FMCG Retail · Israel
When COVID restrictions collapsed physical shopping, the real problem wasn't building a webshop — it was moving an offline-habituated audience online at scale, without a sudden behavioural break. I architected the transformation: solution logic, project structure, the team, and a staged migration model aligned with customer psychology.
During COVID, Israel's largest FMCG retail chain faced a severe structural shock. Movement restrictions sharply reduced physical shopping, while a large share of customers — especially older, habit-driven buyers — weren't comfortable buying groceries through a standard e-commerce interface. This wasn't a typical digital-growth initiative; it was an emergency transformation: demand still existed, but the habitual buying environment had collapsed.
Across the broader restriction period, sales fell roughly 60–65%, with sharper drops at the extremes. The problem ran deeper than lost traffic: customers still needed essentials, many weren't ready to shop purely digitally, and standard catalogue navigation didn't replicate how FMCG purchases are actually made. The task: move a traditionally offline audience online — quickly, at scale, without forcing a sudden behavioural break.
I architected the transformation process — defining the solution logic, designing the project structure, assembling cross-functional resources, coordinating execution after approval, and reporting to management. My contribution sat at the intersection of anti-crisis sales transformation, product strategy, CX design and digital commerce architecture. The core working group was 10+ people across psychology, UX/UI, analytics, CRM & marketing, and external ML support — launched fast, observed live, and refined iteratively in production.
Rather than forcing customers into a standard catalogue, we introduced a transitional interface that mimicked the familiar logic of physical shopping — virtual shelf navigation inspired by real planogram principles. Not a 3D store, but enough of the visual and behavioural logic of in-store selection to feel intuitive. The real barrier wasn't digital readiness; it was behavioural mismatch. The interface created a "laminar transition": familiar visual structure, lower cognitive friction, easier basket-building, and reduced resistance among older, less digitally-adapted customers. It rolled out through multi-channel activation (social, email and printed guidance, order inserts, in-store messaging), and customers could choose their interface while the system preserved their preferences.
Once users crossed the behavioural barrier, the model shifted from transitional UX to personalized commerce. The chain's key advantage: a high-coverage loyalty ecosystem — 80%+ of customers had purchase history linked to loyalty cards. That enabled behavioural prediction from purchase intervals, seasonality, price sensitivity, household habits and category relationships.
Predictive basket. Returning users found a substantial part of their expected order already pre-filled. In many repeat sessions, customers kept roughly 82–94% of pre-filled items — turning shopping from a manual search into an approval flow. Explainable recommendations then surfaced adjacent products with visible reasoning, onboarding customers into automation gradually rather than as a black box. Migration to the final personalized catalogue was gradual (~3 months of dual interfaces) and highly successful: ~92% of users moved across.
The result came from matching the reality of the moment. The crisis was immediate; customer behaviour couldn't be changed by abstract digital optimization; the business needed a bridge, not a forced jump — and personalization became far more effective once behavioural migration was solved first. The sequence mattered: recreate enough familiar shopping logic to reduce resistance → migrate users into digital behaviour without cultural shock → then use loyalty data and predictive logic to make online buying materially easier than offline.
Beyond immediate recovery, the program established online commerce as a structurally stronger model inside the retail system. As digital demand stabilized, the company gained flexibility to optimize its network toward broader-assortment, more centralized formats — and the ~60% post-restriction online share showed durable behavioural change, not temporary emergency usage.