Loblaws canada
Where does Loblaws lose its customers?
Industry
Grocery / Retail
Timeline
April 2026
Scope
Heuristic Evaluation, Usability Testing
My Role
UX Researcher
Team - 4 Researchers
Saadaf Mohsin, Ariana Yang, Salman Zafar
04 — Recommendations
Short-term and long-term fixes
Most of what we found wasn't a copy or button problem. It traced back to one thing: an aging technical foundation straining to keep pace with what users expect. So we split our recommendations into what design can fix now, and what needs deeper investment.
Quick wins
Surface the $30 minimum and pickup fees during browsing, not at checkout.
Right now the requirement is hidden until the final step.
"I really didn't like the $30 limit cap to reach before hitting the checkout. I had to add two random things to my cart just to hit checkout."
— Participant 8
High severity
Task 3 - Review Cart
100% completion
5.94/7 rating
Give every blocked or disabled action a visible reason and next step.
Checkout stalls without explanation when a required step is missed.
"What am I missing? It says complete steps to proceed. But what steps?"
— Participant 8
High severity
Task 4 - Checkout
75% completion
4.44/7 rating
Fix the substitution selector and add real error messaging
"Choose a substitute for me" fails with no error state.
High severity
Flagged in heuristic audit, not reached in usability testing
Make cart contents visible without scrolling
The cart page shows "3 items" at the top, but users have to scroll past promotional content before seeing what's actually in it.
Medium severity
Task 3 - Review Cart
100% completion
5.94/7 rating
Foundational
Invest in backend infrastructure to resolve 5–7 second load delays
Every cart update lagged enough that it broke the sense of a working site
"It's like doing the groceries twice, because the website is not helping much, it keeps loading and loading."
— Participant 4
High severity
Task 3 - Review Cart
100% completion
5.94/7 rating
Rebuild checkout as a guided, single-direction flow with visible progress
Location and pickup selection in particular left people guessing.
"Honestly, I'm so confused. I'll just select this. Assuming this is the closest one."
— Participant 8
High severity
Task 4 - Checkout
75% completion
4.44/7 rating
Treat performance as a UX metric, not just an engineering one
Load time isn't a backend footnote, it's the first thing users notice.
"It shouldn't be taking that long to be updating anything on a website."
— Participant 1
High severity
Task 4 - Checkout
75% completion
4.44/7 rating
Standardize layout patterns (cards, navigation, promo placement) sitewide.
Product cards and navigation shift structure across sections with no clear pattern.
"There's no design separating each item to make look quickly, I have to really focus to differentiate each item."
— Participant 4
Medium severity
Task 1 - Browse & Add
100% completion
6.25/7 rating
03 — FINDINGS
Findings, ranked by severity and frequency
Task completion & satisfaction
All 8 participants completed browsing, deals, and cart review. Checkout is where it broke.
Patterns Observed
-
Time-on-task nearly doubled between browsing and checkout. The task itself got harder, not just longer.
-
Checkout is the only task where completion dropped below 100%, and where ratings and time both got worse together.
-
The widest spread of all: Task 4 ranged from 3 to over 12 minutes. That's inconsistent, not just slow.
01 — THE PROBLEM
Loblaws' desktop site let people browse, find deals, and build a cart with almost no friction. But at checkout, 25% of participants couldn't finish on their own. That gap is what this study digs into.
02 — OUR PROCESS
Heuristic evaluation
Before talking to a single user, we audited the desktop site ourselves against Nielsen's heuristics. The same problems kept showing up: hidden requirements, inconsistent layouts, and no feedback when something went wrong.
That told us where to look, but not how much it actually hurt people. So we picked the 4 flows with the most flags: browse, deals, cart, and checkout, and tested them with real users.
Moderated usability testing
8 participants, screened to exclude UX or grocery-service backgrounds, walked through 4 tasks solo over Microsoft Teams. Each session ran 45 minutes, with a moderator and a dedicated note-taker on every call.
Affinity mapping
Every observation, quote, and friction point from the 8 sessions got clustered live in FigJam, sorted by task, heuristic, and severity.
Priority Matrix - Severity vs. effort to fix
05 — MY ROLE & CHALLENGES
My Role
Moderated 2 of the 8 usability sessions (P4, P5) and served as note-taker for 2 more (P7, P8), on top of contributing to the heuristic evaluation, affinity mapping, and priority matrix.
Challenges
Moderated 2 of the 8 usability sessions (P4, P5) and served as note-taker for 2 more (P7, P8), on top of contributing to the heuristic evaluation, affinity mapping, and priority matrix.