In Portugal, 4.5% of online buyers reported problems in 2025, the lowest value among European Union countries. In the EU, 35.4% said they had found at least one problem when buying on a website or in an application. Eurostat survey It covers people aged 16 to 74 who bought online in the three months prior to the interview.
Who runs a store can use the inquiry to choose what to investigate in their own orders. The slower delivery than expected was the most common complaint in the EU as a whole. For the team, this puts a concrete question on the table. On what orders did the customer miss the deadline before paying?
See how an online purchase begins in Portugal for the steps shoppers take before payment.
How to read Portugal’s 4.5%
The denominator is the number of people who bought online during the reference period. One person may have made several purchases and reported a difficulty in one of them. Therefore, the 4.5% represent the proportion of buyers who reported problems. To calculate an order rate, the order data and the problems associated with each one would be required.
The indicator also follows the residents’ residence. A purchase of a resident in Portugal may have been made to a foreign seller. And the aggregate series includes purchases of goods and services. A Portuguese clothing store may have problems different from a website that sells tickets or a furniture retailer. series published by Eurostat helps to understand the basis of comparison.
The delivery date on the order confirmation
In all EU buyers, 19.9% reported slower deliveries than they expected. The difficulty in using the site or malfunction was reported by 11.5%. Incorrect or damaged articles or services appeared in third place, with 10.4%. These figures describe problems that may occur in the same purchase and should be read individually. Data on online shopping problems, Eurostat
To measure the delays in a store, keep the forecast presented in the purchase and the actual date of delivery, when available. An order from Monday promised until Thursday and delivered on Friday arrived outside the deadline. Another made on Friday, promised until Wednesday and delivered on that day, fulfilled the promise although it may have taken longer calendar days. The average duration explains how long the delivery takes. Compliance with the deadline says if the customer received what was promised.
The forecast also depends on the time when the store delivers the box to the carrier. If orders left the warehouse late in a week of discounts, the first step is to review the preparedness capacity. If they left on time and arrived late to a set of postal codes, the team has a different chance to investigate with the carrier. When the delay is known, it is better to warn the customer with a new forecast and a contact than to let him interpret a generic tracking state.
Find where the site fails
A website can work on the team’s computer and fail on the ad’s mobile phone. First test the route that concentrates complaints. Open the announced variant, try to calculate the delivery and confirm that the form accepts a valid postal code. Then test it with an island address if that destination is covered. An error message that says only “invalid address” does not help when the store actually deletes the region. What to review before launching an online store
For items exchanged or damaged, the best starting point is to associate the occurrence to the order. If the same reference accumulates size changes, compare the form with the label, internal code and list that the warehouse uses to separate the orders. If the damage is concentrated on a fragile product, compare the packaging and route. Keeping a specific reason allows distinguishing the occurrences without depending on a broad category such as ‘returns’.
What to measure during a campaign
A weekly panel can add shipped orders, percentage of deliveries within the forecast and support requests for delays for every 100 orders. For the site, follow the payment errors by method and the number of orders paid. Use comparable periods and register changes as a new carrier or a peak of sales. With orders still in transit, show aside the shipments whose deadline has already expired.
A hypothetical example illustrates the difference: if complaints about delays go from eight to twelve and the volume of orders doubles from 200 to 400, the proportion drops from four to three per 100 orders. There are more requests for the team to respond and fewer complaints per order. They are two useful data for different decisions. The purchase and delivery processes can thus be evaluated with numbers that the store itself can verify.