Free A/B testing toolkit
Know when B really beats A
ABFinal is a set of free calculators for planning and reading A/B tests: how many visitors you need, how long to run, whether the split is healthy, and what the result actually says. Everything runs in your browser. There's no account and nothing to install.
Example: checkout button test
10,000 visitors per arm: 320 vs 380 orders
Seven calculators for the whole test
Use the first four before launch, the SRM checker while the test runs, and the rest when it's time to decide. Each page explains its formula with a worked example.
Calculator
Sample size
How many visitors each variant needs to detect the change you care about, at your chosen confidence and power.
Calculator
Significance
Paste visitors and conversions for A and B. Get the p-value, confidence interval for the lift and the Bayesian chance that B beats A.
Calculator
Test duration
Turn the required sample size into days and full weeks, based on your daily traffic and the share of it in the test.
Calculator
MDE
Given your traffic, find the smallest lift (and drop) a test can reliably detect, before you spend weeks on it.
Calculator
SRM checker
Check whether the visitors in each group match the split you planned. A mismatch means something in the setup is broken.
Calculator
Price test
Compare two prices on conversion, revenue per visitor and gross profit per visitor, with the break-even conversion rate for the new price.
Calculator
A/A simulator
Simulate thousands of tests where nothing changed and watch how often a false winner appears, with and without peeking.
For X (Twitter)
Did your second post really do better?
X doesn't offer split testing for organic posts, so most people compare two posts by eye. Paste the impressions and engagements from each post's analytics and the comparer tells you whether the gap in engagement rate is larger than chance would produce. Treat it as a strong hint, not proof: the two posts went out at different times to different slices of your audience.
X tool
Post A/B comparer
Compare the engagement rates of two X posts and see whether the gap is bigger than chance would produce.
X tool
Variant planner
Plan a fair test of two post variants on X - impressions needed, number of posts, a balanced schedule and a character count for each draft.
Guide
How to A/B test posts
A step-by-step method for comparing post styles on X fairly, from choosing one variable to reading the totals.
Get your numbers from post analytics on x.com. ABFinal is not affiliated with X Corp.
How a test goes wrong, and which tool catches it
| Problem | What it looks like | Check with |
|---|---|---|
| Too little traffic | A "winner" that disappears next month | Sample size, MDE |
| Stopping early | You checked daily and stopped on the first good p-value | A/A simulator, Duration |
| Broken split | A 50/50 test that delivered 52/48 | SRM checker |
| Wrong metric | Conversion went down after a price rise, but profit went up | Price test calculator |
| Misread result | "95% significant" read as "95% chance B is better" | Significance + Bayesian |
Guides
Longer pages for when you need the reasoning, not just the number.
Pillar guide
A/B testing for CRO
The complete guide to using A/B tests in conversion rate optimization, from research and hypotheses to analysis and rollout.
Program
Testing roadmap
Turn a pile of ideas into a testing roadmap you can actually run, with an ICE scorer and a capacity check.
Pricing
Price A/B testing
How to test prices on an online store, which metric to judge by, the break-even math and the legal and customer-trust risks.
Checklist
CRO checklist
An interactive 24-point conversion audit for online stores. Fix the basics before you spend traffic on tests.
AI
AI and A/B testing
What AI features in testing tools actually do - generated variants, multi-armed bandits, personalization, automated analysis - and the trade-offs of each.
Paid social
Meta ads A/B tests
How to A/B test Facebook and Instagram ads with Meta's built-in tool, what to test, how to size the budget and how to read the results.
Honest comparisons
ABFinal doesn't run experiments, so it isn't a replacement for a testing platform. These pages explain what each platform does today and where a free calculator fits: Optimizely, VWO, AB Tasty, Intellimize and Mutiny.
From the blog
- How Long Does It Take to Run an A/B Test?As long as it takes to reach the planned sample, rounded up to whole weeks - usually two to six weeks. Here's how to calculate yours.
- 12 Common A/B Testing Mistakes (and How to Avoid Them)The mistakes that produce false winners and wasted tests, from peeking and underpowered samples to broken splits and metric fishing.
- What Is the Significance Test for A/B Testing?For conversion rates, it's usually a two-proportion z-test (or the equivalent chi-square test). Here's how it works and how to read it.
- How to Optimize Your Prices on ShopifyKnow your margins, do the break-even math, choose a testing approach and judge results by profit per visitor.
- What Is the Difference Between A/A and A/B Testing?An A/B test compares two different versions; an A/A test compares a version with itself to check your setup. Here's when to run one.
- How to Improve Shopify Conversion RatesWhere Shopify stores usually lose buyers, the fixes to make first, and the changes worth testing - in order of impact.
Common questions
Is ABFinal free?
Yes. Every calculator and guide is free, with no account, no trial and no paid tier. The calculators run in your browser, so there is nothing to install and nothing to sign up for.
Does ABFinal run A/B tests on my store?
No. ABFinal does not split traffic, change your pages or connect to Shopify. It helps you plan a test before you launch it in whatever tool you use, and read the numbers once it ends. You bring the visitor and conversion counts; the calculators do the math.
Are the numbers I enter stored anywhere?
No. All calculations happen in JavaScript on your device. Nothing you type is sent to us, and the site sets no cookies and loads no analytics or ad scripts.
Which statistical methods do the calculators use?
The significance calculator runs a pooled two-proportion z-test (equivalent to a 2x2 chi-square test) and an exact Beta-Binomial probability that B beats A. Sample size follows the same normal-approximation formula as R's power.prop.test. The methodology page lists every formula and the reference values we test against.
What happened to the old AB Final Shopify app?
This domain used to host a Shopify A/B testing app. That app and its team are not connected to the current site. ABFinal today is an independent, free toolkit; we can't help with accounts, billing or data from the old app.