Contents6
Nobody leaves a store thinking “it’s slow”. People leave because the photo lags, the “add to cart” button responds half a second too late, the page jumps just as they touch the screen. Slowness does not show in sales reports: it reads as small losses at every step of the journey.
This guide is for teams running an online store who wonder whether speed deserves a project. It does not promise a percentage of extra sales: it shows what studies actually measure, what weighs a store down, and how to size the stakes with your own data. For the measurements themselves (LCP, INP, CLS) and their thresholds, our article on Core Web Vitals for business leaders covers the ground.
In short
The most quoted study, run by Deloitte for Google, measures the effect of a tenth of a second, not a full second, and on large brands.
The effect shows at every step of the journey, from product list to order.
What slows a store down: images, apps, reviews, carousels, videos and tracking tools.
The stakes are sized with your mobile visits, conversion rate and average order value, then checked with a before-and-after measurement.
What the most quoted study measures
In 2019, Deloitte observed for Google the mobile sites of 37 large European and American brands, across more than 30 million sessions. The study, published as Milliseconds Make Millions, measured what a tenth of a second gained on several speed metrics changes. For retail sites, the effect reads step by step:
Increase in progression from one step to the next, for 0.1 second gained
At the end of the journey, the same study measures 8.4% more conversions on retail sites, and shoppers spending 9.2% more. These figures call for two precautions. They come from large brands and from a study commissioned by Google, which has an interest in a fast web: they give a direction, not a promise for your store. And they concern a tenth of a second: nothing allows multiplying them by ten for a whole second.
What the study does say clearly is where speed matters most: between the product page and the basket, the moment the buyer decides.
The pages that matter in a store
A store is not a home page. Visits often land on a product page, from Google or an ad, then go through the collection, the cart and checkout. Four pages therefore deserve their own measurement:
- The collection, loaded with images, filters and sometimes promotional labels.
- The product page, with its gallery, reviews, recommendations and widgets.
- The cart, where an upsell app can add its own code.
- Checkout, often hosted by the platform, worth checking all the same.
The measurement that counts is the phone’s, on an ordinary mobile network. The Core Web Vitals report in Search Console provides these field figures, drawn from real visits, by group of pages; our article on Core Web Vitals explains how to read it.
What weighs a store down
Shopify says it itself, in the help that comes with its performance dashboard:
Each feature that you add to your online store can impact your store’s web performance. Some features can slow down your site or create unexpected behaviors.
The same page names the usual suspects: theme code, apps, images, videos, carousels, social media feeds and analytics. None is useless in itself; it is their sum, rarely decided by anyone, that weighs.
From the field
A 4.2 MB store
On an online plush toy store we are putting back on its feet, the mobile diagnosis found a Lighthouse score of 65, main content displayed at 8.4 seconds and a 4.2 MB page. Four to five review systems were loaded, with no review displayed: a cost in speed without the benefit of social proof. On Palmora Property, a property listings site we run, the image pipeline converted 3,892 originals into 7,784 files in the right formats in 16 minutes, and a thumbnail went from 353 KB to 17 KB.
The first two items to look at are therefore almost always the same: images, served too heavy or in the wrong format, and apps, each of which can load its code on every page. The second is the subject of our article on the cost of Shopify apps, which recalls that Shopify itself caps at ten Lighthouse points the loss accepted for an app with its label. Reviews, for their part, deserve to be displayed rather than loaded for nothing, as our article on customer reviews explains.
Sizing the stakes with your own numbers
The useful question is not “how much does a fast store sell”, but “how much does my mobile store represent, and what improvement would make the project pay off”. The calculation takes four steps:
- Measure today.
A dated lab reading of the four key pages, and the field figures from Search Console.
- Set the stakes.
Mobile visits for the month, multiplied by the mobile conversion rate and average order value: that is the revenue speed touches.
- Find the break-even point.
Set the project’s cost against that figure: it tells you what rise in conversion would pay it back, without promising anything.
- Measure after.
Same tool, same pages, same conditions, then mobile conversion over several comparable weeks.
Conversion rate and average order value are two of the five numbers to track every week; if you already track them, step two takes five minutes. One precaution: a rise after a speed project can also come from a season, a campaign or a promotion. Note them next to the measurement.
Key takeaway
Speed is won through decisions, not settings
Every app, every carousel, every tracking tool added is a decision that costs a little speed. Writing them in a list, with what each brings in, is a store’s best performance tool.
Who decides on speed
Technical causes almost always trace back to business decisions. The choice of theme and its visual effects. An app added for one promotion and never removed. The full-resolution photo sent by the supplier and published as is. The choice of platform itself, which decides what you can control: our Shopify or Medusa comparison discusses it, and a store served as ready-made pages starts with a head start, as our guide to static site hosting explains.
Speed also serves the store’s other projects. The product page that converts only converts if it displays. The paid click of a Google Shopping campaign leads to a page that must respond at once, and the product data sent to Google is prepared with our Google Shopping and marketplaces guide. An added site search engine or one more payment module (see real payment fees) each bring their own code. Emails, whether abandoned cart reminders or welcome, post-purchase and repeat-purchase messages, bring customers back to those same pages.
Other channels follow their own rules: on a marketplace like Amazon, page speed is not in your hands, but stock and listings are (see inventory sync). A store in several languages multiplies the pages to watch (the multilingual, multicurrency store), the returns page and its forms are measured like the rest (the hidden cost of returns), and product pages enriched at scale must be enriched without weighing pages down (a catalogue of thousands of products under control). The other decisions about a store are gathered in our ecommerce section.
If your store feels slow on a phone, our ecommerce support starts with a dated measurement of the pages that sell, then the list of what weighs them down and what each element brings in.
Frequently asked questions
One second less loading time, how many more sales?
Nobody can promise that for your store. The Deloitte study carried out for Google measured 8.4% more conversions for a tenth of a second gained on mobile retail sites of large brands, but that result does not multiply by ten for a second. The right figure is yours: measure mobile conversion before and after a project.
Which store pages should be measured first?
The four pages that sell: a collection, a product page, the cart and checkout, on a phone. The home page matters less than people think, because many visits land directly on a product page, from Google or an ad. Take the pages with the most visits, and always measure them under the same conditions.
Do Shopify apps slow the store down?
They can, and Shopify acknowledges it in its help: every feature added can weigh on performance. An app that loads its code on every page, or that leaves code behind after uninstalling, costs speed without bringing anything in. Take stock once a year, and remove what no longer serves.
Do I need to change platform to get a fast store?
Rarely as a first step. Heavy images, surplus apps and third-party scripts can be fixed on the current platform, and that is where to start. Changing platform makes sense when slowness comes from what you cannot control, or when other reasons already push you to migrate: cost, features, dependency.
Sources
Jim · founder of Brain Plus AI
Also in e-commerce
Choose and price5 min
Shopify: when one more app costs more than it earns
Every Shopify app promises one more sale; each one also adds a subscription, code on the store and a dependency. How to count what it really costs, and when removing it pays more than adding it.
Practical guide6 min
Online store: five numbers to track every week
An online store dashboard shows dozens of metrics; five are enough to decide every Monday. Which ones, how to read them together, and what each one prompts you to do.
Practical guide5 min
Product page that converts: what buyers look for
Before paying, buyers check four things: what exactly the product is, what they will pay in total, when they will receive it and what happens if it does not suit them. A page that answers those questions sells better than a page full of fine phrases.