Case study · Consumer goods
AB InBev Global Resource Allocator
Maximising marketing ROI with real time visibility into global spend, predictive modelling and collaborative decision making built in.
- Client
- Anheuser-Busch InBev
- Service
- Design and development
- Sector
- Consumer goods
- Scale
- 100+ brands, 50+ countries
AB InBev, impact
What the portal changed
Data driven marketing budget allocation across the global portfolio.
Manual processes and decision lag cut by more than half.
Actionable intelligence through predictive modelling and scenario simulation.
Global collaboration and spend transparency in a single place.
The brief
A global portfolio decided in spreadsheets
AB InBev allocates marketing budget across more than a hundred brands in more than fifty countries. Decisions of that shape have to reconcile local market knowledge with global portfolio strategy, and they were being made through manual processes, which meant the answer often arrived after the moment that needed it.
The problem was not a shortage of data. It was that the data, the modelling and the people who had to agree on the outcome were in different places, so every allocation round paid a coordination tax before anyone got to the actual judgement call.
One portal, so the numbers, the forecast and the conversation about them finally sit in the same place.
The build
Visibility first, then foresight
The portal puts global marketing spend in one view, at the level a decision actually gets made: by market, by brand, against the constraints a planner is working inside. That alone removes most of the assembly work that used to precede an allocation conversation.
On top of that sits the modelling. Predictive analytics turn spend history into a forecast, and scenario simulation lets a team ask what a different allocation would do before committing to it. The result is a portal people decide in, not a report they read afterwards, which is where the 60% reduction in decision latency comes from.
Under it: Python with Flask for the services, Celery and RabbitMQ for the modelling jobs that cannot run inside a request, and ReactJS with FusionCharts for the interactive front end. A companion project, the MROI Modeller Portal, gives the short cycle analytics team the marketing mix modelling side of the same picture.
Modules
Five things the portal does
Smart budget allocation
Historical data and predictive analytics drive an allocation aimed at maximising net revenue, rather than repeating last cycle's split.
Scenario planning and forecasting
Teams simulate multiple budget strategies and see the projected outcome before any money is committed.
Customisable inputs and objectives
Priorities and market specific factors are set by the teams who know them, so a global model still respects local conditions.
Executive decision support
A centralised dashboard with interactive charts and performance comparison, built for the people signing off rather than the people modelling.
Market-wise collaboration
Global executives, regional leaders and finance teams work from one set of numbers instead of a spreadsheet per region.
Inside the portal
Allocation, modelling and reporting
AB InBev Global Resource Allocator screens, covering allocation summaries, market level breakdowns and scenario comparison views
Stack
What it was built with
Questions
The Global Resource Allocator, answered
What is the AB InBev Global Resource Allocator?
A strategic portal Wow Labz designed and built for Anheuser-Busch InBev that gives the business real time visibility into global marketing spend, with predictive modelling and collaborative decision making built in. It covers more than 100 brands across more than 50 countries.
What changed after it shipped?
Marketing budget allocation became data driven across the global portfolio, and decision latency fell by more than 60% as manual processes and decision lag came out of the cycle. Global collaboration and spend transparency moved into a single centralised portal.
How does the predictive side work?
The portal turns marketing spend history into actionable intelligence through predictive modelling, and lets teams run scenario simulations to compare allocation options before committing budget. Modelling jobs run asynchronously through Celery and RabbitMQ so the interface stays responsive.
What is it built with?
Python and Flask on the backend, Celery with RabbitMQ for asynchronous modelling work, and ReactJS with FusionCharts for the interactive dashboards.
Let's talk
Got a developer product nobody finishes integrating?
Tell us what you are building. We will tell you how we would ship it.
