खोजी
Reimagining the In-Store Shopping Experience
A proof-of-concept AR navigation & smart checkout system designed to eliminate friction inside large urban supermarkets — from aisle confusion to billing queues.
TOOLS
Figma · Maze · Miro
TYPE
POC · Mobile App
TIMELINE
8 Weeks
ROLE
UX Researcher & Designer
खोजी
Reimagining the In-Store Shopping Experience
A proof-of-concept AR navigation & smart checkout system designed to eliminate friction inside large urban supermarkets — from aisle confusion to billing queues.
TOOLS
Figma · Aftereffects · Blender
TYPE
POC · Mobile App
TIMELINE
8 Weeks
ROLE
UX Researcher & Designer
SCROLL
People weren't struggling to shop.
They were struggling to find.
"What should've been a 20-minute grocery run took me over an hour. I still missed half my list."
— Observed shopper, Mumbai hypermarket
It was a Saturday morning. The supermarket was packed. People zigzag-ed through aisles, looking at boards overhead, carts bumped, voices overlapped, and the air felt heavy with impatience. What should've been a simple grocery run had turned into a maze of confusion and chaos.
That small difference — between shopping and finding — created a big problem. Not just for the shoppers, but for the business too. Longer dwell times, abandoned carts, frustrated customers, and declining repeat footfall were all symptoms of the same root cause: navigation failure inside the store.
Khojee (खोजी — Hindi for "the one who seeks") was born from this observation: what if your phone could guide you through a supermarket the way Google Maps guides you through a city?
Five forces making shopping miserable
Difficulty Locating Products
No clear aisle mapping or digital directory
Inaccurate Staff Assistance
Staff unsure of product locations
Overcrowded Aisles
Peak hours create physical congestion
Abandoned Purchases
Frustration leads to cart abandonment
Slow Checkout
Long queues erode purchase intent
👁
Shadow Observation
Followed shoppers with consent using natural shadowing across 3 hypermarkets. Captured navigation patterns, hesitation points, and emotional reactions without interference.
🎤
Intercept Interviews
Short 5–8 minute conversations post-shopping to understand their mental model of the store and what moments caused the most friction.
👥
Staff Interviews
Spoke to floor staff and managers to understand their perspective — how many direction queries they handle per hour, and what gaps they felt in current signage.
At the site
Together with colleagues, I used natural shadowing to observe users as they shopped — a bit like being undercover cart detectives. It may sound sneaky, but it revealed genuine frustrations and patterns that surveys alone could never capture.
It was quite useful to understand their pain points — right from the aisle confusion to the "where on earth is the atta?" moment. We clocked average navigation time of 18 minutes for a 12-item list — nearly 1.5× longer than it should've been.
Getting out of the building
Desk research can only take you so far. To understand real friction, we had to be where shoppers were — in the aisles, at the billing counters, watching the chaos unfold in real time.
The problem
AVG. TIME WASTED NAVIGATING
~18 min
SHOPPERS WHO SKIP ITEMS
58%
REPEAT VISITS LOST DUE TO FRUSTRATION
19%
Urban shoppers in large supermarkets often spend excessive time searching for specific products due to poor signage, overcrowded aisles, and limited staff assistance during peak hours.
This not only leads to customer frustration and longer shopping durations but also results in potential loss of sales — as people avoid or abandon crowded stores. There is a need for a seamless, technology-driven solution that simplifies and optimises the overall shopping experience.
Spectrum of friction
Across interviews and observations, pain points clustered into four critical areas — each with both shopper and business-side impact.
BUSINESS IMPACT
These behavioural patterns translate directly into measurable revenue loss — lower basket sizes, fewer repeat visits, and increasing preference for online grocery alternatives.
walk away without purchase
58%
28%
said they prefer quick online grocery shopping
The journey
Shoppers enter with a clear list in mind but quickly hit friction — unclear store layouts, inconsistent aisle logic, and hard-to-spot products slow them down. Frustration compounds at every stage.
GOAL
Buy all items on the list efficiently
PAIN POINT
Enters without map; relies on memory
EMOTION
Curious, hopeful
GOAL
Help with brands and aisle numbers
PAIN POINT
Can't see upper shelves; aisle logic unclear
EMOTION
Mildly confused
GOAL
Less wait time
PAIN POINT
Too-long billing queue; waste of time
EMOTION
Impatient, stressed
GOAL
Not to forget anything
PAIN POINT
Forgets half the things bought; misses items
EMOTION
Dilemma
GOAL
Good follow-up if needed support for missing items
PAIN POINT
Took too long; items out of stock
EMOTION
Mixed
Key insight from journey mapping: Inefficiencies in navigation, product discovery, and checkout flow directly undermine both the shopping experience and store performance — proving the need for a smarter, guided in-store system that works at every stage.
