Kroger
2026
Kroger Internship/ MCSX Team
Co-op 2026 Project
Overview
Understanding Kroger MCSX Team
Kroger's MCSX team wanted to understand where self-checkout was breaking down for customers, specifically around BOGO deals, coupon redemption, and loyalty sign-in and identify where friction was costing them completed transactions.
As UX Researcher on the MCSX team, I ran a primary survey (n=100) alongside walk-a-mile immersion research and affinity clustering to surface the core pain points, then translated those findings into kiosk flow prototypes addressing scanning, deal application, and account sign-in.
Problem
Where Self-Checkout Breaks Down
Kroger shoppers are loyal and habitual deal-seekers, 89% always or often look for savings, and 88% shop in-store. But that habit doesn't translate smoothly through checkout. Multi-item deals like BOGO caused the most friction of anything we tested.
Survey Findings (n=100)
87%
Encountered multi-item/BOGO deals but only 49% rated clarity a 5
32%
Experienced an actual issue redeeming a deal
78%
Rated overall satisfaction 4–5, yet frustration is bimodal: 45% low, 29% high
Paint Points Identified
Multi-item/conditional deals fail at checkout
Confusing wording and signage
Deals take too much effort to useConfusing wording and signage
Uncertainty about how many items are needed
Low confidence going into checkout
Price-layering complexity in-store
Goals
Designing for Business & User Value
Project Goals | User Goals |
|---|---|
Reduce checkout failures caused by unclear multi-item/BOGO deal requirements | Know exactly how many items are needed to qualify for a deal, before reaching checkout |
Close the gap between deal confidence (79% find deals easy) and checkout confidence (only 49% rate clarity high) | Trust that a deal will actually apply when scanned |
Surface deal status and item requirements earlier in the flow, not just at checkout | Get clearer in-store signage and simpler wording (the top two requested fixes, at 49% and 34%) |
Support customers across channels (app, signage, checkout) consistently, since shoppers stitch deals together across touch points rather than relying on one | Feel confident going into checkout instead of uncertain |
Research & Insights
Understanding the Opportunity
To understand where self-checkout was failing customers, I ran a primary survey (n=100) covering deal awareness, digital coupon usage, and checkout confidence, then synthesized the open-ended responses into affinity clusters to identify root causes rather than just symptoms.
Key Insights
01
Deal confidence doesn't carry through to checkout
Shoppers are confident finding deals (79% rate it easy) but far less confident meeting requirements before checkout (only 49% rate clarity high). The gap sits between discovery and redemption, not discovery itself.
02
Failures are structural, not user error
Of the 32% who hit an issue, the top causes were the deal not applying at checkout (43%) and not knowing how many items qualified (26%). Problems with the system's clarity, not shopper mistakes.
03
Digital coupons work; awareness doesn't
78% have used digital coupons and rate them easy to find (79%) and use (88%). The 22% who haven't cite awareness gaps (45% "didn't know about them") An onboarding problem, not a design problem.
04
What customers say would fix it
Clearer in-store signage: 49% Progress indicators (e.g. "1 more item needed"): 40% Simpler wording: 34% Better app experience: 33% More associate support: 27%
My Approach
Turning Research into Actionable Design Direction
To move from raw survey data to something actionable, I followed a structured research process designing the questionnaire, fielding it, then synthesizing responses through affinity clustering to separate structural failures from one-off complaints.
Throughout the project, my focus stayed on identifying where the checkout experience broke trust, and translating those breakdowns into clear, testable design directions the MCSX team could act on.
Process
🔍 Research
Designed and fielded a primary survey (n=100, 18 questions) covering deal discovery, digital coupon usage, and checkout confidence across Kroger's shopper base.
💡 Define
Synthesized open-ended and rated responses through affinity clustering to separate structural failures (deal didn't apply, unclear item counts) from awareness gaps (didn't know about coupons).
🎨 Design
Translated the top requested fixes, clearer signage, progress indicators, simpler wording, into kiosk flow prototypes (start-scanning → deal application → sign-in) addressing where confidence dropped in the checkout process.
Validate
Putting the Prototype to the Test
With the kiosk flow prototyped (scan, unlock, redeem) we ran a validation study with 100 participants (ages 18–55, 62% regular coupon/deal app users) across the full flow: start-scanning → first-list → bogo-unlock → sign-in → welcome-back → my-deal → deal-details.
Quantitative Findings (n=100)
Overall usability score: 84/100 Clarity score: 4.2/5 Emotional trust score: 4.0/5 Perceived speed: 4.4/5 Likelihood to use: 7.6/10
What Worked
96% understood the primary action on the start-scanning screen 100% recognized the sign-in requirement 98% understood redemption instructions on the deal-details screen 78% said the flow felt "simple"; 71% said scanning felt "fast"
What Didn't
41% felt sign-in was "poorly placed in the middle of the flow" The top friction point 27% misinterpreted the "BOGO Unlock" screen as already unlocked 22% expected a progress indicator during the unlock step 28% felt the deal list lacked visual hierarchy
Design Recommendations
Move sign-in to the beginning or end of the flow Add progress indicators to unlock steps Strengthen visual hierarchy between locked and unlocked deals Add clearer expiration labels on deal details Provide an optional "save for later" function
Reflect
Reflection & Next Steps
Reflection
This project reinforced that usability issues aren't always about comprehension, they're often about placement and timing. Nearly every screen scored high on understanding (96–100%), but the moments of friction came from wheresomething happened in the flow, not whether users understood it. Sign-in placement is the clearest example: 100% recognized it was required, yet 41% still called it disruptive.
Running a primary survey and a validation study back to back also sharpened how I think about research sequencing, the survey told us what was breaking (multi-item deal clarity, checkout confidence), and the validation study told us why the redesigned flow still had gaps. Neither alone would have given the full picture.
Looking Ahead
If this work continued, the next phase would focus on resolving the sign-in placement issue and re-testing with a revised flow. Future directions could include:
Re-testing a revised flow with sign-in moved to the start or end
Adding progress indicators to the unlock step and measuring their effect on the 22% who expected one
Strengthening visual hierarchy for locked vs. unlocked deals and re-measuring the 28% who cited hierarchy as an issue
Exploring alternative interaction models (e.g. a persistent task bar) as a follow-up direction beyond the kiosk-style flow
