home / compare / switch one experiment

Switch one experiment. Not your whole program.

ABTestly plan details on this page reflect the published pricing page as of 6 September 2026.

You already run A/B tests in Convert, VWO, Optimizely, AB Tasty or PostHog. You do not need another demo account full of someone else's data. Pick one experiment that is live right now, send us its setup, and we recreate it in ABTestly with you. Then you judge the switch on one real piece of your own work.

The offer, in full. Send one experiment. We map it, write the variation code, build the experiment with you, and hand back three plain lists: what carries across, what has to be re authored, and what ABTestly cannot reproduce at all. It costs nothing, we never ask for a login to your current tool, and nothing has to be cancelled anywhere until you decide.
Send one experiment Or start the trial yourself

The button opens your mail client with the checklist already written out. Nothing reaches us until you press send.

Who this is for

Who this is not for

What we recreate

Here is the mapping, part by part. Nothing imports automatically. There is no connector to any other vendor and we are not going to imply one. The recreation is us rebuilding your experiment by hand from what you send.

Part of your experimentIn your current toolIn ABTestly
The variation itselfVisual editor edits stored as generated selectors and CSS, or custom JavaScriptRe authored as variation code in the Monaco editor. The JavaScript and CSS become explicit and reviewable, and you can copy them into your own repository
Primary goalA click, pageview, custom event, or revenue goalRecreated as a goal with an activation trigger: click, view, or a manual call from your own code. Triggers other than on load are a Pro feature, as the targeting row notes for geo. Two system goals attach to every new experiment automatically
Secondary goalsAdditional tracked metricsAttached to the same experiment alongside the primary goal. Conversion and revenue are both supported
TargetingURL, device, query parameter, cookie, behavioral, and custom conditionsRecreated as targeting rules: URL, device, and query parameters. URL rules are case insensitive by default. Geo targeting is a Pro feature. If a rule of yours does not map, we tell you instead of approximating it
Traffic splitPercentages per variant, plus an overall share of trafficVariant weights and traffic allocation, both in basis points, so 5000 is 50 percent
Checking it before launchPreview or debug modePreview links that force a specific variant on your live site, so you see the real page before any visitor is bucketed
ResultsYour vendor's reportingFrequentist on every plan, sequential and Bayesian on Pro, applied to conversion and revenue. The Bayesian view reports a credible interval, which is not the same thing as a significance test
Past resultsLive inside your current toolNot imported, and no tool can import them honestly. Export your history before you cancel anything. New data starts fresh here

What you need to send

One email. The button on this page pre writes it, but here is the whole checklist so you can see there is nothing hidden in it.

  1. Which tool you are on now. Convert, VWO, Optimizely, AB Tasty, or PostHog.
  2. The page or URL pattern the experiment runs on.
  3. What the variation actually changes. A sentence is enough. A screenshot works too, it just gets you a vaguer answer. Your existing variation code is the most exact thing you can send.
  4. The primary goal: the exact action that counts as a conversion, and where on the page it happens.
  5. Any secondary goals you watch alongside it.
  6. The targeting rules: URL conditions, device, audience, anything that decides who qualifies.
  7. The traffic split between variants, and what share of traffic enters the test.
  8. Rough monthly traffic on that page. This is what tells us whether one experiment can read a result for you at all.

What you do not need to send. A login. We never ask for access to your Convert, VWO, Optimizely, AB Tasty, or PostHog account, and you should not give it to any vendor pitching you. Strip the URLs and the numbers if you want to. The structure is what we read.

What we do, and what stays with you

What ABTestly does for you

  • Maps the experiment, including the parts that do not map. You get three lists: carries across as is, has to be re authored, cannot be reproduced.
  • Writes the variation JavaScript and CSS that reproduces your change, and hands you the code to read before it runs anywhere.
  • Builds the experiment configuration: variants, weights, goals, activation triggers, targeting rules.
  • Sets up the preview links so you can check each variant on your own live page.
  • Checks the instrumentation with you once it is live: that the right visitors qualify, that exposure fires only after consent, that goals fire on the action you defined, and that the results panel shows no sample ratio warning.
  • Answers your questions directly. The person who built the experiment is the person who replies.

What stays with you

  • The account is yours. You sign up, you hold the site. We work from what you send and hand back the code and a written configuration. If you would rather watch it being built, add us to the site as a seat and remove us afterwards.
  • You install the loader. A few lines in <head>. It pulls a runtime of about 31 KB gzipped, measured on our own snippet, with the loader itself under 1.5 KB.
  • You wire consent to your CMP if you operate under GDPR. The consent guide covers it.
  • You approve the variation code before it touches live traffic. It is your site. We do not publish anything on it.
  • You press start, and you decide. Nothing here obliges you to switch, and nothing here asks you to cancel anything first.

The five steps, with honest timings

1. You send the experiment (10 minutes). The checklist above, in an email. Redact whatever you want to redact.

2. We map it and write back (within two business days). You get the three lists before you install a single line of code. If the honest answer is that ABTestly cannot reproduce your experiment, that is the answer you get, and it has cost you ten minutes rather than a migration.

3. You install the loader (15 minutes). Sign up, add your site, paste the loader into your <head>. The install guide has the exact snippet your site generates.

abtestly.jsloaded
1<!-- ABTestly · paste first in <head> -->
2<script>(function (w, d) {
3  var s = d.createElement('script');
4  s.src = 'https://cdn.abtestly.com/s/YOUR_KEY.js';
5  s.async = false;
6  (d.head || d.documentElement).appendChild(s);
7})(window, document);</script>

4. We build it, you check it (30 to 60 minutes of your time). We hand over the variation code and the configuration. You read the code, open the preview links, and confirm each variant renders correctly on your real page, on the devices you care about. Nothing is live yet.

