Skip to content
Test

Turn your top opportunity into a live experiment

Your audit ranks each opportunity by what it's costing you. Test takes the one at the top, writes the variation code to fix it, runs it as a live A/B test, and measures the result until you know whether the change actually worked.

Built from your own audit. No setup.

Backlog2

Surface reviews next to the buy button

MajorTrust

Show shipping cost on the product page

MinorConfidence
Building1

Add a guest checkout path

BlockerFriction
Variation code drafted
Running1

Move express pay above the fold

MajorFriction
88% chance to win
Winner1

Reorder path for returning buyers

MajorLoyalty
96% chance to win
The variation code

Spar writes the change, you approve it before it ships

For the opportunity at the top of your list, Spar writes the variation as a self-contained snippet that changes the live page, with every headline, color, and label left editable. Nothing goes live until you approve it, so you always control what shoppers see.

OriginalControl
.checkout__gate {
require-account: true;
cta-label: "Create account to pay";
}
Spar's variationTreatment
.checkout__gate {
guest-checkout: enabled;
cta-label: "Continue to payment";
express-pay: "above-fold";
}

Approval gate before launch

Edit the copy, swap an image, or send it back. The test sits in Draft until you move it to Running.

Approved to run

Yes, we know guest-checkout: enabled isn't real CSS. We weren't sure if you knew, so: it's for illustration. The variations Spar actually ships are real code.

Reading the result

Watch the evidence add up, stop when it crosses your threshold

Spar measures the test itself rather than handing the numbers to GA4 or Clarity. It tracks each variation's chance to win as sessions accumulate, using sequential statistics so you can check progress whenever you want without the repeated peeking that normally inflates false positives. When the leader crosses your confidence threshold, Spar declares the winner.

Chance to win

Express pay above the fold

Winner declared
85% thresholdDay 1Day 12
  • The threshold is yours to set

    Pick how confident a variation has to be before Spar declares it the winner. The dashed line is that bar.

  • No fixed end date required

    Sequential statistics stay valid no matter when you stop, so a clear result can end the test early instead of burning two more weeks of traffic.

  • Inconclusive is a valid outcome

    If the curve never crosses, Spar says so rather than forcing a winner out of noise.

Going live

Ship it through Spar, or the platform you already run

The same approved variation goes live either way: Spar serves it directly, or pushes it into the experimentation platform your team already uses.

Spar ships it

Theme extension or embed script

On Shopify, the variation runs through Spar's theme app extension, with no theme edits and no code to paste. On any other stack, paste one Spar script into the page head. Either way, Spar serves the variation and runs the test for you.

<script src="serve.spar.../e.js" async>

Your platform ships it

Push to the tool you already run

Already running an experimentation platform? Spar pushes the approved variation into it and keeps the two in sync, so the test runs in the tool your team already uses.

ConvertGrowthBookIntelligems

Put your biggest opportunity to the test

Connect your store and Spar audits it, drafts a variation for your highest-impact opportunity, and runs the test for you. You approve what ships and read the result.