What the data told us
01
58% of customers walk away
Long queues significantly impact sales. 58% of customers walk away without completing their purchase, and barely 19% come back later — making wait-time one of the biggest drivers of lost revenue.
02
Crowding triggers non-return
The moment checkout lines feel crowded, shopper discomfort rises sharply, and many opt not to return — proving how critical queue management is for retaining customers. Our observation showed this tipping point at queues of 6+ people.
03
Basket size shrinks in crowded stores
When a store feels crowded, shoppers tend to explore less, avoid impulse purchases, and cut their visit short — ultimately reducing average basket size by an estimated 15–22% and negatively affecting the store's overall perception.
04
Staff time spent on directions: 34%
Through staff interviews, we found that floor staff spend approximately 34% of their shift answering directional queries — time that could be used for restocking, customer service, and inventory management.
Khojee — the fixes
Rather than redesigning the physical store (expensive, slow, limited), we designed a phone-first companion that layers intelligence on top of the existing environment. Four features addressing the four biggest friction points.
Smart Shopping List Integration
Import or build a list. Khojee auto-sorts it by store aisle sequence, so you always walk an optimised route — no backtracking. Integrates with WhatsApp lists via share.
Smart Search & Shelf-Level Accuracy
Search 'Amul Butter 500g' and Khojee shows you not just the aisle but the exact shelf position — left side, third shelf from top. Reduces product search time by 30%.
Self Check-Out
Scan items as you shop. When done, generate a QR code and pay at the kiosk — skipping the traditional billing queue entirely. Reduces average checkout time from 9–10 minutes to 4–5 minutes.
AR-Based Navigation
Point your camera at any aisle and Khojee overlays a live path to your next item. No more asking staff, no more zigzagging. Uses indoor positioning to guide users turn-by-turn within the store layout.
The app in action
The Khojee interface was designed to feel familiar — grounded in the mental model of a maps app but tuned for the supermarket context. Three primary screens handle 80% of use cases:
Home: personalised greeting + smart shortcuts
Search: item lookup with shelf-level accuracy
AR View: live overlay navigation
Self Checkout: scan-as-you-go flow
How the self-checkout flowed
Self-checkout was the most complex flow to design — it needed to handle errors gracefully, support partial lists, and never make users feel trapped. Here's how we mapped it.
Sign In
Phone number or loyalty card
Home Screen
List imported or created fresh
Cart Finalised?
Decision: ready to scan or still shopping
Search for Item
If not found → Ask for help (?) prompt
Scan Product Barcode
Camera-based scan; auto-adds to cart
All Products Added?
Loop until list complete
Verify Cart
Review items + quantities
Cart Ready to Checkout?
Option to continue shopping
Pay
UPI / card / wallet options
Download Receipt
PDF or WhatsApp share
Receipt downloaded & ready?
Confirmation gate before exit
Exit
Show QR code at gate
.png)
Reduction in time spent locating products
With AR navigation, product discovery time dropped to ~6–9 minutes from a baseline of 14+ minutes.
↓
30%
Reduction in checkout time
With self-checkout, shoppers completed billing in nearly half the time — dropping from 9–10 minutes to 4–5 minutes.
↓
25%
Increase in average basket size
Faster product discovery and reduced in-store fatigue meant shoppers picked up 1–2 additional items they might otherwise have ignored.
↑
10%
↓
18%
Drop in abandoned purchases during peak hours
Fewer abandoned purchases directly translated into higher completed sales during high-traffic periods.
Numbers that moved
After a controlled in-store POC trial across two hypermarket locations, the results validated our hypothesis. Every major friction point improved meaningfully.
SHOPPER SENTIMENT
Would use Khojee again
87%
Felt less stressed shopping
79%
Preferred self-checkout
72%
QUALITATIVE FEEDBACK FROM TRIAL
Physical context matters
Designing for physical environments is fundamentally different from screen-only products. Lighting, noise, crowd density, and physical fatigue all shape how users interact with a digital overlay. I had to test in-store, not just in a lab.
The best UX is invisible
When Khojee worked well, users didn't notice it — they just found their items. The moment they had to think about how to use the app, we'd already failed. Every friction point in the digital layer compounded the physical one.
Business and UX goals align
This project reinforced that good user experience is good business. Every metric that improved for the shopper — faster navigation, quicker checkout — also translated into a direct revenue improvement for the retailer. Empathy scales.
What I learned
What's next for Khojee?
The POC validated the core hypothesis. A full product would need deeper retail partnerships for real-time inventory APIs, indoor positioning infrastructure (BLE beacons or WiFi triangulation), and a store-side dashboard for inventory management. The opportunity is significant — India's modern retail market is growing at 15%+ annually, and the in-store experience hasn't kept pace.
Multi-store onboarding
Voice-guided navigation
Personalised deal alerts
Retailer analytics dashboard