60-second story
From 71 findings to a prioritized action plan for a travel booking funnel
Problem
The team knew about individual UI frictions but had no reliable picture of which ones hindered offer evaluation or whether competitors handled these moments better.
Approach
15 unmoderated comparative desktop sessions: each participant used the platform under review and one of two established comparison platforms, from search to checkout.
Key insight
Users didn’t need more information everywhere. They needed the right evidence at the right moment.
Decision
I distilled 71 findings into four decision barriers and five opportunity areas. Product, Design and I prioritized 47 recommendations by user impact, effort, feasibility and dependencies.
Outcome
The findings fed into a product detail page redesign, clearer pricing on the hotel list and a multi-step mobile checkout. No isolated business impact was measured.
71
Findings and pain points
5
Opportunity areas
47
Documented recommendations
Not every pain point was roadmap-relevant
The team was aware of individual UI frictions. However, there was no reliable picture of which ones actually hindered offer evaluation and whether competitors handled these moments better.
What was needed was not another issue catalog, but a prioritization that connected user impact with effort, technical feasibility, existing planning and external OTA dependencies.
Critical uncertainty
Which frictions impaired not just usability but the booking decision and were significant enough for the roadmap?
I didn’t just test tasks. I tested competing decision logics
Each participant completed the platform under review and one of two comparison platforms on desktop, from search through checkout. The order was randomized and comparison platforms were distributed approximately equally.
Platform A
Platform under review
Search to checkout
Platform B
Established comparison platform
One of two comparison platforms per participant
Order randomized
Counterbalanced sequence
To limit potential order effects
15 desktop sessions
One additional session was excluded due to unusable audio quality
One week in 2024
Test period
Unmoderated
No spontaneous follow-up questions possible
External recruitment
Via RapidUserTesting; screening for experience with online travel search and booking
The comparison revealed a trade-off: low complexity vs. strong decision support
The platform under review often made search and navigation easier. Comparison platforms, however, provided more filters, fare variants, price details and risk information at key decision moments.
Relative strength
Less complexity at entry
The reduced interface made it easier to get started with search and navigation.
Relative trade-off
More evidence for price, value and risk
Filters, offer details and variants helped users compare options directly.
Anonymized reconstruction: the schematic screens illustrate observed UI patterns, not original interfaces and not a general superiority of any platform.
Users didn’t need more information everywhere. They needed the right evidence at the right moment
The pattern ran across four areas: price changes needed explanation, offers needed comparison benchmarks, system states needed clear feedback and before checkout, payment options and obligations had to be transparent.
01
Price and offer transparency
Price jumps without explanation undermined confidence in offer reliability.
02
Comparability
Fares and included services could not be directly weighed against each other.
03
Orientation and feedback
Filter states and inputs did not always provide clear feedback on the next step.
04
Confidence before checkout
Payment options, required fields and dense information increased uncertainty before booking.
Decision shift
From a list of isolated UX issues to four connected decision barriers along the funnel.
I turned 71 findings into a prioritized action plan
I structured observations by funnel page. Recurring patterns were inductively coded and distilled into four decision barriers and five product-level opportunity areas. Product, Design and I jointly prioritized the 47 recommendations based on user impact, effort, feasibility and dependencies.
Analysis: pain points initially structured along funnel pages. Original artifacts, blurred for confidentiality.
Inductive synthesis: recurring patterns distilled into opportunity areas. Original artifacts, blurred for confidentiality.
Evidence triangulation: findings cross-referenced with prior surveys and heuristic evaluation. Original artifacts, blurred for confidentiality.
01
Analysis
Recurring patterns were inductively coded across funnel pages and comparison platforms.
02
Synthesis
Four decision barriers were translated into five product-level opportunity areas.
03
Prioritization
We assessed user impact, effort, feasibility, roadmap alignment and external dependencies.
Result
From 47 recommendations emerged a differentiated action plan: fix immediately, bundle into a redesign, validate experimentally or consciously defer.
My responsibility
Research framing, study design, recruiting setup, execution, analysis, synthesis, recommendations, stakeholder communication and follow-up
Product responsibility
Joint prioritization, technical and external alignment and the final roadmap decision
Three product responses show how the synthesis became actionable
The study was one evidence source among several. Its strength lay in bringing specific frictions and an overarching decision problem into the same prioritization.
Anonymized product reconstruction
01
Structural issues were bundled into the product detail page redesign
Information architecture, price transparency, comparability and decision support were brought together as interconnected requirements into a larger product initiative.
Research contribution
The comparative findings helped bundle requirements; the redesign drew on additional evidence sources as well.
Anonymized product reconstruction
02
Hotel list: price information was clarified and a color concept was validated separately via A/B test
Price and offer information was presented more transparently. A revised color concept was subsequently tested against purchase conversion rate and AOV; the selected variant went live.
Validation
The experiment tested only the color concept, not the impact of the full research package.
Anonymized product reconstruction
03
Checkout: quick fixes first, structural load addressed later
Required fields were marked more clearly before input. The mobile one-page checkout, perceived as overloaded, was later converted into a multi-step process.
Research to product
The product response connected a concrete fix with a larger structural change.
The strongest impact: research became the decision foundation across three product areas
Decision impact
71 findings distilled into five opportunity areas
Product could assess user relevance alongside feasibility, effort and dependencies.
Product impact
47 recommendations translated into an action plan
Findings and recommendations were discussed, prioritized and filtered jointly with Product Management, then fed into a redesign, more transparent pricing, clearer required fields and a multi-step mobile checkout.
Validation impact
Example: one specific solution was validated experimentally
The revised color concept was evaluated against purchase conversion rate and AOV; the selected variant went live.
Evidence boundary
No isolated business impact was measured. This case therefore claims neither a causal revenue contribution nor sole causality for subsequent product changes.
What the evidence chain supports and what deliberately remains open
Supported
– 15 desktop sessions revealed recurring patterns across funnel pages and comparison platforms.
– The synthesis distilled them into four decision barriers.
– Recommendations fed into prioritization, redesign and concrete product changes.
– A color concept was validated experimentally and subsequently implemented.
Transparent boundary
– The test does not isolate which element triggered a platform switch.
– Subsequent product changes drew on multiple evidence sources.
– A separate revenue or conversion contribution was not measured.
– The long-term impact of all recommendations is not documented.
Methodological limitations
Unmoderated desktop format, no spontaneous follow-up questions, potential panel and self-selection effects, possible prior experience with individual platforms and individual price preferences.
My learning: strong synthesis doesn’t just produce findings. It produces better decision units
Users don’t decide linearly between simple and complex. They need low cognitive load at entry and sufficient evidence once they evaluate price, value and risk.
Comparison creates context
Only the direct comparison made visible when reduction helped and when it withheld decision-relevant information from users.
Prioritization creates leverage
I structured findings and insights so that Design, Product Management and Development could choose between fix, redesign, experiment and conscious deprioritization.
