← Research blog

Checkout Bench in Action: See Amtrak from Search to Payment

See how two Pacific Surfliner tickets went from a station search to a $144 checkout, with all 34 screens shown side by side with our annotations and digital twins.

LAX ⇄ SAN

Los Angeles Union Station · San Diego Santa Fe Depot

Dec 4–6 · Two adults · Coach Flex

$144 due / $0 chargedSee checkout

Sign in

We dismissed the cookie banner, opened Guest Rewards and signed in. Amtrak then brought us back to the booking form.

01Cookie choices

Cookie choices — Amtrak
Amtrak

The cookie banner covered the booking controls while a Guest Rewards prompt was also open. We chose Reject All so we could use the form.

Next: Dismiss cookie banner
Cookie choices — CB decision model
CB decision model

Predicted first click: navigation (25%). The model rated action clarity at 61%, hierarchy at 32% and readability at 66%.

View scores
RegionConfidenceFirst click
1 · Primary action19%23%
2 · Secondary action60%10%
3 · Navigation40%25%
4 · Main heading59%20%
5 · Main visual72%22%
Cookie choices — Digital twin
Digital twin

The reconstruction keeps the consent controls along the bottom and the sign-in prompt in the upper corner. The train photograph becomes a dark wash behind them. Match score: 67%.

02The signed-out booking form

The signed-out booking form — Amtrak
Amtrak

Our first attempt to dismiss the banner closed the sign-in prompt instead. A second click dismissed the banner and exposed the booking form. A dismissible rewards-card promotion remained along the bottom.

Next: Open Guest Rewards
The signed-out booking form — CB decision model
CB decision model

Predicted first click: primary action (50%). The model rated action clarity at 81%, hierarchy at 55% and readability at 89%.

View scores
RegionConfidenceFirst click
1 · Primary action49%50%
2 · Secondary action22%3%
3 · Navigation68%15%
4 · Main heading66%21%
5 · Main visual69%11%
The signed-out booking form — Digital twin
Digital twin

With the banner gone, the twin reproduces the station fields and the pale Find Trains button above the promotion. Match score: 67%.

03Opening Guest Rewards

Opening Guest Rewards — Amtrak
Amtrak

Sign In appeared inside Guest Rewards after the menu had been open for about two seconds. Clicking the trigger again during that delay could close the menu.

Next: Choose Sign In
Opening Guest Rewards — CB decision model
CB decision model

Predicted first click: main heading (30%). The model rated action clarity at 34%, hierarchy at 63% and readability at 80%.

View scores
RegionConfidenceFirst click
1 · Primary action36%18%
2 · Secondary action48%2%
3 · Navigation56%25%
4 · Main heading66%30%
5 · Main visual71%25%
Opening Guest Rewards — Digital twin
Digital twin

The twin retains the open menu, including the outlined Sign In control beside Join Now. Match score: 89%.

04Member sign-in

Member sign-in — Amtrak
Amtrak

Amtrak opened a separate sign-in page where the member entered an email or rewards number and a password. Sign In stayed gray until both fields were filled.

Next: Enter member credentials
Member sign-in — CB decision model
CB decision model

Predicted first click: primary action (59%). The model rated action clarity at 96%, hierarchy at 84% and readability at 89%.

View scores
RegionConfidenceFirst click
1 · Primary action83%59%
2 · Secondary action54%18%
3 · Navigation55%11%
4 · Main heading71%12%
5 · Main visual3%0%
Member sign-in — Digital twin
Digital twin

The reconstruction preserves the compact form in the middle of a mostly empty page. Match score: 56%.

05Returning to the home page

Returning to the home page — Amtrak
Amtrak

After sign-in, the member’s first name replaced Guest Rewards in the header. The booking form otherwise looked the same, and we continued to the departure field. A blank band briefly appeared beneath the header while the page loaded.

Next: Open the departure field
Returning to the home page — CB decision model
CB decision model

Predicted first click: primary action (48%). The model rated action clarity at 83%, hierarchy at 48% and readability at 84%.

