How we make this list: we compare pricing pages, product documentation and public user reviews. Ratings come from G2, Capterra or Trustpilot, with the date we checked them.
Key takeaways
- VWO is the complete CRO suite, testing plus heatmaps, recordings and funnels in one login, with a free Starter plan for up to 50,000 monthly tracked users and Growth reported from $314 a month billed annually.
- Convert is the privacy first, published price choice for agencies, though current comparisons report its entry price inconsistently, so confirm the tier on its pricing page before budgeting.
- Zoho PageSense at about $16 to $17 a month and Crazy Egg at $29 a month are the honest small site tier, where test volume is low and big simple changes beat subtle ones.
- Optimizely is the enterprise experimentation programme, visual editor included, at contracts reported from $50,000 a year upward. If your traffic cannot produce several hundred conversions per variant, no tool in this list can rescue the maths.
Quick comparison
| Tool | Best for | Starting price | Free plan |
|---|
 VWO | Full CRO suite with a visual editor | $314 per month | Yes |
 AB Tasty | Personalisation alongside testing | See website | No |
 Convert | Privacy first testing for agencies | See website | No |
 Optimizely | Enterprise experimentation programmes | See website | No |
 Kameleoon | AI assisted enterprise testing | See website | No |
 Zoho PageSense | Small sites on a small budget | $16 per month | No |
 Crazy Egg | Simple tests plus heatmaps | $29 per month | No |
 Unbounce | Landing page testing for campaigns | $74 per month | No |
Is your product missing or out of date?
We review this list regularly. Vendors can request a listing, send updated pricing or ask about featured placement.
#1VWO
Best for: Full CRO suite with a visual editor4.4/5 on
G2 (929 reviews, checked Oct 2026)
Visit vwo.com
Starting price$314 per month
Free planYes
VWO is the mid market default because it refuses to separate finding problems from testing fixes. Most failed testing programmes are not short of tools; they are short of reasons to test anything specific, and VWO Insights manufactures those reasons from recordings and funnel leaks. The visual editor is mature, and the free Starter tier is a genuine way to learn the discipline.
The trade is the merger and the meter. Wingify combined VWO and AB Tasty in 2026 and says the products will converge over time, with no published timeline; a buyer signing a two year deal should ask what convergence means for their contract. And per user pricing means a viral traffic spike is a bill, not just a blessing.
What we like- The whole optimisation loop in one platform: Insights heatmaps, recordings and funnels find the problem, Testing's visual editor builds the fix, and a Bayesian stats engine judges it, a free Starter plan covers up to 50,000 monthly tracked users, split URL and multivariate testing included as tiers rise
Watch out for- Since January 2026 VWO sits under Wingify alongside AB Tasty, with pricing pages redirected to Wingify and tiers now quote led, so confirm your tracked user allowance in writing, multivariate and advanced targeting are gated to Pro, reported from $972 a month, monthly tracked user pricing punishes high traffic, low conversion sites
#2AB Tasty
Best for: Personalisation alongside testing4.4/5 on
G2 (422 reviews, checked Oct 2026)
Visit abtasty.com
Starting priceSee website
Free planNo
AB Tasty built its name where testing meets personalisation: the same platform that splits traffic can also show returning customers a different experience from new ones, and its retail and travel clients use exactly that blend. Its editor is as capable as VWO's, and its targeting is a notch more ambitious.
The trade is the same merger from the other chair. AB Tasty customers are now Wingify customers, the two flagship products are expected to converge, and nobody outside the company knows the merged roadmap. The product today is strong; the contract conversation should include migration protection in writing.
What we like- A strong visual editor paired with AI driven personalisation, product recommendations and feature management, so tests and targeted experiences share one platform, a long enterprise and retail client base in Europe and beyond, Bayesian statistics in current comparisons
Watch out for- Quote based pricing throughout, now combined with VWO under Wingify as of September 2026, so product direction is a live question to put to the sales team, depth is wasted on teams that only need occasional headline tests
#3Convert
Best for: Privacy first testing for agencies4.6/5 on
G2 (98 reviews, checked Oct 2026)
Visit convert.com
Starting priceSee website
Free planNo
Convert is the grown up's testing tool: one job, done carefully, with privacy architecture that predates the industry's compliance panic and pricing published while rivals hide theirs. Agencies like it because client tests can run without a data processing negotiation each time.
The trade is scope and stats taste. Convert will not show you heatmaps or run your personalisation programme, and its frequentist reporting, confidence intervals and p values, demands a slightly more statistical reader than a Bayesian chance to beat control figure. For a disciplined testing team, neither is a defect; for a team wanting an all in one, VWO is the shape.
