, surveys, personalization, and server-side testing.

Pros

• Pricing: more transparent than most enterprise vendors at the entry point, a published Starter tier around $314/month billed annually for 10k monthly tracked users. Vendr's aggregated buyer data puts the median annual contract near $16.7K.
• Support: widely praised in reviews. The merged entity explicitly claims LATAM offices and "global reach with local execution."
• Local customers: real Brazilian adoption, and VWO has published content featuring Brazilian CRO practitioners.
• Community: large global user base, extensive documentation, one of the most-read CRO blogs in the industry.

Cons

• Pricing in USD, and pricing above Starter is quote-only. Buyers report costs scaling fast with monthly tracked users, and the free tier has been wound down in favor of a 30-day trial.
• The merger points upmarket. With ~90% of combined revenue from North America and Europe, Brazilian mid-market accounts are unlikely to drive the roadmap.
• Post-merger overlap between VWO and AB Tasty products creates real uncertainty about which modules survive long-term.
• Support is global-hours, not Brazil-native, so Portuguese is not guaranteed.

Best for: established Brazilian e-commerce and media companies that want testing plus behavioral analytics under one contract and have budget in dollars.

3. Convert Experiences

The pragmatic mid-market choice: enterprise-grade testing mechanics, flicker-free execution, SRM protection, solid stats, with published prices and no mandatory sales call.

Pros

• Pricing: genuinely transparent. Growth around $299/month for 100k tested users, Pro around $499/month billed annually for 250k, roughly $299 per additional 100k. For a Brazilian company, predictable published pricing is worth a lot when you're forecasting in reais against a volatile dollar.
• Support: consistently the most-praised aspect in reviews, real humans, fast, via live chat and email, with explicitly agency-friendly support. US/EU hours work reasonably for Brazil.
• Local customers: heavily used by CRO agencies, including in Brazil, which means local implementation help is available even though Convert itself has no Brazilian office.
• Community: strong privacy posture (ISO 27001 and SOC 2), 90+ integrations, and an unusually good educational blog: their coverage of the 2026 consolidation is some of the clearest published anywhere.

Cons

• USD only, no BRL invoicing, no Brazilian entity.
• No Portuguese support.
• Some reviewers find the tool almost too simple (personalization is lighter than VWO or Croct).
• No local case studies to show a Brazilian board.

Best for: mid-market Brazilian teams and agencies that want serious testing mechanics at a knowable price and can operate in English.

4. Optimizely

Still the reference for large-scale, statistically rigorous experimentation: Stats Accelerator, multi-armed bandits, SRM detection, deep enterprise compliance, and a full DXP around it.

Pros

• Best-in-class statistics engine and governance for organizations running high volumes of concurrent tests.
• Strong LATAM enterprise presence through partners and system integrators, which in Brazil usually means Portuguese-speaking implementation partners even when the vendor isn't local.
• Scales to the largest Brazilian retailers, banks and marketplaces without hitting ceilings.

Cons

• Cost. No published pricing. Vendr data across 109 buyers puts the median annual contract near $78,000, ranging roughly $31.5K to $199K. Web Experimentation standalone typically lands in the $40K–$120K/year band. In reais, that is a board-level decision, not a team-level one.
• Annual contracts with auto-renewal and long notice periods. No self-serve trial.
• Modular SKUs mean the "Optimizely" quote you receive is often two or three products bundled.
• Support is enterprise-tier and English-first. Portuguese depends on your account team and partner.

Best for: large Brazilian enterprises with a dedicated experimentation function and budget measured in hundreds of thousands of reais per year.

5. GrowthBook

Open-source, warehouse-native experimentation and feature flagging. Your experiment data stays in your own BigQuery, Snowflake, Redshift, or Databricks, so GrowthBook reads aggregates.

Pros

• Pricing is the headline for Brazil: the self-hosted open-source edition is free with unlimited users. Cloud Starter is free for up to 3 users with unlimited flags, experiments, and traffic. Pro is $40/seat/month up to 50 users. Because it's seat-based, not traffic-based, a high-traffic Brazilian marketplace doesn't get punished for volume (a meaningful difference against MTU pricing when your revenue is in reais).
• Self-hosting inside your own infrastructure simplifies LGPD conversations, since visitor-level data never leaves your environment.
• Serious stats: Bayesian and frequentist engines, CUPED, sequential testing, bandits, SRM checks. 24+ SDKs.
• Active open-source community (Slack, GitHub), used by 3,000+ companies.

Cons

• No Portuguese support and no Brazilian presence. Free tiers get community support only.
• Requires a data warehouse and engineering capacity, which is not a tool you hand to a marketing analyst on day one. Self-hosting means you own uptime.
• The visual editor is a paid-tier feature, and marketer-led testing is not the primary use case.
• Billing still in USD for paid tiers.

Best for: engineering-led Brazilian product teams that already have a warehouse and want cost control at scale.

6. Statsig: now under Amplitude

Feature flags, experimentation, and product analytics in one developer-friendly platform, with a generous free tier and a Pro tier around $150/month.

Pros

• Excellent value below enterprise scale, unlimited seats, strong sequential testing, warehouse-native option.
• Broad mobile SDK coverage: relevant given how mobile-first Brazilian consumer products are.
• Free tier is large enough for most early-stage Brazilian startups.

Cons

• Continuity risk is the dominant con right now. OpenAI bought Statsig, then Amplitude took over the brand and customer base in May 2026, while the team that built it stayed at OpenAI. Optimizely's CEO publicly compared it to acquiring "a race car without a driver." Amplitude also now has overlapping experimentation and analytics products, raising the odds one gets consolidated.
• No Brazilian entity, no Portuguese support, USD billing.
• Roadmap and pricing stability are open questions for at least the next year.

Best for: technical Brazilian teams already invested in it, but I'd think carefully before signing a new multi-year contract until the Amplitude transition settles.

Quick comparison

Prices verified July 2026 and change frequently.

How to choose

Mid-sized company, growth-led, no dedicated dev. Croct if you want reais, Portuguese, and personalization bundled in. Convert if you're comfortable in English and want the cheapest credible pure testing tool with published pricing.

Scale-up with a real product and engineering team. GrowthBook, especially if you already have a warehouse. The seat-based model protects you from traffic-based bill shock, and self-hosting simplifies LGPD.

Large enterprise with a formal experimentation function. Optimizely, or the merged VWO/AB Tasty entity. Negotiate hard: multi-year terms commonly unlock 20–30% off list, and VWO's fiscal year ends in March, so Q1 deals tend to price better.

Migrating off VTEX's deprecated native tester. Croct has the most direct published path. Otherwise, Convert or VWO with a competent local agency.

Now, three closing warnings.

First, budget the real cost, not the sticker price: USD list + FX + 3.5% IOF + import-service taxes. Second, in a consolidating market, ask every vendor directly what happens to your contract and roadmap if they're acquired, since three of the six platforms here changed ownership in the last 18 months. Third, the tool is the easy part. The Brazilian companies winning at experimentation got there through process and culture. Buy the platform your team will actually use every week, not the one with the longest feature list.