View scores
RegionConfidenceFirst click
1 · Primary action50%48%
2 · Secondary action14%1%
3 · Navigation62%25%
4 · Main heading81%14%
5 · Main visual70%12%
Returning to the home page — Digital twin
Digital twin

The twin returns to the same booking layout. Personal account details are masked in the published views. Match score: 89%.

Stations, dates and passengers

We wanted two coach seats from Los Angeles Union Station to San Diego Santa Fe Depot on Friday, December 4, returning around noon on Sunday, December 6. Before searching for trains, the form had to distinguish cities from stations and a return date from a replacement departure date.

06Choosing Los Angeles Union Station

Choosing Los Angeles Union Station — Amtrak
Amtrak

Searching for Los Angeles offered a station, a city location and an option to enter a station manually. We selected Union Station (LAX) for the departure.

Next: Select Union Station (LAX)
Choosing Los Angeles Union Station — CB decision model
CB decision model

Predicted first click: primary action (50%). The model rated action clarity at 79%, hierarchy at 51% and readability at 86%.

View scores
RegionConfidenceFirst click
1 · Primary action53%50%
2 · Secondary action23%2%
3 · Navigation65%18%
4 · Main heading40%14%
5 · Main visual46%16%
Choosing Los Angeles Union Station — Digital twin
Digital twin

The suggestions appear beneath the departure field in both versions while the rest of the form stays in place. Match score: 67%.

07Choosing a San Diego station

Choosing a San Diego station — Amtrak
Amtrak

The destination field offered Santa Fe Depot (SAN) and Old Town (OLT). We picked Santa Fe Depot for a downtown arrival. A city name alone would have left that choice unresolved.

Next: Select Santa Fe Depot (SAN)
Choosing a San Diego station — CB decision model
CB decision model

Predicted first click: primary action (60%). The model rated action clarity at 78%, hierarchy at 49% and readability at 85%.

View scores
RegionConfidenceFirst click
1 · Primary action39%60%
2 · Secondary action19%2%
3 · Navigation65%15%
4 · Main heading53%13%
5 · Main visual58%10%
Choosing a San Diego station — Digital twin
Digital twin

The twin retains the two San Diego options beneath the active destination field. Match score: 78%.

08Opening the calendar

Opening the calendar — Amtrak
Amtrak

The date picker showed two months with past dates disabled and the current day outlined. At this point the search was still set to one way.

Next: Choose December 4
Opening the calendar — CB decision model
CB decision model

Predicted first click: primary action (53%). The model rated action clarity at 75%, hierarchy at 47% and readability at 89%.

View scores
RegionConfidenceFirst click
1 · Primary action44%53%
2 · Secondary action19%3%
3 · Navigation75%16%
4 · Main heading64%20%
5 · Main visual60%8%
Opening the calendar — Digital twin
Digital twin

The calendar also floats above the promotion in the twin, whose plainer background makes its edge easier to distinguish. Match score: 78%.

09Switching to round trip

Switching to round trip — Amtrak
Amtrak

While the departure field was active, each new date replaced the previous departure. Clicking Return Date changed the trip type to round trip so we could enter Sunday’s return.

Next: Choose December 6
Switching to round trip — CB decision model
CB decision model

Predicted first click: primary action (64%). The model rated action clarity at 73%, hierarchy at 45% and readability at 89%.

View scores
RegionConfidenceFirst click
1 · Primary action61%64%
2 · Secondary action13%2%
3 · Navigation67%15%
4 · Main heading70%14%
5 · Main visual50%5%
Switching to round trip — Digital twin
Digital twin

The reconstructed calendar appears beside the date fields, with the selected departure still in view. Match score: 67%.

10Checking both dates

Checking both dates — Amtrak
Amtrak

December 4 and December 6 appeared together in the search field and were circled in the calendar. We checked both before closing it with Done.

