Cart abandonment and speed: the link nobody talks about
There is a whole industry of advice about cart abandonment — emails, popups, discount codes. Almost none of it mentions the cause that quietly makes every other cause worse.
Roughly seven out of ten online carts are abandoned. That figure has been remarkably stable for years across the studies Baymard Institute aggregates, and it is stable because the underlying reasons are mostly structural rather than fixable by a popup.
When shoppers are asked why they left, the reasons they give are consistent: unexpected shipping costs, being forced to create an account, a checkout that felt too long, not trusting the site with card details, and — some way down the list — the site being slow or erroring.
Here is the argument of this article, and it is a slightly different claim from the usual one: speed is not the top reason people abandon, but it is the reason that makes every other reason worse. That is why it deserves more attention than it gets, and why it is the cheapest lever you have.
Why speed amplifies everything else
Consider a shopper who is mildly annoyed about a shipping cost. On a fast site they click back, look at a competitor's page, come back, decide the total is acceptable, and buy. The friction was real but recoverable.
Now make every one of those steps take four seconds. The same mild annoyance now has thirty seconds of dead time attached in which to become a decision to leave. Nobody in that scenario will tell you they abandoned because of speed — they will say the shipping was too expensive. But they would have bought on the faster site.
This is what makes speed hard to see in survey data and easy to see in analytics. It is not a cause people report. It is a multiplier on causes they do report.
The economics are not close
Set the two standard responses side by side.
Making the site faster
- Helps every visitor on every visit
- One-off work, permanent effect
- Costs no margin per order
- Improves Core Web Vitals and search visibility as a side effect
- Works on the shoppers who never gave you an email address
Abandonment discount emails
- Reach only people who got as far as identifying themselves
- Cost margin on every order they recover
- Teach repeat customers to abandon deliberately
- Require ongoing campaign management
- Recover a fraction of a fraction
Recovery emails are not a bad tactic. They are just a tactic applied after the loss, to the small subset of losses you can still contact, at a cost per recovery. Speed is applied before the loss, to everyone.
Where the time actually goes in a shop
Checkout is frequently the slowest part of a shop, and for a specific reason: it is the one page nobody caches, and it is where every third-party script converges. Payment providers, fraud scoring, address autocomplete, analytics, chat widgets, trust badges — each one an external request on the page where patience is thinnest.
Three things are worth measuring in order:
- The pages leading into checkout. Category and product pages are cacheable and should be fast from anywhere. If they are not, fix that first — it is the easiest win and it affects the most people.
- The transition into cart and checkout. Uncacheable by necessity, so this is about origin speed and payload. Count the third-party requests on your checkout page; the number is usually higher than anyone expects.
- The distance for your actual customers. A checkout that responds in 200ms from your office may take 900ms from the country generating a third of your orders.
Time the whole journey rather than the homepage. Load product, add to cart, open checkout, and measure each step from two or three countries you actually sell to. The slowest step is where the money leaks, and it is almost never the page you would have guessed.
The honest caveat
Speed will not fix a checkout that demands account creation, hides shipping costs until the final step, or asks for information you do not need. Those are the reasons shoppers actually cite, and no amount of milliseconds compensates for them.
So the sensible order is: fix the structural things people tell you about, and fix speed at the same time, because it is the one that quietly determines how much patience they have for everything else. Doing only the second is optimising the wrong thing. Doing only the first leaves a multiplier in place.
Common questions
What is the average cart abandonment rate?
Around 70% across aggregated studies, a figure that has stayed broadly stable for years. It is high because much of it is structural — people using carts as wishlists, comparing totals across sites, or researching without intent to buy immediately. A meaningful portion, however, is recoverable friction: unexpected costs, forced account creation, overlong checkouts and slow pages.
Does site speed affect cart abandonment?
Yes, though less as a directly cited reason than as an amplifier of other reasons. Shoppers rarely report leaving because a site was slow; they report the shipping cost or the long checkout. But slowness extends every moment of hesitation, giving second thoughts more time to arrive, so the same friction converts on a fast site and abandons on a slow one.
Why is my checkout page slower than the rest of my shop?
Because it cannot be cached and it carries the most third-party code. Checkout is generated fresh per shopper, so no edge cache can serve it, and it typically loads payment providers, fraud scoring, address autocomplete, analytics and chat widgets. Counting the external requests on your checkout page is usually a revealing exercise.
Is it better to fix site speed or send abandonment emails?
Speed first, because it applies to every visitor rather than only to those who identified themselves, costs no margin per order, and does not train customers to abandon deliberately in expectation of a discount. Recovery emails are a reasonable supplement, but they act after the loss on a small subset of it, whereas speed acts before the loss on all of it.
None of which shows up in a survey answer. It shows up as a conversion rate that is a bit lower than it should be, month after month, for a reason nobody can point at.
See how NordicCDN does this for your site:
Mads has worked in IT — mostly hosting — since he was 16. He took an early stake in a SaaS company and helped grow it through to its acquisition by Visma, has built and run data-center networks, and served as CTO of a Danish data center. He started NordicCDN to make fast, secure infrastructure simple to use.
