Shushant Kumar — Product Design Manager

All work

Cleartrip Sr. Product Designer Dec 2022 — Mar 2023 Shipped June 2023

Redesigning international flight search

Long-haul search was losing people at the moment they tried to compare. Three shipped iterations later, transactions were up 23% and the total value of international bookings up 200%.

Role
User flows, explorations, interactions, visuals, prototyping
Team
1 design manager · 1 product manager · 6 engineers
Research
20+ user interviews and an online study, run with my PM
Tools
Figma, Principle
Outcome
+23% transactions · +200% booking value

The call I made

Start at the result card, not the search form. The funnel said the search result page leaked hardest; the search form is where everyone has an opinion. I took the argument to the data and spent the first phase where the drop actually was.

The harder call was subtraction. Amadeus exposes dozens of fields and every function had one they wanted on the card. I defended a short primary list — price, times, airports, layovers, airline, duration, seats-left — and pushed the rest into the expanded view, with product, business and engineering in the room.

I owned

  • User flows, explorations and interactions
  • Visual design and prototyping
  • The information hierarchy argument

I partnered on

  • User research, with my product manager
  • Feasibility with six engineers, weekly
  • Direction with my design manager

Where the judgment was

  • Choosing the funnel step to fix first
  • Cutting fields from the card, with the business in the room
  • Shipping and reading one iteration before starting the next
Final international flight search results
The shipped international search results page.

Context

Cleartrip is a website and app that makes it easy for travellers to plan and book trips — flights, hotels and bus rides in one place. Booking an international flight means naming an origin and destination, picking a date, and then choosing among the flights different airlines offer at different times and prices.

If you have ever searched for a flight, you know the song and dance. If I change the date, will it be cheaper? If I fly from a different airport, will the flights be less full? The hunt becomes a loop of viewing results, refining, and changing the criteria again. Domestic search had been tuned for years; international inherited the same interface without inheriting its assumptions.

What was going wrong

Six failures on the search result page, confirmed against the funnel.

Backstage: what the traveller doesn't see

Third-party booking systems. Most flight search runs on a reservation system — Amadeus — which is mostly responsible for which data points are available at all. Airlines can only exert limited influence on a better solution.

Dynamic pricing. Prices are recalculated and availability checked on every search query. In our test, a query usually took ten seconds, which is a long waiting time in observed search behaviour.

Cost per request. Searches are charged per request, so airlines want to reduce search requests — which pushes against both pre-emptive and iterative queries.

Flight searches are structured in such a way that it is unlikely a customer finds a suitable flight straight away, because it presupposes the traveller has entered all price-relevant information before submitting the query. That is the real problem, and it sits upstream of any screen.

What travellers told us

The goal was to improve international booking conversion by 200–300 basis points (relative) from search results through to charged. To find out what was wrong, we ran 20+ interviews with different types of traveller, covering platform preferences, the booking process, desktop versus mobile behaviour, and how a flight fits into planning a trip.

  • The flight gets more expensive with every search. Opaque pricing makes travellers suspicious of cookies and tracking.
  • Flights are a needle in a haystack. One route can mean hundreds of options across dozens of providers.
  • Long waiting times are nerve-wracking. Ten seconds a query, repeated, is the whole experience.
  • The quest for the best deal. Price is still the most important decision criterion — and the lack of transparency undermines trust in it.
  • The feeling of having paid too much. Cheaper flights are often only available on limited dates or in limited numbers, so every search feels like a possible mistake.

Challenge 1 — where to spend the effort

Evaluating each funnel segment before choosing what to fix.

Funnel stepStep-through
Home page → Search results31%
Search results → Itinerary17.4%
Itinerary → Payment17.6%
Payment → Success8.5%
Step-through rates on the international flight funnel before the redesign.

We prioritised needs changes based on research and current statistics: the search result page is where the maximum drop in the funnel occurs, so that is where the first phase went.

Challenge 2 — understanding competitors

A pass over the leading players to see which patterns travellers already read fluently.

Googleflights logo Expedia logo Booking logo Flightnetwork logo Skyscanner logo Priceline logo Lastminute logo Blinctrip logo
Studied for information hierarchy, layout and features — what to adopt, and what was inherited clutter.

Challenge 3 — deciding what is primary

Amadeus makes dozens of fields available: airline name and number, aircraft type, class, departure and arrival city and time, layover count and duration, terminal numbers, baggage rules, transit visa requirements, refund type, price cut, seat left, offer visibility, distance from a nearby airport. Everyone inside the company has a field they want on the card.