Next: Close the calendar with Done
Checking both dates — CB decision model
CB decision model

Predicted first click: primary action (70%). The model rated action clarity at 81%, hierarchy at 47% and readability at 91%.

View scores
RegionConfidenceFirst click
1 · Primary action66%70%
2 · Secondary action10%1%
3 · Navigation65%9%
4 · Main heading65%15%
5 · Main visual59%5%
Checking both dates — Digital twin
Digital twin

The twin retains the marked dates and the controls at the bottom of the open calendar. Match score: 78%.

11Opening the traveler list

Opening the traveler list — Amtrak
Amtrak

The traveler picker separated adults, seniors, youths, children and infants, with discounts selected separately. We increased the adult count for the second passenger.

Next: Add one adult
Opening the traveler list — CB decision model
CB decision model

Predicted first click: primary action (81%). The model rated action clarity at 84%, hierarchy at 41% and readability at 67%.

View scores
RegionConfidenceFirst click
1 · Primary action68%81%
2 · Secondary action9%1%
3 · Navigation61%11%
4 · Main heading31%1%
5 · Main visual32%6%
Opening the traveler list — Digital twin
Digital twin

The reconstruction keeps the passenger categories and quantity controls in the panel below the form. Match score: 67%.

12Searching for two adults

Searching for two adults — Amtrak
Amtrak

The form showed two adults with the stations and dates filled in. After closing the picker, we used the green Find Trains button.

Next: Close the picker and find trains
Searching for two adults — CB decision model
CB decision model

Predicted first click: primary action (80%). The model rated action clarity at 84%, hierarchy at 41% and readability at 73%.

View scores
RegionConfidenceFirst click
1 · Primary action70%80%
2 · Secondary action9%1%
3 · Navigation70%11%
4 · Main heading40%1%
5 · Main visual45%7%
Searching for two adults — Digital twin
Digital twin

The twin preserves the active search button and the selected count while the passenger picker is open. Match score: 67%.

Choose trains

We chose a fare for each direction. Several departures cost the same, so we checked the train number and departure time before adding each one.

13Reading outbound departures

Reading outbound departures — Amtrak
Amtrak

Train 562 left Union Station at 6:10 a.m. and reached Santa Fe Depot at 9:04 a.m. Coach started at $36 per traveler and Business at $55, while the cart was still empty. We opened Coach on the 562 row.

Next: Open Coach on train 562
Reading outbound departures — CB decision model
CB decision model

Predicted first click: navigation (40%). The model rated action clarity at 57%, hierarchy at 52% and readability at 77%.

View scores
RegionConfidenceFirst click
1 · Primary action8%19%
2 · Secondary action21%6%
3 · Navigation67%40%
4 · Main heading27%33%
5 · Main visual12%2%
Reading outbound departures — Digital twin
Digital twin

Both timetables show the departure and arrival times. On the original, the CB decision model marked Revise Search as the main action. We needed the Coach fare farther down the page. Match score: 44%.

14Opening Coach Flex

Opening Coach Flex — Amtrak
Amtrak

The open panel offered an unreserved Coach Flex seat for $36 per traveler, with a full refund before departure and changes without a fee. The fare was preselected and still required Select. Opening the fare panel dimmed the rest of the timetable.

Next: Select the outbound fare
Opening Coach Flex — CB decision model
CB decision model

Predicted first click: primary action (50%). The model rated action clarity at 75%, hierarchy at 73% and readability at 60%.

View scores
RegionConfidenceFirst click
1 · Primary action81%50%
2 · Secondary action6%2%
3 · Navigation73%29%
4 · Main heading62%12%
5 · Main visual8%7%
Opening Coach Flex — Digital twin
Digital twin

The twin preserves the fare card within the expanded row and the Select button in the lower-right corner. Match score: 67%.

15The first return option

The first return option — Amtrak
Amtrak

Adding the outbound fare put $72 in the cart. The return timetable opened on train 761 at 4:01 a.m. and offered 13 trains across two pages. Repeated Lowest Fare and Fastest badges did little to distinguish them.