5. You start it, and you decide. Whether one experiment runs long enough to read a result depends on your conversion rate and the size of the effect you are looking for, so put your own numbers into the A/B test calculator before you set expectations. Config is served from Cloudflare's edge; the bucketing and the variation itself happen in the visitor's browser.

Five step timeline for moving one experiment, marking which steps cost your time and which are ours, with the exit at step two.
The five steps, and whose time each one costs. Step two is an exit, not a formality.

What you can compare at the end

This is the point of the exercise. After one experiment you have direct evidence on all of the following, from your own site rather than from a comparison table we wrote. If you would rather see the shape of it before sending anything, one experiment built end to end walks through a sticky add to cart test in full: the brief, the variation code, the targeting and goals, the QA pass, and the place where this pattern is usually built wrong.

What you cannot compare, and we will not pretend otherwise

A head to head conversion rate race between two tools. Run the same change in both tools at once and some visitors land in both, some in neither, and two scripts fight over the same elements, so the conversion numbers stop meaning anything. ABTestly's exclusion groups coordinate experiments inside ABTestly and have no say in what another vendor buckets, so no audience built here can keep a visitor out of the other tool. A clean split has to happen once, upstream of both tools, with a deterministic rule on a stable visitor id, exposed to each tool as an audience condition. The migration guides set out how to do that if you want it.

An exact numbers match. Even done properly, two tools will not agree to the decimal. Different identity, attribution, consent handling, and goal definitions move the numbers for legitimate reasons. Treat any cross tool check as a plumbing check, not a scoreboard.

Historical performance. Your past results stay in your current tool. Export them before you cancel.

What it costs

The same terms are set out on one page, in a form you can forward to whoever signs off the spend, at evaluate one experiment.

ABTestly prices exclude tax and are billed via Paddle as merchant of record. Applicable tax is handled at checkout.

When we will tell you not to switch

Honesty is the thing we are actually selling, so it applies to this offer too. We will tell you to stay where you are when your experiment depends on a visual editor in the hands of people who do not write code, on multivariate or personalization campaigns, on feature flags or server side experiments, or on session recordings and surveys. ABTestly does none of those. Experiment heatmaps are included with Pro and Business, and the Signals add on brings them to Starter, but recordings and surveys are not part of this product.

The comparison pages carry the same argument in more detail, including where each competitor genuinely wins: Convert, VWO, Optimizely, AB Tasty, and PostHog.

One more thing, before you send anything. Check your renewal date first, because the expensive mistake at this stage is not about us. Optimizely subscriptions renew for twelve months on their own and stopping that takes ninety days written notice before the term ends, so if your renewal is eight weeks out, the decision that saves you money is giving notice today rather than sending us an experiment. VWO quota keeps burning while you evaluate in parallel, and when it runs out VWO pauses the campaign rather than billing you, so a comparison can stop halfway through on its own. Convert bills overage only if you have switched it on. None of that changes what we do here, and all of it changes when you should start.

Common questions

What does it cost to have one experiment recreated?

The recreation itself is free and carries no obligation. Running the experiment needs a plan. Monthly plans start with a 14 day free trial. A card is required at signup and the first charge lands on day 15. Cancel before day 15 and nothing is charged. Annual billing and Enterprise are not eligible for the trial.

Do you need access to my current A/B testing tool?

No. We never ask for a login to Convert, VWO, Optimizely, AB Tasty, or PostHog, and you should not hand one to any vendor pitching you. We work from the description you send: the page, the change, the goals, the targeting, and the split. Redact URLs and numbers if you want to. The structure is what we read.

Can I run the same test in both tools at once and compare the conversion rates?

Not cleanly, and we will not pretend otherwise. Run the same change in both tools at once and some visitors land in both, some in neither, and two scripts fight over the same elements, so the conversion numbers stop meaning anything. ABTestly's exclusion groups coordinate experiments inside ABTestly and have no say in what another vendor buckets. A clean split has to happen once, outside both tools, on a stable visitor id, with each tool configured to serve only its own group. What one experiment does tell you is how the same change is built, checked, measured, and priced here.

Does anything import automatically from my current tool?

No. There is no connector to any other vendor. The recreation is us rebuilding your experiment by hand from what you send: variation code re authored in the Monaco editor, goals, targeting, and traffic split rebuilt in the dashboard. Past results do not come across either, so export your history before you cancel anything.

What if you cannot reproduce my experiment?

We tell you, in writing, before you install anything. ABTestly does A/B and split URL tests with multiple variants. It does not do multivariate, personalization campaigns, feature flags, server side experiments, session recordings, or surveys. If your experiment depends on one of those, that is a reason not to switch, and we will say so.

How long does the whole thing take?

About half a day of your time, spread over a few days: roughly 10 minutes to send the experiment, 15 minutes to install the loader, and 30 to 60 minutes to check the variants with preview links before you start it. The mapping write up comes back from us within two business days.

Can I do this without talking to anyone?

Yes. Start the trial, follow the docs, and rebuild the experiment yourself. The migration guides walk through the same steps unattended. This offer exists for people who would rather have the mapping done for them first.

Sources for the numbers on this page

Checked on 6 September 2026. Prices and product behaviour change, so verify against the source before you rely on a number.

ABTestly is an independent A/B testing platform for CRO developers and technical teams. Convert, VWO, Optimizely, AB Tasty, and PostHog are trademarks of their respective owners. ABTestly is not affiliated with, endorsed by, or sponsored by any of them.

one experiment, not a migration

Send one experiment.
Get it back, built.

One email, with the checklist already written for you. We map it, build it with you, and tell you plainly what does not carry across. Free, no access to your current tool, and nothing to cancel until you decide.

Send one experiment Or start the trial yourself