Customer experience in ecommerce is everything a customer goes through with an online shop, from the search to the refund. Most of it happens after the checkout, out of sight of the website team: the delivery slot, the doorstep, the box, the return label and the money coming back. In 2025, 35.4% of online shoppers in the EU reported a problem with a purchase, and the most common one was slower delivery than expected, at 19.9%. A website that was hard to use came a distant second at 11.5%. This article maps where ecommerce CX breaks, using regulator and ombudsman data from four sources, and names the five moments after the checkout that decide the next order.
Key takeaways:
- The product and the website are rarely the problem. Slow delivery, parcels left in the wrong place, damaged goods and unclear returns top every complaint list from Brussels to London.
- Satisfaction scores stay high while problems climb. In the UK, 37% of people had a problem with their last parcel for the fifth year running, and Ofcom's average satisfaction with parcel firms slipped to 75% in 2026 only after three years at 78%.
- Returns are routine, not exceptional: Dutch consumers send back one parcel in eight, and a third of webshops now charge for it. A return handled well earns more loyalty than a first order that simply arrived.
- Product reviews and CX feedback are different things. Reviews help the next buyer and are now regulated as such. Feedback helps you fix the cause.
- In one Belgian retailer's data on our platform, the webshop scores an NPS of +81 and the parcel drives it: packaging and condition, delivery accuracy and delivery speed carry the biggest potential gain.
What customer experience in ecommerce means
Customer experience in ecommerce is the impression a customer forms of an online shop across every contact, and most of those contacts are physical. The customer touches the website for minutes and the parcel for days. In the Netherlands, 447 million parcels crossed the border in 2025, up 7.4%, and 82.2% of all parcels were delivered at home, with 12.4% collected at a service point and the number of parcel lockers almost doubling to 9,309. In France, FEVAD reports 196.4 billion euros of online sales in 2025, 3.2 billion transactions, and out-of-home delivery now close to one delivery in two. Belgians spent 18.3 billion euros online, up 5.4%, and more than nine in ten bought online at least once.
The shop controls the checkout and outsources the rest, which is why the experience so often ends with a stranger at the door. Coolblue, which runs its own delivery and 38 stores, attributes its NPS of 72 to the quality and efficiency of its processes, "such as our first time right". Few shops will buy vans. Every shop can accept that the parcel is the product, and whoever carries it carries the brand.
Why ecommerce satisfaction numbers hide the risk
Ecommerce scores well in the satisfaction indexes. The American Customer Satisfaction Index puts online retailers at a stable 79, with Amazon, Nordstrom and Chewy tied at 82, and consumer shipping at 78 after gains in app quality, shipment tracking and delivery timeliness. The numbers that should worry a CX lead sit underneath the index. In the UK, Citizens Advice found that 37% of people had a problem with their last parcel in 2025, the fifth year in a row above 33%. Among them, 29% said the driver left before they could get to the door, 24% found the parcel in an insecure location and 24% got it late. The star ratings improved slightly, and the charity's own explanation is uncomfortable: the improvement came from fewer complaints raised on social media and to its consumer service, which "could suggest that people are beginning to simply expect poor parcel service".
Ofcom's 2026 survey says the same thing with a lag. Average satisfaction with parcel firms had held around 78% from 2023 to 2025 and slipped to 75% this year, while 71% of people receiving parcels had a delivery issue in the past six months, up from 68%. Delays were the most common issue at 32%, parcels left in inappropriate locations at 28%. Fergal Farragher, Ofcom's policy director, said people "rightly expect packages to arrive on time and without hassle" and that the regulator will consider further measures. Satisfaction moved three points. Problems moved from a minority to a majority. The gap is where the next lost customer sits.