Next: Scroll to later departures
The first return option — CB decision model
CB decision model

Predicted first click: navigation (44%). The model rated action clarity at 41%, hierarchy at 51% and readability at 76%.

View scores
RegionConfidenceFirst click
1 · Primary action8%25%
2 · Secondary action18%9%
3 · Navigation75%44%
4 · Main heading25%21%
5 · Main visual25%1%
The first return option — Digital twin
Digital twin

The reconstruction keeps the outbound cart state above the first return departure. Match score: 33%.

16Finding the noon departure

Finding the noon departure — Amtrak
Amtrak

Scrolling brought the 8:01 a.m., 10:01 a.m. and 12:01 p.m. departures into view. Each showed $36 for Coach and $55 for Business, so we used train number and time to locate 777.

Next: Open Coach on the noon departure
Finding the noon departure — CB decision model
CB decision model

Predicted first click: navigation (85%). The model rated action clarity at 28%, hierarchy at 22% and readability at 73%.

View scores
RegionConfidenceFirst click
1 · Primary action7%6%
2 · Secondary action6%2%
3 · Navigation68%85%
4 · Main heading21%5%
5 · Main visual22%2%
Finding the noon departure — Digital twin
Digital twin

In the twin the successive rows also repeat the same fares, leaving the timetable as the way to tell the departures apart. Match score: 56%.

17Inspecting train 769

Inspecting train 769 — Amtrak
Amtrak

A click aimed at 777 while the list was moving opened train 769, leaving at 8:01 a.m. Opening the fare panel also scrolled its row to the top. We closed it before making a selection.

Next: Close the 769 fare panel
Inspecting train 769 — CB decision model
CB decision model

Predicted first click: primary action (61%). The model rated action clarity at 78%, hierarchy at 84% and readability at 65%.

View scores
RegionConfidenceFirst click
1 · Primary action79%61%
2 · Secondary action10%1%
3 · Navigation58%30%
4 · Main heading7%4%
5 · Main visual7%4%
Inspecting train 769 — Digital twin
Digital twin

The expanded morning row and its $36 Coach panel remain visible together in the twin. Match score: 56%.

18Selecting train 777

Selecting train 777 — Amtrak
Amtrak

On the reopened fare panel we verified train 777, leaving San Diego at 12:01 p.m. and arriving in Los Angeles at 2:57 p.m., before selecting the $36 fare.

Next: Select the 777 fare
Selecting train 777 — CB decision model
CB decision model

Predicted first click: primary action (52%). The model rated action clarity at 77%, hierarchy at 77% and readability at 52%.

View scores
RegionConfidenceFirst click
1 · Primary action80%52%
2 · Secondary action6%1%
3 · Navigation65%36%
4 · Main heading10%5%
5 · Main visual7%6%
Selecting train 777 — Digital twin
Digital twin

The twin keeps the noon departure aligned with the open fare card, so the itinerary and price can be checked together. Match score: 78%.

Review passengers

With both trains in the cart, we checked the seats and baggage allowance. Amtrak filled in some passenger details from the member profile and asked for the rest.

19Checking the round trip

Checking the round trip — Amtrak
Amtrak

The trip summary showed two unreserved coach seats on train 562 on Friday and train 777 on Sunday. The cart now totaled $144 for both adults and both directions. The fixed Back and Continue bar covered part of the summary at the bottom.

Next: Continue with this itinerary
Checking the round trip — CB decision model
CB decision model

Predicted first click: primary action (47%). The model rated action clarity at 87%, hierarchy at 47% and readability at 81%.

View scores
RegionConfidenceFirst click
1 · Primary action87%47%
2 · Secondary action80%2%
3 · Navigation72%29%
4 · Main heading68%21%
5 · Main visual5%1%
Checking the round trip — Digital twin
Digital twin

The reconstruction retains the two itinerary columns, making the outbound and return departures visible together. Match score: 56%.

