Early access. Every feature on this site carries its status. Where we are today

For stores with 300,000+ visits a month · pilot

Conversion work you pay for when it's proven.

A pilot service: we find what to test, build the variants, check them by hand, and run the tests. Every test is to be written down before it starts and settled by a neutral readout; those parts are being built, as tagged below. You pay a small monthly fee, and a share of the extra profit only when a change wins.

Why this model

Paying for tests that prove nothing is your risk today. Here, we share it.

Paying for months of conversion work with nothing proven is the risk you carry today. With this model you pay a share only for changes a neutral test says won.

How a test runs

Written down, built, run, settled.

  1. 01

    Write down

    Hypothesis, measure, sample size and stop date, before it starts.

  2. 02

    Build

    AI drafts the variant; a person checks it.

  3. 03

    Run

    Through Shopify Rollouts or your testing tool.

  4. 04

    Settle

    Won, lost or can't tell, from a neutral readout.

  • A/B tests with honest statistics Building
  • Every test written down before it starts, shown to you Pilot
  • Results read from Shopify Rollouts or your own testing tool Pilot
  • No inflated wins Building
  • Billing on verified extra profit, from your order data Pilot

Price

A small fee, and a share of verified profit.

Price on request

Terms are agreed with each store before the pilot starts.

Fit check

We'll tell you before you sign.

Below about 300,000 visits a month, most tests take too long to read, and you'd pay the fee for little.

  • A fit check before you sign: is there enough traffic to read a test in time? Pilot

Ask about the pilot.