We'll claim the speed, not the lift.

There’s a chart every performance vendor eventually shows you: pages got faster, and look, conversion went up and to the right. It’s a persuasive chart, and it’s usually not honest, because a storefront is a moving target. Between the “before” and the “after” you changed a headline, ran a sale, swapped a hero image, and any one of those moves conversion at least as much as speed does. I know, because I tried to build that chart for LayerKick, recording a Shopify baseline and then comparing it once we were live, and watched a client change something mid-measurement and quietly skew the whole thing. So I stopped. We don’t claim we lifted your sales. We claim the one thing that is cleanly ours to claim, and we prove it on your own store.

The claim we can stand behind

It’s a plain, almost boring sentence, and that is the point: this share of your pages, over this window, was served this fast, and against your normal Shopify baseline that is this much faster. Every piece of it is measured on your own traffic, attributable to us, and hard to argue with. Faster is the thing we actually do, so faster is the thing we will put our name on.

We even keep that number honest against our own interest. We measure how much of a page came from cache by bytes, not by file count, because counting files lets one heavy uncached image hide behind forty tiny cached icons. And we caught ourselves counting our own platform scripts as cache hits, our code grading its own homework, and took them out. A speed number is only worth showing if it would still embarrass us when it’s bad.

The link from faster pages to more orders isn’t ours to assert, but it isn’t a mystery either. The industry has spent a decade showing that slower pages lose sales; you don’t need a vendor to hand you a number on top of that. We make the page faster and give you the receipt. What that does to your conversion you will read in your own dashboard, uncolored by our marketing.

Why we won’t draw the last line for you

The honest reason we won’t claim a lift is that we can’t cleanly separate our effect from yours. The only rigorous way to do it is a live experiment, splitting real traffic in the same moment so nothing else can drift between the two, and that is exactly what we are building. But even a clean split runs into something particular to how this works: a page we haven’t served yet isn’t fast yet, so early on a real slice of traffic lands on pages that aren’t warm, and pinning an order to “the fast one” or “the slow one” gets muddy in a hurry. We would rather tell you that plainly than paper over it with a conversion figure that wouldn’t survive you asking how we arrived at it.

What this layer is really for

Speed is one of three things worth having a thin layer between Shopify and your shopper. It’s also where insight lives, a real view of how shoppers move through your store that Shopify doesn’t hand you on its own. And it’s where testing happens server-side, real experiments that decide the page before it’s sent instead of flickering it in the browser. The faster pages are the part we can put a hard number on today. The rest is what the same layer is quietly built to carry, and most of the entries around this one are about one corner of it.