What we like- A testing specialist with a privacy led design: no personal data collection by default, which simplifies GDPR conversations, published pricing and a flicker free testing tag are its stated selling points, A/B, split and multivariate testing plus full stack experiments, with 90 plus integrations, a favourite of agencies running tests for clients
Watch out for- Current comparisons report its pricing inconsistently, from $99 to $399 a month for entry tiers, so the live pricing page is the only safe source, a frequentist stats engine, which some teams read less intuitively than Bayesian probability, less behaviour analytics depth than VWO
Starting priceSee website
Free planNo
Optimizely sells experimentation as an organisational capability: many teams, one platform, every result recorded, no duplicate tests, no orphaned variants left live by accident. Its visual editor is genuinely marketer friendly, but the product's centre of gravity is the programme around the editor.
The trade is simple arithmetic. Divide a $50,000 plus contract by the number of meaningful tests your traffic can actually support; if the answer per test has too many zeros, the programme is a costume. Enterprises with real traffic and many teams should shortlist it by default. Everyone else should admire it from the cheaper seats.
What we like- The enterprise experimentation standard, covering web testing with a visual editor, feature experimentation, flags and personalisation under one governance model, both Bayesian and frequentist statistics options in current comparisons, the programme management layer, roadmaps, results libraries, that large organisations need to run hundreds of tests without chaos
Watch out for- Contracts reported from $50,000 a year upward, a sales led purchase with implementation to match, absurd overkill for a marketing site running four tests a year
#5Kameleoon
Best for: AI assisted enterprise testing4.6/5 on
G2 (137 reviews, checked Oct 2026)
Visit kameleoon.com
Starting priceSee website
Free planNo
Kameleoon competes on intelligence: AI that predicts which visitors a variant will move, layered over a conventional visual editor and flag platform. For large retail and media sites, that targeting can find lift in segments a flat 50/50 split would wash out.
The trade is the same volume law that governs this entire category, sharpened. Machine learning on thin traffic learns nothing, confidently. Kameleoon's advantages are real exactly where its price is aimed, at high volume programmes, and theoretical everywhere else.
What we like- An enterprise platform combining web testing, personalisation and feature flags, with AI assisted audience targeting as its signature, both stats approaches available per current comparisons, a strong European presence and privacy posture
Watch out for- Quote based pricing at enterprise level, less mindshare in North America than Optimizely or the Wingify pair, AI targeting features need traffic volume to learn anything, which small sites cannot supply
#6Zoho PageSense
Best for: Small sites on a small budget4.1/5 on
G2 (38 reviews, checked Oct 2026)
Visit zoho.com
Starting price$16 per month
Free planNo
PageSense is the correct first purchase for a small site: real tests, real heatmaps, a real editor, for the price of a team lunch. It teaches the discipline, what a hypothesis looks like, how long a test actually needs, at a cost where the lesson is cheap.
The trade is the ceiling, which for most small sites is not a trade at all. A site with 5,000 monthly visitors cannot feed a bigger platform's statistics anyway; PageSense's limits will bind around the same time traffic outgrows them, and by then the upgrade conversation is funded by results.
What we like- The cheapest credible way to start testing, at about $16 to $17 a month in current comparisons, A/B and split URL testing with a visual editor, plus heatmaps, funnel analysis and session recordings in the same small bill, part of the Zoho suite, so it sits comfortably beside Zoho CRM and Campaigns for existing customers
Watch out for- Statistical depth and targeting sophistication trail the specialists, traffic based limits on lower tiers cap serious programmes, few agencies build their practice on it, so peer knowledge is thinner
#7Crazy Egg
Best for: Simple tests plus heatmaps4.2/5 on
G2 (128 reviews, checked Oct 2026)
Visit crazyegg.com
Starting price$29 per month
Free planNo
Crazy Egg is a heatmap tool that learned to test, and for its audience that order is right. Most small site improvements are not subtle: the button nobody sees, the form field everyone abandons. Crazy Egg shows you the problem in colour and lets you test the fix before lunch.
The trade is that serious programmes outgrow it in a quarter. There is no experimentation roadmap, no audience platform, no governance. Buy it to start, learn what a real test cycle feels like, and graduate with evidence about what you actually need, which is a better use of $29 than most marketing spend.
What we like- The friendliest entry in the category: heatmaps and scrollmaps show where visitors click and stop, and a simple visual editor turns the obvious fix into a test without code, plans from $29 a month, setup measured in minutes, which is why bloggers and small businesses have used it for years
Watch out for- Testing depth is basic: no multivariate, no server side, frequentist statistics without the programme machinery, not a platform for a dedicated CRO team
#8Unbounce
Best for: Landing page testing for campaigns4.3/5 on
G2 (386 reviews, checked Oct 2026)
Visit unbounce.com
Starting price$74 per month
Free planNo
Unbounce fits the paid media workflow: a campaign needs five landing pages by Friday, each with a test running, and no developer in sight. Its builder is the product; the testing is the quality control on what the builder produces, including AI routing that shifts traffic toward winners automatically.
