Shedd Aquarium’s out-of-state visitors pay double the local ticket price and convert 1.6 points less. Sixteen eye-tracking sessions showed why: the prices were on the page, but people’s eyes slid right past them. 94% came away unclear what their visit would actually cost. Closing that gap is worth roughly $95 to 107K in recoverable revenue every month.
Spring 2025
UX Researcher
Hridya Nadappattel, Iris Sun, Saskia Suherman, Conor Mack
UX Research
Chicago residents buy a fixed-price ticket. Everyone else gets variable pricing, and has to click through four more screens than a local before seeing a real number. Google Analytics for March 2025 put the cost of that opacity in plain sight: fixed-price visitors converted at 4.7%, while variable-price visitors, who pay $37 to 46 a ticket instead of $19.99, converted at 3.1%, despite browsing 56% more.
Research question: how might Shedd’s ticket-purchasing process better support individual and family visitors?
I ran 16 moderated eye-tracking tests (7 desktop, 9 mobile), synthesized them with a rainbow spreadsheet (16 participants across roughly 45 observed behaviors), and prioritized issues with teammate-averaged RICE scoring. Price transparency scored 62.5, nearly double the next issue.
Prices were shown, but never seen. Gaze plots showed attention scattering across the calendar and time-slot pages without ever fixating on cost. 94% of participants were unclear about pricing between the date and time steps.
The calendar demanded the most work at the worst moment. The time-slot screen logged the highest fixation count of the entire flow: 624 fixations, 243 seconds of total fixation time, because dates, times, prices, and eligibility were crammed into one view coded with both colors and icons, with no legend. “I didn’t understand what the colours were or what all the text about holiday period or discount periods were.” 14 of 16 participants hesitated here, and by the time they reached the time-slot page they were cognitively exhausted enough to skim right past per-slot price changes. “That’s why it was a very quick scroll through.”
The flow only moved one way. 15 of 16 participants hit navigation dead ends: no back buttons inside the flow, green CTAs that meant “navigate” on one screen and “purchase” on the next, add-ons that couldn’t be removed once added. “If none of these dates work for me, I can’t close the window. That’s annoying.”
“I picked $38, but then the next screen says $41. I’m like… wait, what’s going on?” A study participant
Recommendations, in RICE order:
Surface real prices before the funnel: a dynamic price calendar on the “Plan Your Visit” page, so out-of-state visitors see cost without entering checkout. RICE 62.5.
Decompress the calendar: one encoding system with a legend, calendar above the fold, dynamic pricing moved to the time-slot step. RICE 33.3.
Make the flow two-way: consistent back and next buttons and a sticky progress stepper. RICE 32.
Then a designed A/B test to check the biggest recommendation before Shedd builds around it: 4 weeks, non-Illinois visitors only, control versus a price calendar added to Plan Your Visit. The key metric is reduced funnel entry from that page, because the study’s sharpest insight was that a lot of “abandoning” visitors were never buying at all. They entered checkout just to find out the price, which quietly distorts every funnel metric downstream from it.
The baseline SUS score of 69.4, alongside real strengths like guest checkout and a “Get Tickets” button that 14 of 16 people found instantly, is why the recommendation is to support the existing flow rather than rebuild it. Shedd’s foundation already works. What’s leaking is specific, and it’s testable.
Sixteen remote participants, not on-site Chicago visitors. SUS data from only 13 of 16. One month of GA4. And the revenue figure is an opportunity estimate: it assumes price transparency closes the conversion gap, which is exactly what the A/B test exists to check rather than what this study alone proves. What I’d want to know next: across e-commerce funnels generally, how much of what looks like drop-off is actually just failed price discovery wearing a purchase-intent costume?
Shedd Aquarium’s out-of-state visitors pay double the local ticket price and convert 1.6 points less. Sixteen eye-tracking sessions showed why: the prices were on the page, but people’s eyes slid right past them. 94% came away unclear what their visit would actually cost. Closing that gap is worth roughly $95 to 107K in recoverable revenue every month.
Spring 2025
UX Researcher
Hridya Nadappattel, Iris Sun, Saskia Suherman, Conor Mack
UX Research
Chicago residents buy a fixed-price ticket. Everyone else gets variable pricing, and has to click through four more screens than a local before seeing a real number. Google Analytics for March 2025 put the cost of that opacity in plain sight: fixed-price visitors converted at 4.7%, while variable-price visitors, who pay $37 to 46 a ticket instead of $19.99, converted at 3.1%, despite browsing 56% more.
Research question: how might Shedd’s ticket-purchasing process better support individual and family visitors?
I ran 16 moderated eye-tracking tests (7 desktop, 9 mobile), synthesized them with a rainbow spreadsheet (16 participants across roughly 45 observed behaviors), and prioritized issues with teammate-averaged RICE scoring. Price transparency scored 62.5, nearly double the next issue.
Prices were shown, but never seen. Gaze plots showed attention scattering across the calendar and time-slot pages without ever fixating on cost. 94% of participants were unclear about pricing between the date and time steps.
The calendar demanded the most work at the worst moment. The time-slot screen logged the highest fixation count of the entire flow: 624 fixations, 243 seconds of total fixation time, because dates, times, prices, and eligibility were crammed into one view coded with both colors and icons, with no legend. “I didn’t understand what the colours were or what all the text about holiday period or discount periods were.” 14 of 16 participants hesitated here, and by the time they reached the time-slot page they were cognitively exhausted enough to skim right past per-slot price changes. “That’s why it was a very quick scroll through.”
