Is your online store too slow? How page speed costs sales
Shoppers on phones do not wait for slow product pages, and most small stores lose that time to apps, tags and oversized images.

- Speed matters most on the pages where shoppers decide: collection, product and cart pages, viewed on a phone.
- Start with field data from real visitors in PageSpeed Insights, then use lab data to find what to fix.
- On small stores, the usual causes are apps, leftover app code, tracking tags, heavy theme sections and oversized images.
- The best known speed and revenue study was commissioned by Google, so treat its uplift figures as direction, not a forecast for your store.
Your online store is too slow if real visitors on phones wait noticeably for collection, product or cart pages to appear and respond. The quickest check is the field data in Google's PageSpeed Insights, which shows how real Chrome users experienced the page over the last 28 days. On small stores, the cause is usually apps, tracking tags, heavy theme sections or images that are far larger than they need to be.
Speed is rarely the only reason a store underperforms. Price, shipping costs, product information and traffic quality all matter more on some stores. The commercial question is whether slow pages are costing you sales you have already paid to attract, and whether fixing them is cheaper than buying more traffic.
What the research says, and what it does not
The most cited study is Milliseconds Make Millions. web.dev's summary explains that Google commissioned it and that 55 and Deloitte ran it across 37 European and American brand sites and more than 30 million sessions, monitoring mobile load times for 30 days at the end of 2019. For retail sites, a 0.1 second improvement across four speed metrics came with an 8.4% higher conversion rate and 9.2% higher spending.
The same summary reports that progression from product page to add to basket rose 9.1%. That matches what most store owners see: the product page is where speed and buying intent meet.
ConversionNest's view: the study is useful direction, not a promise. It was commissioned by a company that benefits from faster web pages, it covered large brands, and the summary does not discuss how much of the change speed alone caused. Use it to justify checking your speed, not to forecast your revenue.
Which pages to check first
Not every page deserves the same attention. Check the pages where shoppers make decisions, and always check the mobile result first.
Page | What slow looks like | Why it matters |
|---|---|---|
Collection page | Products appear one by one, filters lag | Shoppers browse fewer products |
Product page | Main image loads late, the page jumps | This is where shoppers decide to add to cart |
Cart | Updates lag after changing quantity | Doubt grows just before payment |
Homepage | Slideshow loads before anything else | Matters most for direct and brand traffic |
If your ads send people straight to product pages, the homepage matters less than you might think. Test the pages your traffic actually lands on.
How to read a PageSpeed Insights report
PageSpeed Insights shows two kinds of data. Field data comes from real Chrome users through the Chrome User Experience Report (CrUX). Lab data comes from a single simulated page load on a fixed device and network.
web.dev's guide to lab and field data explains that CrUX covers a rolling 28-day period and reports the 75th percentile, and recommends using field data to decide where to focus. The two can disagree: in its example, the same page showed a 1.8 second Largest Contentful Paint in the field and 3.0 seconds in the lab. Both are valid.
Read the report in this order:
- Field data first: if it shows "Failed" for Core Web Vitals on mobile, real shoppers are having a slow experience.
- No field data: smaller stores often have too few Chrome visits for CrUX. Then use the lab result as a rough guide and test on your own phone.
- Lab diagnostics second: the opportunities list names the files and scripts slowing the page, which tells you what to fix.
The Core Web Vitals thresholds for "good" are a Largest Contentful Paint of 2.5 seconds or less, an Interaction to Next Paint of 200 milliseconds or less and a Cumulative Layout Shift of 0.1 or less, as web.dev documents.
What usually slows a small online store down
Shopify's guide to improving online store performance names three main factors: the theme, installed apps, and third-party code added manually, including tag managers and their tags. The same causes apply on most platforms.

- Apps: each app can add scripts to every page. Shopify advises removing apps that are not essential for sales, and notes that uninstalling an app can leave its code in your theme.
- Tracking tags: old pixels, chat widgets and testing tools pile up in the tag manager. Remove tags nobody uses.
- Theme sections and effects: long page templates, slideshows and animations add weight. Shopify suggests limiting sections and comparing performance with animations on and off.
- Large collections: turn on pagination so fewer products load at once.
- Images: size images to the space they fill. A product photo uploaded at camera size is several times heavier than it needs to be.
A worked example: what a speed fix could be worth
Consider a hypothetical store with 20,000 mobile sessions a month, a 1.5% conversion rate and a €50 average order value. These figures are hypothetical, not a client result.
| Conversion rate | Orders a month | Revenue a month |
|---|---|---|---|
Today | 1.5% | 300 | €15,000 |
After a speed fix | 1.6% | 320 | €16,000 |
A rise from 1.5% to 1.6% is a modest assumption, well below the study's figures. It still adds €1,000 a month, or €12,000 a year, from traffic the store already has. If removing three unused apps and resizing images takes a developer a day, the fix pays for itself quickly. If your store has far less traffic, the same logic tells you speed may not be the first thing to fix.
Fix it in this order
A speed clean-up for a small online store
- Test the pages that matterRun your top product page, a collection page and the cart through PageSpeed Insights on mobile. Note field and lab results. Allow 20 minutes.
- Remove what you do not useList every app and tag. Remove the ones that do not earn sales or answer a question you still ask, then check the theme for leftover code. Allow one to two hours.
- Fix the heaviest imagesResize hero and product images to their display size and compress them. Allow one to two hours, depending on catalogue size.
- Simplify the heavy templatesCut slideshows, autoplay video and extra sections from the homepage and product template. Allow an hour.
- Test again and watch the field dataLab results change at once. Field data takes up to 28 days to reflect the change, so check it again in a month.
When speed is not the problem
If your pages pass Core Web Vitals and sales are still weak, look elsewhere. Our guide to product page views with no add to cart covers the most common product page leaks, and our diagnosis of shipping costs and checkout abandonment covers the last step. For a full review, work through the e-commerce website checklist.
Before you spend more on traffic, check that your product page loads and responds quickly on a phone, using real visitor data where you have it.

Sima Damar Beygu
Founder of ConversionNest. 10+ years in growth marketing, managing 300K euro monthly media budgets and scaling acquisition across 15+ markets. MSc in Digital Marketing from University College Dublin. Google and Meta certified.
Sima on LinkedIn
Does your homepage say what you do in five seconds?
New visitors decide fast whether your homepage is worth their time. A simple five-second test shows whether yours explains what you do, for whom, and what to do next.

What is conversion rate optimisation, and where should you start?
A worked CRO example takes you from a conversion-rate calculation to a test brief, with practical checks for measurement, customer evidence and commercial value.

Website UX audit: what to check before a redesign
A UX audit should follow important visitor tasks across the website. Use reproducible findings to decide which problems deserve a fix, research or a test.