5 Things Customers Expect From Your eCommerce Store
- Baymard Institute puts the documented average online cart abandonment rate at 70.22 percent, measured across 50 separate studies.
- Extra costs at checkout (shipping, tax and fees) are the single largest fixable reason for abandonment, cited by 40 percent of US shoppers who abandoned for a reason other than browsing.
- Google's mobile speed research found that 53 percent of mobile site visits are abandoned when a page takes more than three seconds to load.
- Delivery speed and the returns policy are now purchase conditions read before the order is placed, cited by 20 percent and 13 percent of non-browsing cart abandoners respectively.
- Social share buttons have been replaced by social proof: shoppers expect ratings, written reviews and brand replies they can check on the product page.
Customers arrive at an online store with a short list of conditions. They rarely say them out loud. They simply leave when one of them is not met, and the store owner sees a bounce, an abandoned cart, or a first-time buyer who never comes back.
The useful thing about these conditions is that they are documented. Baymard Institute has spent years asking US shoppers why they abandoned a checkout, and the answers are consistent enough to plan against. Getting them right pays: in a 2019 roundup of customer experience research for Forbes, Blake Morgan reported that 84 percent of companies that work to improve their customer experience see an increase in revenue. This article covers the five expectations that matter most, what each one costs when you miss it, and how to find out which of them is hurting your store in particular.
What do customers expect from an eCommerce store?
Roughly four in ten shoppers who abandon a cart were only browsing, and there is little you can do about them. The rest left for a reason you can name and fix. Baymard's cart abandonment research ranks those reasons among US shoppers who abandoned for a reason other than browsing:
| Reason for abandoning | Share of abandoners | Expectation it breaks |
|---|---|---|
| Extra costs too high (shipping, tax, fees) | 40% | The full price up front |
| Delivery was too slow | 20% | Shipping they can live with |
| Did not trust the site with card details | 19% | A payment method they trust |
| Forced to create an account | 18% | A checkout that does not get in the way |
| Checkout too long or complicated | 17% | A checkout that does not get in the way |
| Site errors or crashes | 17% | A site that works |
| Unsatisfactory returns policy | 13% | A way out if it goes wrong |
| Could not see total cost up front | 12% | The full price up front |
| Not enough payment methods | 9% | A payment method they trust |
Read down that list and the five expectations below stop looking like opinions. They are the store failing at things the customer took for granted.
1. A site that loads fast and does not break
Page speed is not only a conversion problem, it is a measurement problem. Most stores test speed on a desk, on office wifi, on a laptop. Customers arrive on a mid-range phone on a mobile network, often from an ad, and the difference between those two conditions is usually several seconds.
Google's mobile page speed research is the figure most often quoted here: 53 percent of mobile visits are abandoned when a page takes more than three seconds. Treat three seconds as a ceiling rather than a target, and measure with PageSpeed Insights, which reports both a lab score and Core Web Vitals collected from real visitors. When the two disagree, believe the field data. Pingdom Tools is a useful second check that shows which individual files slow the page down.
What actually moves the number, in rough order of return:
- Images. Serve modern formats, size them for the slot they occupy, and lazy-load anything below the fold. On most stores this is the single biggest win.
- Third-party scripts. Chat widgets, analytics, heatmaps, review widgets, ad pixels. Audit what is loading and remove what nobody looks at.
- Hosting and CDN. Serve assets from a network close to your customers instead of one data center.
- Caching. Repeat visitors should not download the whole page again.
- Code weight. Minify CSS and JavaScript, and defer anything not needed for the first screen.
Stability matters as much as speed. A checkout that fails intermittently on one browser is invisible in an average page-load report and very visible to the 17 percent who hit it.
2. Shipping and delivery they can live with
Free shipping is usually discussed as a promotion, a hook to catch a browsing visitor. That is only half of it. Shipping is a cost the customer is quietly adding up while they shop, and the damage happens at the moment the real number appears. A $4.95 delivery charge shown on the product page is a fact. The same charge revealed on step three of checkout is a betrayal.
Delivery speed has become the second half of the same expectation, and it has moved fastest of the five. Large marketplaces normalized next-day and two-day delivery, and one in five shoppers who abandon for a stated reason now names slow delivery. You do not have to match a marketplace. You do have to say what will happen and when, in plain terms, on the product page.
Practical ways to meet this without giving away margin:
-
Show the cost before the cartPut the shipping cost, or the threshold that removes it, on the product page. If it depends on location, use the visitor's country to show the right number rather than a range.
-
Show a date, not a duration"Arrives Tuesday 15 September" is a commitment a customer can plan around. "3-5 working days" makes them do arithmetic and doubt the answer.
