Dell search is an AI-powered gateway that helps customers quickly find the right product across a vast and complex catalog, without friction or guesswork.
Gen AI Search at Dell
Role
User Interface Sr Designer
Team Structure
2 Sr Principle Lead Designer
1 Sr Lead Designer
1 User Interface Sr Designer.
Tools
Design: Figma
User Testing: User Testing Platform
Collaborations: Teams
My Responsibilities
Design
Competitive Research
Ideations
Rapid Prototyping
Micro Animations
User Testing & Synthesis
Design Hand off
Cross-Collaborations
Stakeholder presentations
Dev Sync ups
Metrics Achieved
Overall AI Overview engagement is
42.15% CTR
Search at Dell was not optimized for the scale and complexity of its product ecosystem. Disconnected experiences, limited contextual filtering, and inconsistent result relevance were impacting discoverability and slowing user decision-making across key business segments
Context
Goals
Hence, we started working on the search unified vision
UX Goal:
Prioritize high-traffic, high-intent keywords to rapidly design, launch, and validate optimized search experiences through continuous testing.
-
Business Goal:
Start with a focused, high-impact MVP and scale iteratively to unlock new revenue streams and measurable growth.
-
Customers struggle to find the right product quickly.
O1
Search Intent is unclear or ambiguous
-
Customers use vague or imprecise queries leading incomplete or inaccurate results
-
Customers struggle to phrase searches effectively or match product terminology.
O2
Limited Understanding of product benefits and Use cases
-
Customers need help in finding products for specific scenario.
-
Unclear product features and benefits creates uncertainty.
-
Lack of smart recommendation leads to missed bundles and add ons opportunity.
-
Uncertainty about specs, compatibility or fit delays purchase decisions.
O3
Decision Fatigue
-
Too many product choices can overwhelm users, causing them to abandon their search.
-
Users may need guidance on filtering and narrowing down their results.
-
Users struggle to differentiate between similar products with slight variations.
We identified near-term Gen AI use cases to determine the funnel stage and guide users through it.
Mapped search keywords to the user journey (purchase funnel)
Market research to shape our ideations later
Insights were distilled into key themes and how they can be implemented in Dell-Al-Powered Search, Visual Search, Personalization & Smart Suggestions, Save Search, Augmented Shopping, Conversational Al & Voice, Ecosystem Integration, and more.
Out of which we focused on MVP focused key themes such as -
-
Al-Powered Search
-
Conversational Al
The Search interaction model was defined to visualise the flow of components better...
We looked at consumer behaviour and personas
Meet Victoria , who is small business owner looking for a budget friendly laptops to buy.
"Need super powerful laptop which can helps me scale my business work loads without blowing up my budget" - Victoria
User Journey
The purchase funnel and components were mapped through the user journey for ideating a seamless browsing user experience.
Concepts: For the onboarding screen
Ideation 1:
The VA response with prompts with images.
The images adds up a visuals cue to users to jump into the respective catalog.
Ideation 2:
The VA response with prompts with images with Deals context in mind
Ideation 3:
The VA response with prompts with images with trending products context in mind
Ideation 4:
The VA response with prompts without images
While the initial scope was AI-assisted search refinement, we recognized the opportunity to influence the broader browsing architecture.
The design components were structured to scale across entry points such as - search, category pages, and product comparison.
Defining the search experience
I presented the finalised concept to the leadership to narrow down the MVP concept and check on any refinements needed further for us to closely gauge the search user journey.
Building the Top AI Panel Experience
As users types in and searches of a keyword , The AI Panel loads up on top the page as a AI blurb.
It can be expanded and collapsed as per user’s preference without getting in the way of the whole page or flow of the user.
The AI Panel on top enhances a user’s search experience by only providing a summary view with relevant data which will help and empower our users to take inform decisions.
Read more/less
Expands and Minimises the panel content to user’s need.
Contextual Prompts
Ideally was to refine the Virtual assistant response within this panel.
But due to technical capabilities being limited for the MVP. The concluded decision was to open up the Virtual Assistant Chat Panel on the side for the user to futher interact with in a conversational AI mode, while shopping.
Virtual Assistant Response
Data gets populated upon user’s search keyword.
Open/Hide Panel
Collapses the whole panel on top of the page
Main CTA
Which opens up the side panel to
chat further with the Virtual Assistant.
Concept User Testing & Results
All our tests indicated that the top assistant significantly enhanced users' search experience. In this test, we asked the users to engage with the Top Al panel.
One challenge was ensuring AI-generated prompts were in parity with search results.
While user testing showed positive engagement, we identified a structural backend limitation that caused some mismatch. We escalated this as a trust risk and advocated for backend alignment. Ultimately, we shipped with phased refinement, but the experience reinforced the importance of AI transparency and system coherence.
Production Metrics
MVP Production Metrics with 100% Traffic ( After US Launch)
Post-launch, engagement CTR improved to 42.15%, and we saw measurable reduction in search abandonment.
More importantly, users reported higher confidence in product selection, which aligned directly with Dell’s goal of improving high-consideration purchase journeys.
Refinements
We further ideated and tested post launch design enhancements for refinements and scaling the search experience.
Post launch