The third reason the numbers mislead is that online and offline loyalty have different drivers, so a blended retail score hides both. One Belgian retailer with stores and a webshop, anonymised, scores a webshop NPS of +81 from 1,083 responses over the last 12 months on our platform, against +76 in its stores from 3,925. The key driver analysis shows two different businesses. In the stores, people drive loyalty: staff assistance and availability by far, then friendliness and expert advice. Online, the parcel drives it. Packaging and condition, delivery accuracy and delivery speed have the biggest potential gain, and delivery speed is the most mentioned topic. Returns, refund speed and website usability follow. Our guide to retail customer experience covers the store side. This article is about the parcel.
Where ecommerce CX breaks: a complaint map from four sources
Regulators, ombudsmen and consumer bodies publish what customers escalate about online shopping, and the lists agree.
| Source, year | What customers escalate |
|---|---|
| EU, Eurostat, 2025 | 35.4% of online shoppers had a problem. Slower delivery than expected 19.9%, website hard to use 11.5%, incorrect or damaged goods delivered 10.4%. Highest shares in Malta (64.0%), the Netherlands (57.9%) and Luxembourg (51.4%) |
| Belgium, postal and parcel ombudsman, 2025 | 4,390 mediation requests, 2,367 admissible files, 14% more files handled than in 2024. "My parcel is lost" remains the most common complaint, 808 of 2,523. The share of complaints about ecommerce parcels keeps rising, with unsafe delivery methods, untraceable shipments, no proof of handover and unclear return procedures as the weak points. The ombudsman calls out complaints being bounced between webshops and carriers |
| Netherlands, ACM, 2025 | 447 million cross-border parcels, up 7.4%, partly from webshops in Asia. Consumers returned 12.3% of parcels on average over two years, up from 11.4% in 2021, while a third of webshops charge for returns |
| UK, Citizens Advice and Ofcom, 2025 to 2026 | 37% had a problem with their last parcel; driver left too early 29%, insecure location 24%, late 24%. Ofcom: 71% had a delivery issue in six months, delays 32%, wrong location 28%. Evri and Yodel had the least satisfied customers for the fourth year running |
Sources: Eurostat; Ombudsman Post e annual report 2025; ACM post and parcel monitor 2025 and NOS on ACM return figures; Citizens Advice parcels league table 2025 and Ofcom 2026 via PA.
Two details matter for a CX lead. First, the customer does not separate the shop from the carrier. In Belgium, 76% of complainants at the postal ombudsman are recipients, and its 2025 report names the "shrugging off of consumer complaints between webshops and platforms and the carrier" as a bad practice, calling on both to make customer-focused agreements. The shop chose the carrier. The customer holds the shop responsible. Second, the complaints are about the mechanics: lost, late, left somewhere, returned without a trace. None of those shows up in a product review.
The five moments after the checkout, and what to measure at each
The checkout is where the website team stops measuring and the customer starts judging. Five moments follow. Each deserves its own short transactional survey close to the event, with one score and one open question.
1. The promise at checkout
The delivery promise is the first thing the customer will test. Slower delivery than expected is the top complaint in the EU, and "than expected" is the operative phrase. A four-day delivery promised in four days is fine. A three-day delivery promised for tomorrow is a complaint. Measure the promise against the scan: the share of orders delivered on or before the promised date, by carrier and region. Then ask one question at confirmation: is it clear when and how your order will arrive?
2. The doorstep
The doorstep is where most problems happen and where the shop is least present. A driver who left too early, a parcel in the recycling bin, a neighbour who was never told. In the UK, 24% of problem deliveries ended in an insecure location, and Ofcom found 28% of consumers with an issue had a parcel left in an inappropriate place. Measure within hours of the scan: was the parcel where you expected it, in the state you expected? Split by carrier, the answers are the most useful carrier scorecard a shop will own, because they come from the customer and not from the carrier's own data.
3. The unboxing
Packaging and condition is the driver with the biggest potential gain in the Belgian retailer's webshop data, ahead of delivery accuracy and speed. The customer who opens a crushed box has decided the shop does not care before the product is tested. Eurostat counts 10.4% of EU shoppers receiving incorrect or damaged goods. Measure it in the same survey as the doorstep, with one more question and a photo upload. The photos are evidence for the carrier conversation and training material for the warehouse.
