Case Study 04 — AI & Search

From Crazy Choices to Customised & Efficient Search

"Tailored Search: Enhancing User Experience through Customisation"

Duration 4 months
Role UX Lead
Scope AI & Search
Stakeholders Product & Tech partners
Before and after comparison of legacy search criteria versus a streamlined customised search experience
01 / Problem

Problem statement

The existing method of finding relevant information across the product was complex and not user-friendly.

Too many options and difficult navigation led to confusion and frustration — driving an increase in calls to customer service representatives.

The challenge

  • Stakeholder buy-in: convincing multiple stakeholders that so much negative feedback stemmed from users not finding relevant data
  • User bias: users accustomed to the existing behaviour
  • Collaborating with regional teams: as this change would roll out on a global scale

Solution outcome

Customised search with quick filters, recently searched queries, and advanced options

Why not reuse existing search

A fintech B2B platform requires a solution tailored to its specific needs — searching for financial transactions calls for dedicated features around transaction history, entitlement-based search for data security, and integration with existing tools that off-the-shelf solutions can't achieve.

02 / Research

Design process & user research

A design-thinking process from research through to test.

Design process — User research, Define, Ideate, Design, and Test
Design process
08
Stakeholder interviews
10
Client interviews
50
Session observations

Major pain points

No string search: All users wanted to search by a string in combination with other fields.

Cannot reuse a query: Users needed to reuse the same search multiple times a day or week.

Only one way to search: Users wanted a combination of advanced and quick filters suited to their behaviour.

Define

User personas

Two personas represented the core users of the search experience.

Persona 01

Sarah Pinoto

Account Manager — 5 years experience, uses search 3x weekly
GoalsIdentify transactions not going as expected and take action; make informed decisions from gathered data; generate daily reports.
FrustrationsToo many fields to search by; most journeys start with a search that eats up most of her time; can't tell if a search is already applied or where to update it.
NeedsSimple, robust search that's flexible at any time, reusable, and customised per module.
Persona 02

Daniel Lewis

Client-facing team — 6 years experience, daily usage
GoalsMove fast in helping users with queries; guide users properly to what they're looking for; fetch data quickly without breaching corporate data.
FrustrationsUsers wait a long time to find data; can't search with minimal information; unsure about recent changes and why they happened.
NeedsA quick, efficient way to search with a high-level view, able to guide users with minimal effort.
User flow
03 / Ideation & design

Brainstorming

We explored various options for how to approach this solution, narrowed down to one direction, and explored it further through brainstorming and selected-solution ideation with the wider UX community.

Low-fidelity wireframes
Ideation

Final designs after testing

After all the brainstorming we finalized on one concept and developed it further with all the connected user journeys and a clickable prototype.

Usability testing

After narrowing down to 1 concept we conducted usability testing with 10 users and based on the result made all the required changes.

04 / Delivery

Documentation & handoff

To keep collaboration seamless, we shared behaviour documents, scenario-specific specification documents, and usage documentation with both the dev and UX teams to ensure consistency across the product.

Rollout & next steps

To make the pattern more user-centric long term, we proposed rolling it out in multiple phases — so the same search pattern could be reused across products and eventually become a global, product-level search.

Less is more. Allow users to see information on a need basis.

Thank you for reading

Explore more case studies

Back to Case Studies