About

Built by the people who had the question first.

Checkoutlab is made and supported by Root Agency, a Shopify development studio in the Netherlands. It exists because a client asked what the express buttons in their checkout were really doing, and there was no honest way to find out.

Why this app exists

And why it is deliberately small.

Express checkout buttons are one of the few things in a Shopify checkout that split opinion straight down the middle. One camp says they remove friction and lift conversion. The other says they skip the moment where a customer would have added something, or confuses people into a wallet they do not have. Both camps have case studies. Neither has your store.

The honest answer is that it depends on your customers, your market and your device mix, which makes it an empirical question rather than an opinion. But until recently it was almost impossible to test: you cannot A/B test the Shopify checkout with a script, and the obvious workarounds either break the checkout or measure through a pixel that ad blockers and cookie banners quietly eat.

Shopify Functions changed that. A payment customization can hide or move payment methods inside the real checkout, and the cart attribute that decides it travels all the way onto the order, so the measurement can come from order webhooks instead of a browser. That is the whole idea behind Checkoutlab: the variant runs where the checkout runs, and the counting runs where nothing can block it.

Finishing the job

Most A/B tools stop at the number. You get a winner, a p-value and a congratulations, and then the actual change is a ticket that sits in someone's backlog until the next replatform. In a checkout that gap is worse than usual, because the thing you just proved is hidden behind a Shopify Function that someone has to go and configure by hand.

So redeeming is built in. A finished experiment turns into a rule with one action: same configuration, same markets, same conditions, now applied to everyone. And because rules and experiments share one precedence ladder, asking the question again later does not mean unpicking anything. The new experiment simply takes over for as long as it runs, and the rule resumes afterwards.

What we chose not to build

There is no web pixel, no storefront access token and no customer data. Not because it was hard, but because every one of those would have been a thing you had to trust us with, in exchange for numbers we can get another way. The app asks for four access scopes and the narrowest set we could make work.

There is no paywall in front of building, either. You can install the app, configure an experiment and preview both arms on your own storefront without paying us anything. What a plan buys is the moment something reaches a real shopper, which is both the only moment we are actually costing you anything and the only moment there is anything to measure.

And there is no benchmark feature, no “stores like yours see X%”, no aggregated insight product. Your data answers your question and is then deleted. That is the entire deal.

Who you are dealing with

Root Agency, registered in the Netherlands under Chamber of Commerce number 95339043. Support, development and the answer to your privacy question all come from the same small team. Mail hello@root-agency.nl and you get the person who wrote the code, within one business day.

We are not affiliated with, endorsed by or sponsored by Shopify Inc. We just build on it, most working days.

Have a checkout question of your own?

If it is about payment methods in the checkout, the app probably tests it. If it is about something else, tell us anyway. Knowing what merchants actually want to know is how this one got built.