4. The return
Returns are routine. One Dutch parcel in eight goes back, and the rate rose even as a third of webshops started charging for it. The Belgian ombudsman lists unclear return procedures among the weak points it sees in ecommerce files. A return has three steps of its own: finding out how, sending it, and knowing it arrived. Measure with an effort score after the return is registered, and read the comments for the word "label". Most return friction is a label that would not print, a drop-off point that was closed, or a tracking page that stopped updating. The best shops treat the return as a moment of truth. Coolblue replaces first, no questions asked, and then fixes the cause so it does not happen again.
5. The refund
The refund is where the return is judged, and it is judged in days. In the Belgian retailer's data, refund speed sits directly behind returns as a driver. A customer waiting for money is counting, and every day of silence is a day the competitor's site is open in another tab. Measure when the refund lands: one satisfaction question, one open field. Then compare the customers who returned something with those who did not, three months later. If the returners order more, the return process is working. If they order less, it is costing more than the shipping.
Reviews help the next buyer, feedback helps you
A product review and a piece of CX feedback look alike and do opposite jobs. The review is public and written for the next buyer: is this jacket true to size, does this kettle last. It is regulated as such. The UK's Competition and Markets Authority published its guidance on the ban on fake and concealed incentivised reviews in April 2025 under the Digital Markets, Competition and Consumers Act, and review platforms now have to take reasonable steps to keep fake content out. Reviews are evidence for shoppers, and the law treats them that way.
CX feedback is private and written for you: the parcel was late, the label would not print, the refund took nine days. It names the cause rather than the product, and it comes from everyone who went through the process, including the majority who never write a review. The Belgian retailer's most mentioned topic online is delivery speed, which no product review covers. The two sources belong side by side, never merged. Our guide to customer feedback analysis covers how to read both at volume; sentiment per topic is what lets "great jacket, terrible courier" count as two signals, and the second one is the one you can fix.
Why a smooth return earns more than a smooth first order
A first order that arrives on time is expected. A return handled well is remembered, and the memory outlasts the purchase. Michel Stevens described the mechanism on our CX Matters podcast, comparing two Dutch electronics retailers: one that owns its whole supply chain and positions itself on customer-friendliness, and a marketplace that owns little more than the warehouse and the platform. The marketplace scores higher on satisfaction, and his explanation is the service recovery paradox. "Call bol.com about a defective device or a missing product and they will say, without hesitation: 'You know what? Do not worry about it, we will take it from here. A new device is coming by express courier, you will have it tomorrow.' They over-deliver, and the paradox kicks in." Recovery, built into the business model, becomes the advantage.
He also told the story from the inside. Running a customer service department before ecommerce existed, he fought finance to ship a replacement the moment the returned parcel was scanned at the post office, because "that customer will use either our product or our competitor's". Finance feared bricks in the boxes. A few arrived. For the vast majority, the shop did the right thing and kept the customer. His caveat matters as much as the story: recovery works on the functional and emotional level, and the paradox is only complete when the customer sees what the shop learned. Replace first, then fix the cause, then say so. Most shops skip the last step; our guide to closing feedback loops shows what it looks like in practice, and our guide to customer retention strategies puts a number on what a kept customer is worth.
How to measure ecommerce CX so it changes something
Ecommerce CX measurement works in three layers, and most shops over-invest in the top one. The relationship score, a few times a year, says the brand is liked. The transactional layer, a short survey after each of the five moments above, says which carrier, which region and which step is losing customers this week. The open text underneath, from surveys, chats, emails to customer service and the reviews on the carrier's app, says why. Shops that measure only at order confirmation measure the website, and the website is rarely the problem.