20Reviewing baggage and extras

Reviewing baggage and extras — Amtrak
Amtrak

The bicycle and pet quantities stayed at zero. The page separately listed one personal item up to 25 lb and two carry-ons up to 50 lb each per passenger. We continued with the included allowance.

Next: Continue without extras
Reviewing baggage and extras — CB decision model
CB decision model

Predicted first click: navigation (31%). The model rated action clarity at 75%, hierarchy at 30% and readability at 75%.

View scores
RegionConfidenceFirst click
1 · Primary action86%29%
2 · Secondary action93%2%
3 · Navigation68%31%
4 · Main heading61%22%
5 · Main visual26%16%
Reviewing baggage and extras — Digital twin
Digital twin

The twin preserves the add-on selectors and baggage allowances, while simplifying the original luggage illustrations. Match score: 67%.

21Entering passenger details

Entering passenger details — Amtrak
Amtrak

The member profile filled the first traveler’s contact and rewards fields. A person supplied the required date of birth and the second traveler’s details. The page also offered to save the birth date to the profile.

Next: Complete both passenger forms
Entering passenger details — CB decision model
CB decision model

Predicted first click: navigation (57%). The model rated action clarity at 65%, hierarchy at 45% and readability at 82%.

View scores
RegionConfidenceFirst click
1 · Primary action28%16%
2 · Secondary action89%3%
3 · Navigation63%57%
4 · Main heading54%23%
5 · Main visual7%1%
Entering passenger details — Digital twin
Digital twin

Both versions retain the traveler tabs and the birth-date field. Personal values in the forms have been masked. Match score: 67%.

Checkout

The last decision before payment was whether to add travel protection. After declining it, we checked the balance and itinerary, left the terms unchecked and stopped before entering payment.

22Deciding about travel protection

Deciding about travel protection — Amtrak
Amtrak

The payment page required a yes-or-no choice about insurance. Accepting would add $9 per traveler, or $18 for the trip. The recommended option listed benefits, while declining required acknowledging the risk and pressing Confirm. We declined.

Next: Decline protection and confirm
Deciding about travel protection — CB decision model
CB decision model

Predicted first click: primary action (48%). The model rated action clarity at 63%, hierarchy at 58% and readability at 82%.

View scores
RegionConfidenceFirst click
1 · Primary action42%48%
2 · Secondary action66%6%
3 · Navigation70%14%
4 · Main heading35%28%
5 · Main visual52%4%
Deciding about travel protection — Digital twin
Digital twin

The reconstruction keeps the recommended choice, the decline option and the confirmation button in the same dialog. Match score: 56%.

23Reconciling the balance due

Reconciling the balance due — Amtrak
Amtrak

The itinerary and price summary showed $144 due, expressed as two travelers at $72 each. Earlier, $72 covered the outbound trip for both adults; here it covered a round trip for one adult. The required terms box remained empty. A warning about the declined protection remained on the page.

Next: Leave the terms unchecked
Reconciling the balance due — CB decision model
CB decision model

Predicted first click: primary action (31%). The model rated action clarity at 84%, hierarchy at 45% and readability at 78%.

View scores
RegionConfidenceFirst click
1 · Primary action73%31%
2 · Secondary action80%2%
3 · Navigation74%31%
4 · Main heading73%25%
5 · Main visual66%11%
Reconciling the balance due — Digital twin
Digital twin

The twin preserves both legs, the balance and the unchecked terms box. The original card-brand logos beneath the Payment tab become colored blocks. Match score: 78%.

24Stopping at payment

Stopping at payment — Amtrak
Amtrak

The page offered Payment, Voucher and Gift Card tabs. The account had no saved card, and the new-card and PayPal options stayed closed. Purchase remained disabled; we submitted no payment and received no reservation.

Next: Stop before purchase
Stopping at payment — CB decision model
CB decision model

Predicted first click: primary action (34%). The model rated action clarity at 88%, hierarchy at 45% and readability at 85%.