-
Offer a fast paid option next to a slower free oneThis converts the impatient without subsidizing everyone. It also tells you, from the split, how much your customers actually value speed.
-
Set the free-shipping threshold from your dataSet it just above your average order value so it lifts basket size, not below it, where it only costs you margin on orders you would have won anyway.
-
Keep promotions honestDiscounts and deadlines still work, because people work harder to avoid losing something than to gain it. They stop working the moment a customer notices the same "ends tonight" banner three weeks running.
Test a shipping threshold, a delivery message or a promotion before you commit to it store-wide.
See how A/B testing works in Omniconvert Explore →3. Help from a real person when something goes wrong
Two moments decide whether support helps or hurts. The first is pre-purchase, when a shopper has a question the product page did not answer: sizing, compatibility, ingredients, whether it will arrive before a birthday. If that question is unanswered, the order does not happen and you never learn why. The second is post-purchase, when something has already gone wrong. That is where a customer decides whether there will be a next order.
What customers expect, concretely:
- Contact details that are easy to find. If the only route to a human is a chat widget that loops back to an article, you have hidden your support desk on purpose and customers can tell.
- An honest response time. "We reply within one working day" beats a chat bubble that says "typically replies instantly" and then does not.
- A FAQ that answers real questions. Build it from your actual support tickets and return reasons, not from what you wish people asked.
- Recognition. A repeat customer should not have to re-explain their order history. This is the point where customer data stops being a reporting exercise and becomes service.
- An answer, not a deflection. Automation is fine for order status. It is not fine for a complaint.
Service quality is also the cheapest source of product insight you have. Support tickets and return reasons tell you which product pages are lying, which sizes run small, and which delivery promise your carrier cannot keep. Feeding that back into the site is usually worth more than the tickets themselves. If you want the wider picture, see our guides to consumer behavior patterns and eCommerce customer experience.
4. Social proof they can check for themselves
This is the expectation that has changed most. In the mid-2010s the advice was to put Facebook, Pinterest and Twitter buttons on every product page so customers could spread the word. That advice has aged badly. Sharing still happens, constantly, but it happens by pasting a link into a group chat or a story, not by clicking a widget. Some of it now ends inside the social platform itself, a shift Floship tracks in its overview of social commerce trends. On the product page, what has taken the place of share buttons is proof.
Proof the customer can check means:
- Reviews with volume and range. Twelve reviews all at five stars reads as suspicious. A hundred reviews averaging 4.4, including some two-star ones, reads as real.
- Critical reviews left visible. Removing them removes the reason anyone believes the good ones.
- Brand replies to the bad reviews. A reply that explains a misuse or offers a fix is read by every later shopper, not only the complainant.
- Photos and video from customers. In apparel, home and beauty this does more work than the studio photography above it.
- Proof placed where the doubt is. A review about sizing belongs next to the size selector, not at the bottom of the page.
Reviews are also a feedback instrument, not just a conversion device. A recurring complaint in your reviews is a product or copy problem you can fix, and the fix is usually worth more than the review widget. If you also measure sentiment formally, our guide to the Net Promoter Score scale covers how to score and act on it.
5. The full price up front, and a payment method they trust
Customers are not only price-sensitive, they are surprise-sensitive. A total that grows between the product page and the payment step is read as a trick, even when every individual charge is reasonable. The fix is not always a lower price. It is an earlier one.
Three things to get right:
- Total cost, early. Show shipping, tax and any handling fee before the customer invests effort in a checkout. If tax depends on location, estimate it from the visitor's country and label it as an estimate.
- Payment methods they already use. Card logos, a digital wallet such as Apple Pay or Google Pay, and whatever dominates locally in your main markets. Show them on the product page and in the cart, not only on the payment step.
- A returns policy stated as a condition of sale. How long, who pays for return shipping, and how the refund arrives. This is what a shopper is really asking when they hesitate over an unfamiliar brand, and it is why the policy belongs near the buy button.
Discounts belong in this section too, and they behave differently from what most stores assume. A discount shown clearly against a real reference price supports the decision. A permanent "50% off" that never expires teaches customers that your list price is fiction, which makes every future full-price sale harder. Our guide to checkout optimization goes further into the mechanics of the last step.
What has changed about these expectations
It is worth being explicit about this, because advice written for an earlier internet is still circulating.
What has not changed is the underlying rule: customers expect a store to be predictable. Every item on this list is a form of predictability, and every failure on it is a surprise the customer did not want.
