← Back to work
UX Research · Ecommerce · 2026

Ecommerce Analysis WM

Digital Tours research study across 16 sessions in Costa Rica and Guatemala, diagnosing why Walmart CA's ecommerce CSAT scores 6.4 vs the 8.9 industry benchmark — and identifying the 4 critical bugs behind purchase abandonment.

16
Sessions
9 CR · 7 GT
6.4
WM CSAT
vs 8.9 competition
4
Critical Bugs
blocking checkout
−3.9
CR Gap
pts vs benchmark
Context

Why is Walmart CA's e-commerce underperforming?

🌎

The Study

Estudio Experiencia — Tour Digitales 2026 was commissioned to understand real end-to-end shopping flows in two key markets. Sessions were conducted as moderated Digital Tours — guided shopping tasks performed by participants on Walmart's live ecommerce platform, recorded and analyzed for friction points, task completion rates, and sentiment.

📍

Two Markets, Two Stories

Costa Rica (9 sessions) and Guatemala (7 sessions) painted starkly different pictures. While GT users showed relatively positive satisfaction (CSAT 8.1), CR users averaged only 5.1/10 — a full −3.9 points below the competitive benchmark. Understanding why required going beyond surveys: watching real purchases break in real time.

CSAT Results

The satisfaction gap

Overall Walmart CA CSAT: 6.4/10 against a competitive benchmark of 8.9/10. The split between markets tells the real story.

🇨🇷
Costa Rica
9 sessions
5.1
Repurchase intent (high)11%
Repurchase intent (low)56%
Gap vs benchmark−3.9 pts
🇬🇹
Guatemala
7 sessions
8.1
Repurchase intent (high)57%
Repurchase intent (low)0%
Gap vs benchmark−0.6 pts
Total WM CA6.4 / 10
0Benchmark: 8.9 ▲10

Key Insight

16/16 sessions were affected by the associate discount bug — the single most widespread failure across both countries.

Methodology

Digital Tours — Seeing real purchases break

1

Session Design

Structured 16 moderated sessions around a standardized task flow: browse product categories, add items to cart, apply discounts/coupons, proceed to checkout, and complete payment. Each session was screen-recorded and annotated for friction events.

2

Digital Tour Execution

Participants navigated the live Walmart CA ecommerce platform on their own devices. Researchers observed in real time via screen share, noting hesitations, error encounters, and workarounds without intervening — allowing authentic failure patterns to surface.

3

Post-Reception Analysis

A subset of 5 participants received their orders and were followed up on product quality, price accuracy, and order completeness. This validated whether issues were UX-layer problems or deeper supply-chain failures.

4

Competitive Benchmark

Parallel sessions evaluated Uber Eats, Amazon, WM USA / Canada / Mexico, PriceSmart, and local competitors across the same task flow. CSAT scores and friction taxonomies were compiled to produce a gap analysis.

Session Breakdown

Costa Rica (CR)9 sessions
Guatemala (GT)7 sessions
Post-reception follow-up5 participants
✓ Quality OK
✓ Prices OK
⚠ 1 incomplete order
Critical Findings

4 bugs blocking purchase completion

The Digital Tours revealed that low satisfaction wasn't caused by product quality or pricing — it was caused by four specific technical failures at checkout.

1

Misleading tiquete/factura error message

Users selecting factura (tax invoice) received an error indicating the option was unavailable in Costa Rica — but the message was confusing enough that users interpreted it as a payment failure. CR-only issue.

3 purchase abandonments observed across sessions
2

Associate discount not applying

Employee discount codes failed to apply during checkout in all 16 sessions across both CR and GT markets. The UI accepted the code without error but the cart total remained unchanged.

16/16 sessions affected — both countries
3

Coupon 7AHORRO broken

The promotional coupon code 7AHORRO consistently failed validation in Costa Rica sessions. Users who entered it received a generic 'invalid code' response despite the promotion being actively advertised.

CR sessions only — erodes trust in promotions
4

Frozen payment screen

During Costa Rica sessions 5 and 9, the payment confirmation screen stopped responding after form submission. No success or error feedback was displayed, leaving users uncertain if the purchase was completed.

2 CR sessions — complete checkout block
💡

The Core Insight

Post-reception results showed 5/5 participants rated product quality as acceptable and 5/5 found prices correct. Only 1 order was incomplete at delivery. The problem was never the product. It was the checkout bugs. Fixing these four issues would address the majority of the CR satisfaction gap without changing a single SKU.

Competitive Benchmark

How the market compares

Sessions evaluated Walmart CA against regional and global ecommerce players across the same Digital Tour task flow and CSAT methodology.

Competition avg.
8.9
PriceSmart
9.1
Uber Eats
8.8
Amazon
8.7
WM USA
8.5
WM Mexico
8.3
WM Canada
8.2
Walmart CA (GT)
8.1
Walmart CA (total)
6.4
Walmart CA (CR)
5.1
Recommendations

Prioritized remediation roadmap

P0 — Immediate
  • Fix frozen payment screen (CR)
  • Fix associate discount logic (CR + GT)
  • Repair 7AHORRO coupon validation (CR)
P1 — Short term
  • Rewrite tiquete/factura error message copy
  • Add real-time discount feedback in cart
  • Implement payment confirmation toast
P2 — Medium term
  • Repeat Digital Tours post-fix (CR focus)
  • CSAT tracking dashboard (weekly cadence)
  • Competitive parity review quarterly

Expected impact after P0 fixes

5.1 → 7.5+
CR CSAT projection
+45%
CR repurchase intent
0
checkout abandonments from bugs
16/16
discount sessions resolved
Learnings & Reflection

What this study taught me

Running research at this scale — live sessions, two markets, real checkout flows — surfaced things a survey would never catch.

CSAT alone doesn't explain "why"

A 5.1 score is a symptom. Only by watching real sessions did we see that checkout bugs — not product quality or pricing — were the actual cause of abandonment.

Market segmentation reveals hidden signals

The CR/GT split (5.1 vs 8.1) showed the problem was localized, not systemic. Without segmenting by country, the aggregate 6.4 would have hidden the severity in Costa Rica.

Post-reception validation eliminates noise

5/5 post-purchase participants found quality and price acceptable. This ruled out supply-chain as a cause and pointed squarely at the UX layer — a crucial finding for prioritization.

Observational research catches what surveys miss

No survey would have revealed a frozen payment screen or a discount code silently failing. The Digital Tour format — watching real users on live systems — made invisible failures visible.

Protected under NDA

Due to legal and privacy arrangements established in the Walmart employment contract, the full research report, session recordings, raw data, and internal analysis documentation cannot be shared publicly. The findings presented on this page represent a curated summary approved for portfolio use.