View scores
RegionConfidenceFirst click
1 · Primary action87%34%
2 · Secondary action93%2%
3 · Navigation54%32%
4 · Main heading64%23%
5 · Main visual26%9%
Stopping at payment — Digital twin
Digital twin

The twin retains the payment choices and the disabled Purchase control beside Back. Match score: 89%.

After checkout

We opened Guest Rewards to see whether a reservation appeared, which passenger details were saved and how much trip history the member could see.

25Looking up a reservation

Looking up a reservation — Amtrak
Amtrak

Even after sign-in, My Trip asked for a six-digit reservation number and another search field. The member’s own trips were accessible through a different menu.

Next: Open the member menu
Looking up a reservation — CB decision model
CB decision model

Predicted first click: navigation (70%). The model rated action clarity at 63%, hierarchy at 36% and readability at 77%.

View scores
RegionConfidenceFirst click
1 · Primary action77%20%
2 · Secondary action94%2%
3 · Navigation68%70%
4 · Main heading30%7%
5 · Main visual12%1%
Looking up a reservation — Digital twin
Digital twin

The reconstruction keeps the reservation lookup fields in a panel over the payment page. Match score: 89%.

26Opening the member menu

Opening the member menu — Amtrak
Amtrak

The Guest Rewards menu contained links to the account, profile, trips, coupons, lounge access and sign-out. We used My Trips to inspect the account after checkout.

Next: Open My Trips
Opening the member menu — CB decision model
CB decision model

Predicted first click: navigation (82%). The model rated action clarity at 11%, hierarchy at 26% and readability at 70%.

View scores
RegionConfidenceFirst click
1 · Primary action9%1%
2 · Secondary action11%0%
3 · Navigation32%82%
4 · Main heading52%15%
5 · Main visual9%2%
Opening the member menu — Digital twin
Digital twin

The twin preserves the menu columns and account destinations. Member identifiers are masked. Match score: 56%.

27Checking upcoming trips

Checking upcoming trips — Amtrak
Amtrak

The upcoming list contained no trip. Together with the disabled Purchase button, this shows where the session ended: the itinerary reached the cart without becoming a reservation.

Next: Check past trips
Checking upcoming trips — CB decision model
CB decision model

Predicted first click: navigation (73%). The model rated action clarity at 20%, hierarchy at 68% and readability at 81%.

View scores
RegionConfidenceFirst click
1 · Primary action35%3%
2 · Secondary action15%1%
3 · Navigation67%73%
4 · Main heading47%14%
5 · Main visual64%9%
Checking upcoming trips — Digital twin
Digital twin

The twin keeps the selected upcoming tab and its empty state beneath the member summary. Match score: 56%.

28Checking past trips

Checking past trips — Amtrak
Amtrak

The past tab was also empty and explained that reservations disappear five days after a trip ends. An assistant trying to infer a member’s usual route from this list would have limited history.

Next: Open My Profile
Checking past trips — CB decision model
CB decision model

Predicted first click: navigation (80%). The model rated action clarity at 17%, hierarchy at 55% and readability at 87%.

View scores
RegionConfidenceFirst click
1 · Primary action6%3%
2 · Secondary action18%2%
3 · Navigation59%80%
4 · Main heading67%14%
5 · Main visual9%1%
Checking past trips — Digital twin
Digital twin

The reconstruction preserves the past-trip tab and the explanation beside the empty list. Match score: 56%.

29Reviewing account details

Reviewing account details — Amtrak
Amtrak

The profile brought together member information, password, email and phone settings. Most changes had an edit control, while changing the member name required contacting Amtrak by phone. A banner prompted the member to complete the profile.

Next: Scroll through the profile
Reviewing account details — CB decision model
CB decision model

Predicted first click: navigation (77%). The model rated action clarity at 19%, hierarchy at 54% and readability at 88%.

