CASE STUDY

Designing search across the website and app

ROLE

Lead Product Designer

TIMELINE

Sept 25 - Jan 26

PLATFORM

Web & App (Mobile-first)

Search is the feature that works best when you don't notice how good it is.

Search is the feature that works best when you don't notice how good it is.

Search is the feature that works best when you don't notice how good it is.

'Dry skin, bakuchiol, or "the blue one"': this project was about designing search for how people actually shop.

'Dry skin, bakuchiol, or "the blue one"': this project was about designing search for how people actually shop.

'Dry skin, bakuchiol, or "the blue one"': this project was about designing search for how people actually shop.

Overview

Overview

Overview

On Rituals website, over 11% of customers use search, with those customers converting at roughly 5x the rate of non-search users.

During Rituals' move to a headless architecture, I led the end-to-end redesign of the search experience across web and app. I managed stakeholders across development, commercial content, customer support, and product teams. The challenge was to support product findability, inspiration, and customer support while working across multiple platforms and content types.

The approach

The approach

The approach

I mapped search behaviour through customer interviews. This revealed three distinct patterns:

  1. Customers who knew exactly what they wanted

  2. Customers using search to explore the range

  3. Customers looking for information rather than products (returns, delivery, loyalty support).

To create alignment early, I facilitated a cross-functional sketching workshop involving designers, engineers, and business stakeholders. Beyond generating ideas, the session helped align teams around requirements, surfaced constraints early, and created shared ownership of the solution.

The preliminary work wasn't visual. It was defining behaviour:

  1. How should suggestions be ranked?

  2. When should search predict intent?

  3. How should editorial content appear alongside products?

  4. What should happen when no results are found?

These decisions required close collaboration with development to balance customer needs, business goals, and the back-end platform's capabilities.

Testing

Testing

Testing

Through multiple rounds of testing, I evaluated whether customers could find what they expected and whether search results felt trustworthy.

One key finding was that customers rarely discovered support content, FAQs, and editorial content as early as intended. As a result, I revised how non-product content was surfaced throughout the experience rather than treating it as secondary content within results.

WHAT GOT BUILT

The Outcome

The final redesign covered the full search experience across web and app: the open state, predictive suggestions, active search, results page, and no results page.

First-click | autocomplete

Results | autocomplete

Search | results page

Edge cases | loading state, no results, content-only results

The response

The response

The response

Customer feedback since the feature launched on the app has been direct:

"Easy to use, search function is very good!"

— NL Customer, November 2026

“Search function works great, quick to product."

“Search function works great, quick to product."

— NL Customer, September 2026

My reflections

My reflections

My reflections

Designing search well is an exercise in designing for edge cases as much as the happy path.

The customer who finds exactly what they wanted on the first try doesn't need much from you. The one who types something ambiguous, or something the catalogue doesn't recognise, or who arrives via search looking for help rather than a product, that's where the design earns its place.

Teams: Design, Trading & Commercial, Technical Teams, Algolia, SEO