Answer 46 of 50
Building Trust Through Design in Ecommerce Sites
In Baymard Institute’s survey of US online shoppers, 19% of those who abandoned a checkout said they didn’t trust the site with their credit card information. Two reasons ranked higher: extra costs such as shipping, tax, and fees (40%) and slow delivery (20%). Trust design matters, but it works alongside cost and delivery transparency, not instead of them. For a store a visitor has never heard of, trust can be built from a handful of specific, testable page elements.
Fix the reasons Baymard ranks next to trust
Baymard’s cart abandonment list (last updated September 22, 2025) first sets aside the 42% of shoppers who abandoned because they were “just browsing / not ready to buy”. Among the remaining reasons, seven map directly to design work:
| Reason given for abandoning | Share | Design response |
|---|---|---|
| Extra costs too high (shipping, tax, fees) | 40% | Show shipping, tax, and fees on the product or cart page, before checkout starts |
| Delivery was too slow | 20% | Show a delivery estimate next to the price |
| Didn't trust the site with credit card information | 19% | Frame the card fields, add a recognizable seal or icon, offer familiar payment methods |
| The site wanted me to create an account | 18% | Offer guest checkout |
| Too long or complicated checkout process | 17% | Cut and simplify form fields |
| Website had errors or crashed | 17% | Test the checkout on every code change |
| Returns policy wasn't satisfactory | 13% | State the returns policy in plain terms and link it from the cart |
Read the table as a priority list. By Baymard’s numbers, surprise costs drive more abandonment than distrust of the payment form. The form side of that work is covered in how to design form fields that prevent errors.
Make the card fields look as protected as they are
Baymard’s checkout testing found that shoppers judge security field by field. In its article on how users perceive security during checkout, Baymard reports that test participants “generally had less confidence in the security of their credit card details” on pages where the card form didn’t look more secure than the fields for less sensitive information, such as the address. Its recommendation is visual encapsulation: give the card fields a distinct treatment, such as a framed area with a different background, that is not reused for less sensitive fields elsewhere in the form, and place one or two icons inside that frame.
Which seal you use matters less than older advice suggests. The familiar “use Norton and McAfee” recommendation traces back to Baymard’s January 2013 survey of 2,510 US adults, where Norton drew about 36% of votes and McAfee about 23%. In Baymard’s 2023 results, the Norton SSL seal still led, but a homemade seal with no meaning beyond the icon “performed significantly better than the SSL seals issued by established vendors except Norton.” Baymard’s reading is that, beyond strong brand recognition, the important thing is having a visual cue around the card fields. Baymard also notes its 2013 survey did not test one seal against several, so seal count is something to settle with your own A/B test. The finding concerns a visual cue, not a credential: a homemade seal should not imitate a certification mark or suggest a security audit the site has not had.
The browser padlock is no longer yours to lean on
HTTPS is the floor, not a trust signal. Google’s security team announced that starting with Chrome 68 in July 2018, Chrome would mark all HTTP sites as “not secure” (Google Security Blog, February 2018). Then the padlock itself was retired: in May 2023 the Chromium team announced it would replace the lock icon in Chrome’s address bar with a “tune” icon in Chrome 117, scheduled for early September 2023, noting that “only 11% of study participants correctly understood the precise meaning of the lock icon” (Chromium Blog). The trust cues you control are on the page, not in the browser’s address bar.
Show that a business stands behind the store
A first-time buyer is asking whether anyone will answer if the order goes wrong. Put the answers where they look before paying: a way to contact you, the returns policy, delivery times, and the full order cost. Baymard’s list puts “I couldn’t see / calculate total order cost up-front” at 12%, a separate complaint from high costs.
Polish counts too, though not in the way “make it pretty” implies. Baymard reports that layout bugs during checkout led some test participants to suspect the site had been hacked, and that it has seen participants abandon even major retailers’ checkouts over layout quirks. The practical rule is to test the checkout across browsers and devices every time its code changes.
Reviews: what helps, and what US rules prohibit
The Medill Spiegel Research Center’s study of online reviews, first published in 2017, found that purchase likelihood “typically peaks” at ratings from 4.0 to 4.7 and then decreases as ratings approach 5.0. It also reported that “negative reviews can have a positive impact because they establish credibility and authenticity”, and that verified-buyer badges improved the odds of purchase by 15%. A wall of five-star reviews is not the goal; a believable mix is.
Hiding the negatives carries legal risk in the United States, not only reputational risk. The FTC’s rule on consumer reviews and testimonials, effective October 21, 2024 (Federal Register, August 22, 2024), makes it a violation for a business to imply that the reviews it displays represent most or all submitted reviews while it suppresses reviews based on their rating or negative sentiment (16 CFR 465.7). The rule allows withholding reviews under criteria applied equally to all reviews, such as defamatory or abusive content, reviews the seller reasonably believes are fake, and reviews unrelated to what you sell.
Asking people you know for early reviews needs care. It is not banned outright, but the connection has to be disclosed:
- The FTC’s Endorsement Guides say material connections “can include a business, family, or personal relationship” and “the provision of free or discounted products”, and that a connection the audience would not expect must be disclosed clearly and conspicuously (16 CFR 255.5).
- Under the reviews rule, an officer or manager who asks immediate relatives or employees for reviews violates it if those reviews appear without disclosure and the officer or manager did not instruct reviewers to disclose the relationship, among other conditions. General requests to all purchasers are excluded (16 CFR 465.5).
So if you send free product to friends, family, or early supporters, ask each of them to state the relationship in the review itself, for example: “I received this free from the founder, who is my sister.” This is general information about US federal rules, not legal advice; a store selling into other countries or with an unusual review program should confirm its setup with counsel. Where and how to display reviews and testimonials on the page is covered in where to place social proof for credibility.
A starting checklist for a store with no reviews
| Element | What to put on the page | What to avoid |
|---|---|---|
| Costs and delivery | Shipping, tax, fees, and delivery estimate before checkout | A surprise line at the final step |
| Payment | Familiar payment methods and a framed card section with one or two icons | A dozen badges stacked near the button |
| Policies | Returns and guarantee terms in plain language, with any conditions stated | "Risk-free" wording the terms do not back up |
| People | An about page with the founder or team and a working contact route | A contact form as the only way to reach you |
| Early reviews | Reviews from real buyers, with any relationship disclosed in the review | Undisclosed reviews from friends, family, or staff |
| Press | Coverage that exists, linked to the article | Logos of outlets that never covered you |
Selling across borders
A store that ships abroad needs trust cues that work for each market: the payment methods buyers there use, prices in their currency, and copy written for them rather than machine-translated. Your payment provider’s country data and your own checkout logs are better guides than general claims about national preferences. The language side of this is covered in the design and content challenges of multilingual sites.
How to tell whether trust is the problem
Abandonment numbers do not say why someone left. Four checks narrow it down:
- Step-by-step drop-off. If visitors leave at the payment step more than at shipping, look at the card form and payment options first.
- New versus returning buyers. A large conversion gap between the two can point to first-visit trust.
- Moderated tests of the checkout. Comments like “this feels weird” at the payment step are the signal to note.
- A post-purchase question. Ask buyers, “What almost stopped you from buying?”
If you test removing trust badges and see no measurable difference, they may not be doing work on that page; if conversion drops, they were, and placement next to the card fields is the next variable to test.
Fix cost and delivery transparency first, then the card form and guest checkout, then the review program, and collect every early review with its relationship disclosed.