View scores
RegionConfidenceFirst click
1 · Primary action40%2%
2 · Secondary action24%1%
3 · Navigation70%77%
4 · Main heading58%19%
5 · Main visual8%1%
Reviewing account details — Digital twin
Digital twin

The twin keeps the profile tabs and edit layout. Account and contact values are masked in both views. Match score: 44%.

30Checking what the profile saved

Checking what the profile saved — Amtrak
Amtrak

The profile described verification by email or text, showed an empty billing address and displayed the date of birth entered earlier. Marketing choices remained off. Reaching payment had already changed one profile field.

Next: Open Payment Methods
Checking what the profile saved — CB decision model
CB decision model

Predicted first click: navigation (85%). The model rated action clarity at 25%, hierarchy at 32% and readability at 72%.

View scores
RegionConfidenceFirst click
1 · Primary action46%10%
2 · Secondary action4%2%
3 · Navigation70%85%
4 · Main heading10%2%
5 · Main visual6%1%
Checking what the profile saved — Digital twin
Digital twin

The reconstruction retains the authentication, billing and preference sections. The saved personal values are masked. Match score: 44%.

31Reviewing payment methods

Reviewing payment methods — Amtrak
Amtrak

The account had room for up to three saved cards but held none, and the preferred payment method was unset. The page also listed digital payment options, vouchers and gift cards.

Next: Open Preferences
Reviewing payment methods — CB decision model
CB decision model

Predicted first click: navigation (64%). The model rated action clarity at 21%, hierarchy at 42% and readability at 86%.

View scores
RegionConfidenceFirst click
1 · Primary action35%4%
2 · Secondary action14%2%
3 · Navigation61%64%
4 · Main heading51%30%
5 · Main visual4%0%
Reviewing payment methods — Digital twin
Digital twin

The twin retains the payment settings and their empty states, while simplifying the row of brand marks. Match score: 56%.

32Checking station preferences

Checking station preferences — Amtrak
Amtrak

Favorite departure and arrival stations were both empty, and the default discount type was Adult. Signing in could fill traveler details but could not settle which San Diego station to choose.

Next: Open My Transactions
Checking station preferences — CB decision model
CB decision model

Predicted first click: navigation (78%). The model rated action clarity at 15%, hierarchy at 47% and readability at 75%.

View scores
RegionConfidenceFirst click
1 · Primary action44%2%
2 · Secondary action26%1%
3 · Navigation64%78%
4 · Main heading72%19%
5 · Main visual8%0%
Checking station preferences — Digital twin
Digital twin

The reconstruction keeps the empty station fields within the Preferences tab. Match score: 44%.

33Reading the transaction window

Reading the transaction window — Amtrak
Amtrak

The transaction list showed zero results for its selected date window. That filter did not establish the outcome of this checkout; the payment controls and upcoming trips supplied the direct evidence.

Next: Open My Coupons
Reading the transaction window — CB decision model
CB decision model

Predicted first click: navigation (69%). The model rated action clarity at 24%, hierarchy at 60% and readability at 84%.

View scores
RegionConfidenceFirst click
1 · Primary action64%6%
2 · Secondary action8%1%
3 · Navigation71%69%
4 · Main heading50%17%
5 · Main visual36%7%
Reading the transaction window — Digital twin
Digital twin

The twin retains the date filters and transaction columns, so the scope of the empty result remains visible. Match score: 44%.

34Checking coupons

Checking coupons — Amtrak
Amtrak

The member’s coupon page showed no available coupons, completing the account review after the booking session.

Next: End of session
Checking coupons — CB decision model
CB decision model

Predicted first click: navigation (73%). The model rated action clarity at 17%, hierarchy at 63% and readability at 81%.

View scores
RegionConfidenceFirst click
1 · Primary action4%1%
2 · Secondary action8%1%
3 · Navigation69%73%
4 · Main heading13%17%
5 · Main visual31%8%
Checking coupons — Digital twin
Digital twin

The reconstruction retains the empty coupon state beneath the account header. Match score: 67%.