FMCGCase study

Global Food & Lifestyle Platform: How I ran it

Led the end-to-end reinvention of a community webapp for a global beverage brand, from user research to a tag-driven personalisation system, scaling across 8 countries and 4 continents.

2023 – present · 8 markets · 4 continents

How I ran it

Cadence, roles, risks and the decisions behind the delivery.

Governance

  • Designs reviewed in sprints in Figma, with feedback from about 30 stakeholders
  • Close, ongoing collaboration with the SEO and editorial agencies
  • A defined rollout process with stage gates and timeframes for each market
  • Jira conventions and a standing meeting cadence to keep the stakeholders aligned
  • Monthly executive and top-line reports per market, plus a quarterly business report, each ending in insights, challenges and next steps
  • Priorities set around “must-win battles” per market

Who did what (RACI)

RACI matrix: for each activity, who does the work (R), approves (A), is consulted (C) or informed (I)
ActivityClientUX/UIDevDataSEOEditorialMarkets
User research and discoveryClient: AUX/UI: RDev: not involvedData: not involvedSEO: not involvedEditorial: not involvedMarkets: C
Design in Figma (sprint reviews)Client: AUX/UI: RDev: not involvedData: not involvedSEO: not involvedEditorial: not involvedMarkets: C
Webapp and PWA developmentClient: AUX/UI: CDev: RData: not involvedSEO: not involvedEditorial: not involvedMarkets: not involved
Tagging taxonomy and personalisationClient: AUX/UI: not involvedDev: RData: RSEO: not involvedEditorial: CMarkets: not involved
Datalayer, dashboards and KPI reportingClient: AUX/UI: not involvedDev: not involvedData: RSEO: not involvedEditorial: not involvedMarkets: not involved
SEO audit, metadata and hreflangClient: AUX/UI: not involvedDev: CData: not involvedSEO: REditorial: not involvedMarkets: not involved
Editorial content and chef profilesClient: AUX/UI: not involvedDev: not involvedData: not involvedSEO: not involvedEditorial: RMarkets: not involved
Campaign landing pagesClient: AUX/UI: RDev: RData: not involvedSEO: not involvedEditorial: not involvedMarkets: C
Quarterly business reportClient: AUX/UI: not involvedDev: not involvedData: RSEO: not involvedEditorial: not involvedMarkets: not involved

R does the workA approvesA/R approves and doesC consultedI informed

Main risks and how I handled them

Aligning about 30 stakeholders on design decisions

Scope

Designs reviewed in sprints through Figma, with structured feedback loops.

Dependency on external agencies’ timelines (SEO, editorial)

Vendor

Close, ongoing collaboration with both agencies.

Losing search visibility when migrating off the legacy platform

Technical

Page audits, metadata and hreflang optimisation.

Reliance on third-party data and APIs for the map

Integration

Integration with an external restaurant database and the Google Places API.

Decisions

Drive personalisation from the tagging system

Context
The brand needed predictable, editorially controlled personalisation.
Decision
A deep tagging system (awards, cuisine type, location) powers dynamic personalisation.
Outcome
Personalisation that editors can steer.

Offer lite and full registration

Context
Registration growth was a core goal.
Decision
A dual registration system with a lite path.
Outcome
Registrations grew +32% year over year.

Deliver as a progressive web app

Context
The community used the product across many devices and markets.
Decision
Build the new platform as a PWA.
Outcome
A single product across devices.

Go social-first and use push notifications

Context
Engagement and retention beyond search traffic.
Decision
Early social-first approach and a push notification strategy.
Outcome
Engagement channels beyond organic search.

What I learned

  • Registration was the weakest KPI against its annual target; we formed three hypotheses (value proposition, registration nudges, usability) and responded with gated content, conversion-flow analysis and extra datalayer events.
  • Personalisation works when editors control the taxonomy behind it.
  • Stage gates made each market launch repeatable.
  • Measuring from day one turned the registration gap into an action plan instead of a surprise.

Working on something like this?

Download my CV or message me on LinkedIn.