The trade is that automatic shifting is early stopping wearing a suit. On healthy traffic it finds lift faster; on thin traffic it can crown noise. Use Smart Traffic for low stakes campaign pages, run disciplined fixed split tests for decisions you will build on, and never use Unbounce's numbers to judge a page it did not build.
What we like- A landing page builder with testing built into the page workflow, so campaign pages are created and A/B tested in one place, plans from $74 a month in current comparisons, Smart Traffic, its AI routing, sends visitors toward the variant converting best for their attributes, no developer needed for paid media landing pages
Watch out for- It tests pages it builds
- It is not a tool for testing your existing website, AI traffic routing optimises on small samples by design, which needs the statistical caution in the guide below, per page and traffic limits shape the tiers
How we chose
We picked current visual testing platforms across the market's real tiers, from $16 a month to enterprise contracts, favouring tools with a genuine point and click editor, since that is the keyword's promise. The Wingify merger is stated in both affected entries because it changes what a buyer is signing. Pricing is published figures from 2026 comparisons and is See website where vendors quote only or where comparisons conflict. We did not claim hands-on testing, and no vendor's claimed average lift is repeated anywhere in this guide, because average lift claims are marketing, not measurement.
Sample sizes, peeking, and other ways to lie to yourself
The arithmetic is unforgiving. To detect a 10 percent relative lift on a 3 percent baseline conversion rate, you need on the order of tens of thousands of visitors per variant. Most small sites cannot detect small effects at all, which dictates strategy: test big, legible changes, the offer, the headline, the page structure, where the plausible effect is large enough for your traffic to see. Button colour tests on 2,000 monthly visitors produce only superstition.
Peeking is the second failure. Checking the dashboard daily and stopping when a variant crosses significance guarantees false winners, because repeated looks multiply the false positive chance far beyond the stated level. Set the sample or the runtime in advance, at least one full business cycle, usually one to two weeks, to cover weekday patterns, and do not touch the stop button before it, whatever the graph is doing. Every tool here will let you peek; discipline is not a feature tier.
Third, run an A/A test once: the same page against itself. It should never reach significance. If it does, something in your setup, tracking, assignment, sample ratio, is broken, and every result the tool has given you is suspect. Vendors suggest this themselves; almost nobody does it. Be the team that does.
Last, record everything. A results library, hypothesis, screenshot, sample, outcome, including the losses, is what turns testing from a slot machine into a research programme. The enterprise platforms sell this bookkeeping as a feature; a small team can keep it in a shared document. Either way, the team that writes down what it learned compounds; the team that only remembers its wins repeats its losses at full price.
Frequently asked questions
How much traffic do I need to A/B test properly?
Enough to produce several hundred conversions per variant for the effect you hope to detect. High traffic sites can detect small lifts; low traffic sites should test large changes only, run tests longer, and accept that some questions are unanswerable at their volume. A sample size calculator, built into several tools here, will give you the honest number before you start.
What happened to Google Optimize?
Google shut it down in September 2023. Sites that relied on the free tool moved to the platforms in this list, or to the experiments features inside analytics and product platforms. Any current guide recommending Google Optimize as a free option is out of date; the small site role it played is now covered by VWO's free Starter tier and the budget tools here.
Bayesian or frequentist statistics, which is better?
Both are valid; they answer slightly different questions. Bayesian engines report a probability that a variant beats the control, which most stakeholders read naturally. Frequentist engines report confidence against a null hypothesis, the classical standard. What matters more than the school is behaviour: pre set samples, no early stopping, and full business cycles. Either engine punishes peeking.
Are VWO and AB Tasty the same product now?
They are sister products under Wingify, which combined them in 2026, announcing the merger in January and relaunching under the Wingify brand in September 2026. Both products continue to run during a transition and are expected to converge over time, with no published timeline. Buyers should get roadmap and migration commitments in writing before signing a multi year term with either.
Can a visual editor break my site?
It can if variants are built carelessly: heavy edits on complex pages can flicker, where visitors briefly see the original before the variant loads, or clash with the site's own scripts. Reputable tools use small, fast tags designed to minimise flicker, and server side testing removes it for critical pages. Test every variant on real devices before launch, and keep the first tests simple.
Sources we checked (0)
Is your product missing or out of date?
We review this list regularly. Vendors can request a listing, send updated pricing or ask about featured placement.
Shihab Mia · Founder, Rankite
Shihab has 5+ years in B2B and SaaS SEO. He writes Rankite's software buying guides from pricing pages, documentation and public user reviews. More about Shihab.