Quick answer: Four online sellers tested a faster page against a slower one, and each one saw a measurable lift. The size varied a lot, from 1% to more than 50%, and these are large companies. The lesson for a small store is what to test and how to test it, not the size of the numbers.
We kept only tests whose results the company published on web.dev, Google’s site for web developers. Below: the four tests, what they share, and how to run your own.
What did the four tests find?
| Company | What it sells | What it changed | Result in the A/B test | Page date |
|---|---|---|---|---|
| Rakuten 24 | Daily necessities | Faster landing page, nothing else | 53.37% more revenue per visitor, 33.13% higher conversion rate | 2022 |
| Vodafone | Telecom plans | Landing page with 31% better load speed | 8% more sales | 2021 |
| Farfetch | Luxury fashion | Product pages more than 600 ms faster | 1% to 5% higher conversion rate | 2022 |
| Trendyol | General ecommerce | Listing pages that respond 50% faster to taps and clicks | 1% more clicks from listing pages through to product pages | 2023 |
What can a small store learn from each one?
Rakuten 24: change one thing and split the traffic evenly
Rakuten 24 tested one landing page for a month. Half the visitors got the faster version and half got the original, and speed was the only difference. That clean setup is why the result means something.
Vodafone: watch the steps, not just the sale
Vodafone also saw 15% more leads per visit and 11% more cart activity per visit. Those earlier steps show where the gain came from, and they build up faster than sales do.
Farfetch: expect a range, not one number
Farfetch reported its lift as 1% to 5%. That is an honest way to report a test. Be wary of any single number that sounds too good, including one from your own test.
Trendyol: small lifts still count
Trendyol cut its response delay in half and gained 1% on clicks from listing pages through to product pages. A 1% lift on a page people use all day is worth having, and small gains add up.
Do these results apply to a small store?
The method does. The percentages may not. Rakuten 24 says it picked a page with significant traffic so that the result would be meaningful, and most small stores do not have that traffic. They are also older tests of large sites, and they may not repeat on yours.
What holds up: all four tested a real change on a real page against the old version at the same time. All four measured money or clicks, not opinions. In our view, that is the part to copy.
How do you run a small version of this?
- Pick your busiest product or landing page. Test where the traffic already is.
- Measure it first. Our guide to seven hosting speed checks shows how to find what is slow.
- Change one thing. Compress the images, or remove a slow plugin.
- Split the traffic if you can. If you have too few visitors, compare two similar periods instead, and treat the result as a clue rather than proof.
- Choose your measure before you start. Our guide to A/B test metrics for online stores covers which ones to use.
- Write down the result, whatever it is. A test that changes nothing still tells you where not to spend time.
Where does your hosting fit in?
All four tests are about speed, and your hosting sets the floor for it. If your store runs on WooCommerce, Turbify WooCommerce hosting includes a free domain, free SSL, Business Mail accounts and backup and restore tools. Backups matter here too, because you can undo a test that breaks something.
Quick answers
What is an A/B test?
You show two versions of a page to different visitors at the same time and compare what they do. The comparison is fair because both groups visit in the same period.
How long should an A/B test run?
Rakuten 24 ran its test for a month. We would plan for at least that long on a small store, so that weekdays, weekends and paydays are all included.
Why are all four tests about speed?
Because these are the ones with published numbers we could open and check at the source. Many other case studies come from companies selling testing tools, and we left out what we could not verify.
Your next step
Open your busiest product page on your phone and time how long it takes to appear. If it feels slow to you, that is your first test.
Launch your online store on Turbify, with secure checkout, inventory tools and room to grow. See WooCommerce hosting.