Three capabilities carry most of the weight. Key driver analysis ranks which topics move loyalty, so delivery speed, packaging or refund time show up as drivers with a weight rather than a complaint in the Monday meeting. Forward-looking alerts catch a sentiment shift on one carrier or one depot before it becomes a week of chase emails. Customer interactions bring the chats and emails about "where is my order" into the same view as the surveys. That is how Hello Customer is set up for retailers that sell in stores and online and need each channel measured on its own drivers. The scoreboard matters because 48% of companies never calculate the financial impact of CX, as our 2026 study found. In ecommerce the calculation is short: reorder rate of customers with a problem, by whether the problem was fixed.
If you are choosing software, our comparisons of feedback software for ecommerce and feedback software for retail set out the criteria. Telecom operators face a close cousin of this problem in the technician visit; our guide to customer experience in telecommunications applies the same method there, and the banking guide does it for the fraud alert. The series also covers customer experience in insurance and customer experience in utilities.
A 90-day start for an ecommerce CX lead
Start with the delivery and returns line, where the complaints are, and run it for a quarter.
- Map the five moments in the customer's words: "when will it come", "where did they leave it", "the box was crushed", "how do I send it back", "where is my money".
- Send one question after the doorstep scan and one after the refund, each with an open field, and an effort question when a return is registered. Keep every survey under a minute.
- Measure the promise against the scan for four weeks: on-time share by carrier and region. That is the baseline.
- Read every open answer, including the chats about "where is my order". Tag the words that recur: "late", "neighbour", "label", "refund".
- Rank the drivers and fix the top one. In most shops it is a carrier, a depot, a return label or a refund delay rather than anything on the website.
- Replace first, then fix the cause, then tell the customers who complained what changed. The service recovery paradox only pays out when the customer sees the learning.
- Report two numbers at day 90: the on-time share and the reorder rate of customers who had a problem against those who did not.
Then add the second carrier, the marketplace orders and the store pickups, each with its own survey and its own owner.
Frequently asked questions
What is customer experience in ecommerce?
Customer experience in ecommerce is the overall impression a customer forms of an online shop across every contact, from the search and the checkout to the delivery, the return and the refund. Most of it happens after the checkout. Eurostat found that 35.4% of EU online shoppers had a problem in 2025, and the most common one was slower delivery than expected, well ahead of problems with the website itself.
Why are delivery and returns the main source of negative ecommerce feedback?
Because they are the physical part of the experience and the part the shop outsources. Regulators in the EU, Belgium, the Netherlands and the UK all report late deliveries, parcels left in the wrong place, damaged goods and unclear return procedures as the main complaints. The customer holds the shop responsible for the carrier it chose, and the Belgian postal ombudsman specifically criticises complaints being passed between webshops and carriers.
What is the difference between product reviews and customer feedback?
A product review is public and written for the next buyer; it describes the product. Customer feedback is private and written for the shop; it describes the process, from the delivery promise to the refund. Fake and concealed incentivised reviews are banned in the UK under the Digital Markets, Competition and Consumers Act, with guidance published in April 2025. Reviews help the next shopper choose. Feedback tells you what to fix.
How do online shops measure customer experience after the checkout?
With a short transactional survey after each of five moments: the delivery promise at checkout, the doorstep, the unboxing, the return and the refund. Each has one score and one open question. Key driver analysis then ranks which topics, such as delivery speed, packaging or refund time, move loyalty most, so the shop fixes the heaviest driver first and tells customers what changed.
Final thoughts
Online shops lose customers at a parcel left in the rain, a box that arrived crushed, a return label that would not print and a refund that took a week longer than the order. The website worked the whole time. The complaint maps from Eurostat, the Belgian ombudsman, the Dutch regulator and the British watchdog agree: the experience is decided after the checkout, by people the shop does not employ. Measure the five moments on their own, keep reviews and feedback apart, and handle the return as the sale it is. The parcel is the product. Treat it that way and the second order follows.
To see how a retailer measures the webshop and the stores on their own drivers, book a demo and we will walk through it with your own carriers and return flows.
Bram De Vos