How to find out which expectation is costing you
The table below is the diagnostic we use with eCommerce teams. It is deliberately qualitative: the point is to match a symptom you can see in your own data to the expectation behind it, not to compare yourself with an average.
| Expectation | Signal that you are missing it | What to do first |
|---|---|---|
| A site that works | Mobile conversion far below desktop; high exits on the first page from paid traffic | Measure Core Web Vitals on field data, then cut third-party scripts and image weight |
| Shipping they can live with | Drop-off concentrated between cart and the shipping step | Move the shipping cost and a delivery date onto the product page |
| Help from a real person | Repeat purchase rate falls for customers who contacted support | Read the last 100 tickets and fix the three most repeated causes on-site |
| Social proof they can check | Long product-page dwell time with low add-to-cart; questions in support that reviews should answer | Collect reviews with photos and place them next to the doubt they resolve |
| The full price up front | Abandonment spikes on the step where shipping or tax first appears | Show the estimated total earlier and display payment options before checkout |
Two sources of evidence make this concrete. The first is what customers say. An on-site survey triggered on exit, and a post-purchase question asking what almost stopped the order, will name the failing expectation faster than any analytics report. The second is what customers do. Comparing repeat buyers with one-time buyers shows which experience produced a second order. RFM segmentation is the standard way to draw that line.
Nexus by Omniconvert builds those segments from your customer data, so you can see whether your best customers came through the fast checkout or the slow one, bought at full price or on discount, and what the ones who never returned had in common. Nexus draws on 7,000+ websites across 15+ industries, 248+ audit criteria and 13 years of data.
Frequently Asked Questions
Customers expect five things from an online store: a site that loads fast and does not break, shipping and delivery terms they can live with, help from a real person when something goes wrong, social proof they can check for themselves, and the full price shown before they reach checkout. These are not preferences. Baymard Institute's checkout research shows that unexpected extra costs, slow delivery, payment worries, a poor returns policy and site errors are among the most common reasons shoppers abandon a cart.
Aim for a page that becomes usable in under three seconds on a mobile connection. Google's mobile speed research found that 53 percent of mobile site visits are abandoned if a page takes longer than three seconds to load. Measure with PageSpeed Insights and track Core Web Vitals on real visitor data rather than lab scores alone, because the lab number often flatters a site that is slow for actual customers.
Baymard Institute puts the documented average cart abandonment rate at 70.22 percent across 50 studies. Among shoppers who abandoned for a reason other than browsing, 40 percent left because extra costs such as shipping, tax and fees were too high, 20 percent because delivery was too slow, 19 percent because they did not trust the site with their card details, 18 percent because they were forced to create an account, 17 percent because of site errors or crashes, and 13 percent because of an unsatisfactory returns policy.
They expect shipping cost and delivery date to be predictable, and they treat surprise fees at checkout as a reason to leave. Free shipping is the simplest way to meet that expectation, but it is not the only one. A clearly stated flat rate, a free-shipping threshold shown on the product page, or a paid express option next to a free slower one all work, provided the cost appears before the checkout step rather than inside it.
It is a purchase condition, not an after-sales detail. Baymard's research lists an unsatisfactory returns policy among the reasons US shoppers abandon carts, cited by 13 percent of those who abandoned for a reason other than browsing. Shoppers read the returns terms before they buy, especially in apparel and other categories where fit is uncertain, so the policy belongs on the product page and not only in the footer.
Offer the methods your customers already use, and show them before checkout. Baymard's data shows 19 percent of non-browsing cart abandoners left because they did not trust the site with their credit card information and 9 percent because there were not enough payment methods. Card logos, a wallet option such as Apple Pay or Google Pay, and any local method that dominates your main markets cover most of that gap.
It affects the second purchase more than the first. A shopper who cannot get an answer before buying usually leaves without telling you, and a customer whose problem is handled badly after buying rarely returns. Because repeat customers carry most of the profit in eCommerce, the cost of poor service shows up in customer lifetime value long before it shows up in conversion rate.
Ask them and watch them. Run an on-site survey at the point where people leave, add a post-purchase question about what almost stopped them from ordering, read the reasons behind returns and support tickets, and compare the behavior of repeat customers with one-time buyers. Nexus by Omniconvert segments customers by recency, frequency and monetary value so you can see which expectations your best customers care about, and Omniconvert Explore lets you test the fixes before rolling them out.
Open your own store on a phone, on a normal mobile connection, and buy something. Time how long the product page takes to become usable. Note the exact moment the shipping cost appears, and whether the delivery date is a date or a vague promise. Try to find the returns policy without using the footer. Then try to reach a human. Most stores fail at least two of those tests, and the failures are cheap to fix compared with the traffic they waste. Fix the ones you find, then use survey answers and customer segments to decide which of the five expectations is costing you the most, rather than guessing.
See which expectations your best customers care about
Nexus by Omniconvert segments your customers by recency, frequency and monetary value, so you can compare what repeat buyers experience with what one-time buyers experience, and fix the gap that costs you the most lifetime value.