Quick answer
Baymard found that 47% of benchmarked sites that allowed guest checkout did not make it prominent enough, leaving 53% outside that group. In a separate benchmark, 42% prompted account creation before the shopper paid. Make the guest path clear and check it still works when the shopper enters a known email or cannot remember their password.
Baymard Institute, Make Guest Checkout the Most Prominent Option · Baymard Institute, Save Account Creation for the Confirmation Step
Key findings and sources
53% of guest-enabled sites fell outside the insufficient-prominence group
Basket Rescue’s calculation from Baymard’s guest checkout study puts 53% of guest-enabled benchmark sites outside the insufficient-prominence group. The published failure rate was 47%, so the two groups were six percentage points apart.
Baymard Institute, Make Guest Checkout the Most Prominent Option
Baymard’s sites that already allow guest checkout. Subtract the published 47% from 100%, then subtract 47% from the remainder.
58% were outside the early account prompt group
Baymard reported that 42% of assessed sites suggested account creation before or during checkout. The complementary 58% were outside that group, which does not establish that all of them offered account creation later.
Baymard Institute, Save Account Creation for the Confirmation Step
Sites assessed on the timing of account creation. Subtract the published 42% from 100%.
18.75% sign-in test abandonment equals about 188 per 1,000 at the same rate
Baymard observed an 18.75% average checkout abandonment rate in its existing-account sign-in testing and attributed it to forgotten passwords and reset-email problems. This is equivalent to about 188 per 1,000 test sign-ins at the same rate, rather than an actual count or general checkout rate.
Baymard Institute, Avoid Unnecessarily Complex Password-Creation Requirements
Usability test participants who tried to sign in to preexisting accounts. Multiply the published percentage by 10, then round 187.5 to the nearest whole attempt for the cited passage.
Phone and app examples made up 51.49% of gallery records
Baymard’s public account-selection gallery contained 502 mobile examples and 69 app examples out of 1,109 total. Combined, those device groups made up 51.49% of screenshot records, rather than 51.49% of benchmark retailers.
Baymard Institute, Account Selection design examples
Public account-selection design gallery, read November 2026. Add mobile and app records, divide by total records and convert to a percentage.
Compare the results
Nearly half of guest-enabled sites obscured the guest option
Offering guest checkout only helps if shoppers can find it.
A guest option can exist and still be missed
A shopper trying to complete a one-off purchase usually reads the first thing checkout asks them to do. If the page opens on a sign-in form, they can reasonably assume they need an account, even if a guest link sits below it. Baymard found insufficient prominence on 47% of benchmarked sites that offered guest checkout. That leaves 53% outside the reported group. The two shares differ by only six points, even though the benchmark already excludes sites without a guest path.
Open your checkout on a phone without a session. Look for a clearly labeled action to continue as a guest before you fill anything in. Check the same flow in a reopened cart or recovery link. The label and location need to make sense after a shopper returns, when a stored email or expired session might cause a different screen to appear.
Test what happens when the email already belongs to an account
Baymard reported 18.75% average checkout abandonment among participants who tried signing in to preexisting accounts in its test. It attributed those exits to forgotten passwords and reset-email problems. That result is for the tested sign-in situation. It does not mean 18.75% of your checkout traffic will leave. It identifies a case to test directly, because a shopper who enters an address they used before may be pushed toward a password they do not remember.
Try the full path with a new email, an email on an existing account and an account whose reset email never arrives. The order can still need an email for confirmation and delivery without requiring account creation. If guest checkout is supported, make its destination clear and preserve the cart when switching. If the shopper genuinely needs their account to use a benefit, say which benefit and let them decide whether signing in is worth it.
Follow the route a recovery message actually opens. A link that lands on a generic login page can strand someone even when the normal cart has a guest button. Check on a signed-out phone that the link restores the right items, keeps the total visible and allows the shopper to proceed.
Offer the account after payment and measure what changes
A separate Baymard item reports that 42% of assessed sites asked users to create an account before they placed the order. That leaves 58% outside the early-prompt group. Some of those shops may offer optional registration on the confirmation page, while others may never ask. The takeaway for an experiment is straightforward. A shop can measure account creation after payment separately from whether the buyer completes the order.
Compare completion at the account step and paid orders per checkout start before and after changing the placement. Break the results out by new and recognized email, and track resets that do not return to the checkout. Watch for orders that lose discounts or shipping details when the shopper switches to guest. If paid orders rise, measure account creation after purchase and check whether the shopper still gets the features they need.
Study data
| Measure | Value | Context |
|---|---|---|
| Guest option not made prominent enough | 47% | Failed the reported prominence criterion |
| Guest-enabled sites outside the insufficient-prominence group | 53% | Calculated remainder of the reported criterion |
| Measure | Value | Context |
|---|---|---|
| Guest option difficult to locate | 44% | Public benchmark observation on locating the guest option |
| Sites represented in the parent UX benchmark | 344 sites | Number of ecommerce sites in the stated parent benchmark |
| Account selection gallery examples | 1,109 screenshots | Public gallery count |
| Account selection desktop examples | 538 screenshots | Public gallery count |
| Account selection mobile examples | 502 screenshots | Public gallery count |
| Account selection app examples | 69 screenshots | Public gallery count |
| Measure | Value | Context |
|---|---|---|
| Account prompt before the order was placed | 42% | Asked the shopper to create an account before or during checkout |
| Sites outside the early-account-prompt group | 58% | Calculated remainder; it includes sites with no account offer |
| Measure | Value | Context |
|---|---|---|
| Unnecessarily complex password requirements | 82% | Baymard benchmarked password-creation requirements |
| Checkout abandonment attributed to reset email in the sign-in test | 18.75% | Baymard attributed the observed abandonment to forgotten passwords and reset-email problems. |
What changes when you compare the figures
| Comparison | Result | Calculation |
|---|---|---|
| Guest-enabled sites outside insufficient-prominence group | 53% | = 100 - 47 |
| Difference between remainder and insufficient prominence | 6 percentage points | = (100 - 47) - 47 |
| Sites outside early-account-prompt group | 58% | = 100 - 42 |
| Phone and app share of account selection gallery | 51.49% | = (502 + 69) / 1109 * 100 |
| Sign-in test abandonment attributed to reset email, per 1000 | 187.5 per 1,000 test sign-ins | = 18.75 * 10 |
Methodology and sources
We transcribed the guest prominence, account selection, delayed registration and password figures from Baymard’s public research. We calculate complements within each published benchmark item without assuming the samples are identical. Abandonment attributed to password reset is an observed sign-in test outcome, rather than a benchmark share of retailers. Gallery records support the device breakdown only, and 344 is the parent benchmark size rather than a per-question denominator. We report no pooled prevalence or causal conversion forecast.
Download the source data (CSV)
Download the calculations (CSV)
Related reading
Cart abandonment statistics and the problems behind them
Checkout form error statistics and how shoppers recover
Return policy checkout statistics and what buyers need to see before paying
Cart abandonment reasons merchants miss
Cart editing statistics for orders shoppers can still correct
Sources and definitions
Baymard Institute, Make Guest Checkout the Most Prominent Option
Baymard Institute, Account Selection design examples
Baymard Institute, Save Account Creation for the Confirmation Step
Baymard Institute, Avoid Unnecessarily Complex Password-Creation Requirements
The source data includes the population, period, denominator and limitations for every figure.