So we settled it explicitly, with product managers, business representatives and engineers. Primary on the card: total price and coupon callout, departure and arrival time and date, departure and arrival airport, number of layovers, layover type, airline name, total duration, seats-left callout, layover airport. Secondary: no baggage, travel visa requirement, terminal change, airport change, refund type, more options at the same price, price breakup.

In the expanded view the priority inverts toward verification — segment-level airline, airport and time, leg duration, stopover duration and airport, terminal and airport change, travel visa requirement, baggage check-in and cabin, cancellation and amendment info.

How the design got there

Three iterations. Each shipped and read before the next started.

Iteration 1 — redefining the result card

The old card confused users. Departure, arrival and duration times were easily conflated because of their similarity, potentially leading to oversights. It was challenging to intuitively grasp airport codes, unfamiliar icons made the design hard to work with, and it was difficult to comprehend and navigate seamlessly.

I collaborated with stakeholders, presented the problem and the approach, and through constructive dialogue we agreed a direction. Over two days I went through more than ten iterations of the card, then walked three or four selected versions through to engineering for feasibility. Four considerations survived: time clarity — departure and arrival positioned side by side, with layover information made prominent; icon and text integration — labels alongside icons, so comprehension doesn't depend on guessing; transparent pricing — the search page shows the final amount after applicable coupons, not a number that changes later; and comparative display — flights at similar prices shown together, with variations in stops or duration highlighted so the trade-off is an informed one.

Before
The old one-way and round-trip cards

The old one-way and round-trip cards

After
The redefined cards, with final price and seats-left

The redefined cards, with final price and seats-left

Iteration 2 — filter and sort

The old system presented challenges in both filter navigation and comprehension: the arrangement and depth of filters, plus difficulty understanding and using sorting options, which lacked intuitiveness and visibility alike.

Recognising that users typically think in terms of speed and cost, I implemented sort options that let them arrange flight sequences on those two axes — fastest and cheapest. Then I strategically prioritised primary filters: stops, airlines and time of flight, the ones used most frequently, moved upfront. Streamlining the filtering process addresses both the comprehension and the visibility concerns raised in the earlier evaluations.

Before
Filters and sort before

Filters and sort before

After
Sort reduced to two axes, frequent filters promoted

Sort reduced to two axes, frequent filters promoted

Iteration 3 — the expanded view

The old expanded section had several usability issues. Excessive scrolling was required to access information within the in-page expanded section, which made it hard to read and compare two flight options. The airport code shown in the layover lacked clarity. Crucial information such as cancellation and amendment policies, or baggage details, was hidden behind non-intuitive prompts. Overall, the expanded view took up considerable space while providing questionable utility.

I moved the detail into a side sheet, so opening one option no longer means leaving the comparison. Information scanning improved, layover information and baggage details became clearer, and a dedicated fare rules tab put cancellation and date change policies upfront — where they empower a decision rather than explain a mistake.

Before
The old in-page expanded section

The old in-page expanded section

After
The side sheet, with a fare rules tab

The side sheet, with a fare rules tab

Working through it

Design review session with the team
Reviewing card iterations with engineering
Reviewing tuple iterations with product and engineering — weekly through implementation.

Outcome

Launched the new version by the end of June 2023.

200%Increase in total value of international flight bookings
23%Increase in number of transactions
5.2%Increase in charged users
32.7%Of charged users interacted with the expanded card
Funnel stepBeforeAfter
Home page → Search results31%73%
Search results → Itinerary17.4%45%
Itinerary → Payment17.6%23.7%
Payment → Success8.5%16.8%
Step-through rates after launch, against the pre-redesign baseline.

The shipped experience

Results with low-cost fare comparison
Results with fare inclusions
Fare rules and price breakup on mobile
Search, comparison and fare rules across web and app.

Closing notes

Initially the team wasn't entirely sure about making changes — thinking we were doing well in the international market. When we looked closely at how many people were actually completing their bookings, we were in for a surprise. The numbers showed a 200% increase in the total value of international flight bookings once the redesign shipped, which convinced us there was a real opportunity to grow in that market. We decided to invest more in the non-domestic market, especially in making international flight experiences even better.

Happy to walk through the parts that didn't work, too.

shushant0657@gmail.com

Open to full-time roles and freelance projects