The flow only moved one way. 15 of 16 participants hit navigation dead ends: no back buttons inside the flow, green CTAs that meant “navigate” on one screen and “purchase” on the next, add-ons that couldn’t be removed once added. “If none of these dates work for me, I can’t close the window. That’s annoying.”
“I picked $38, but then the next screen says $41. I’m like… wait, what’s going on?” A study participant
Recommendations, in RICE order:
Surface real prices before the funnel: a dynamic price calendar on the “Plan Your Visit” page, so out-of-state visitors see cost without entering checkout. RICE 62.5.
Decompress the calendar: one encoding system with a legend, calendar above the fold, dynamic pricing moved to the time-slot step. RICE 33.3.
Make the flow two-way: consistent back and next buttons and a sticky progress stepper. RICE 32.
Then a designed A/B test to check the biggest recommendation before Shedd builds around it: 4 weeks, non-Illinois visitors only, control versus a price calendar added to Plan Your Visit. The key metric is reduced funnel entry from that page, because the study’s sharpest insight was that a lot of “abandoning” visitors were never buying at all. They entered checkout just to find out the price, which quietly distorts every funnel metric downstream from it.
The baseline SUS score of 69.4, alongside real strengths like guest checkout and a “Get Tickets” button that 14 of 16 people found instantly, is why the recommendation is to support the existing flow rather than rebuild it. Shedd’s foundation already works. What’s leaking is specific, and it’s testable.
Sixteen remote participants, not on-site Chicago visitors. SUS data from only 13 of 16. One month of GA4. And the revenue figure is an opportunity estimate: it assumes price transparency closes the conversion gap, which is exactly what the A/B test exists to check rather than what this study alone proves. What I’d want to know next: across e-commerce funnels generally, how much of what looks like drop-off is actually just failed price discovery wearing a purchase-intent costume?
Shedd Aquarium’s out-of-state visitors pay double the local ticket price and convert 1.6 points less. Sixteen eye-tracking sessions showed why: the prices were on the page, but people’s eyes slid right past them. 94% came away unclear what their visit would actually cost. Closing that gap is worth roughly $95 to 107K in recoverable revenue every month.
Spring 2025
UX Researcher
Hridya Nadappattel, Iris Sun, Saskia Suherman, Conor Mack
UX Research
Chicago residents buy a fixed-price ticket. Everyone else gets variable pricing, and has to click through four more screens than a local before seeing a real number. Google Analytics for March 2025 put the cost of that opacity in plain sight: fixed-price visitors converted at 4.7%, while variable-price visitors, who pay $37 to 46 a ticket instead of $19.99, converted at 3.1%, despite browsing 56% more.
Research question: how might Shedd’s ticket-purchasing process better support individual and family visitors?
I ran 16 moderated eye-tracking tests (7 desktop, 9 mobile), synthesized them with a rainbow spreadsheet (16 participants across roughly 45 observed behaviors), and prioritized issues with teammate-averaged RICE scoring. Price transparency scored 62.5, nearly double the next issue.
Prices were shown, but never seen. Gaze plots showed attention scattering across the calendar and time-slot pages without ever fixating on cost. 94% of participants were unclear about pricing between the date and time steps.
The calendar demanded the most work at the worst moment. The time-slot screen logged the highest fixation count of the entire flow: 624 fixations, 243 seconds of total fixation time, because dates, times, prices, and eligibility were crammed into one view coded with both colors and icons, with no legend. “I didn’t understand what the colours were or what all the text about holiday period or discount periods were.” 14 of 16 participants hesitated here, and by the time they reached the time-slot page they were cognitively exhausted enough to skim right past per-slot price changes. “That’s why it was a very quick scroll through.”
The flow only moved one way. 15 of 16 participants hit navigation dead ends: no back buttons inside the flow, green CTAs that meant “navigate” on one screen and “purchase” on the next, add-ons that couldn’t be removed once added. “If none of these dates work for me, I can’t close the window. That’s annoying.”
“I picked $38, but then the next screen says $41. I’m like… wait, what’s going on?” A study participant
Recommendations, in RICE order:
Surface real prices before the funnel: a dynamic price calendar on the “Plan Your Visit” page, so out-of-state visitors see cost without entering checkout. RICE 62.5.
Decompress the calendar: one encoding system with a legend, calendar above the fold, dynamic pricing moved to the time-slot step. RICE 33.3.
Make the flow two-way: consistent back and next buttons and a sticky progress stepper. RICE 32.
Then a designed A/B test to check the biggest recommendation before Shedd builds around it: 4 weeks, non-Illinois visitors only, control versus a price calendar added to Plan Your Visit. The key metric is reduced funnel entry from that page, because the study’s sharpest insight was that a lot of “abandoning” visitors were never buying at all. They entered checkout just to find out the price, which quietly distorts every funnel metric downstream from it.
The baseline SUS score of 69.4, alongside real strengths like guest checkout and a “Get Tickets” button that 14 of 16 people found instantly, is why the recommendation is to support the existing flow rather than rebuild it. Shedd’s foundation already works. What’s leaking is specific, and it’s testable.
Sixteen remote participants, not on-site Chicago visitors. SUS data from only 13 of 16. One month of GA4. And the revenue figure is an opportunity estimate: it assumes price transparency closes the conversion gap, which is exactly what the A/B test exists to check rather than what this study alone proves. What I’d want to know next: across e-commerce funnels generally, how much of what looks like drop-off is actually just failed price discovery wearing a purchase-intent costume?