Case study · Consumer goods
AB InBev MROI Modeller Portal
Marketing mix modelling for short cycle analytics teams.
- Client
- Anheuser-Busch InBev
- Service
- UI/UX design, full-stack development
- Sector
- Consumer goods
- Users
- Short cycle marketing analytics
At a glance
What the portal standardised
Analysts had been spending a significant share of every cycle on data collection and preparation by hand.
One workflow the team runs, rather than reassembling the analysis and its interface conventions each time.
Historical and live campaign data in the same view, so a campaign can be responded to while it is running.
One of three Wow Labz built for AB InBev's marketing organisation, alongside the Global Resource Allocator.
The brief
Analysts doing data entry instead of analysis
Short cycle analytics is a rhythm, not a project. AB InBev's team derives learnings from past marketing spend on a repeating schedule, and the value of that work depends on it being fast enough to inform the plan it feeds.
Most of the cycle was going somewhere else. Analysts spent a significant amount of time manually handling data collection and preparation, workflows were inconsistent between runs, the interfaces were complex, and the feedback loop was slow enough that responding to a campaign while it was live was not really on the table.
Automate the parts that never change, so the cycle spends its time on the part that does.
The build
Where this sits next to the Allocator
AB InBev's marketing organisation runs on two related questions. Where should the money go, and what did the last round of spending actually teach us. The Global Resource Allocator answers the first, across more than 100 brands in more than 50 countries. The MROI Modeller Portal answers the second, for the analytics team whose job is turning spend history into learnings.
They are deliberately different products. The Allocator is a decision surface for a wide group of stakeholders. The Modeller Portal is a working tool for specialists who run the same analysis on a cycle, which is why automated pipelines and a simplified workflow matter more in it than breadth of access.
Python handles the backend data processing, ReactJS the dynamic interface. Wow Labz did the UI/UX design and full-stack development.
Modules
What the portal added
Integrated data pipelines
Automated ingestion and processing, which is the piece that gives analysts their cycle back.
Streamlined insight generation
Multi-channel marketing impact tracked in one tool rather than assembled per channel.
Custom dashboard views
Tailored metric displays, so the people running this every cycle see their numbers rather than everyone's.
Real-time feedback loop
Historical and live campaign data together, enabling response during a campaign instead of a post-mortem after it.
A usable interface
Simplifying the analytics workflow was an explicit goal, and adoption went up because of it.
Stack
What it was built with
Questions
The MROI Modeller Portal, answered
What is the MROI Modeller Portal?
An internal tool Wow Labz built for Anheuser-Busch InBev that lets its short cycle marketing analytics team derive learnings from past marketing spend, with automated data pipelines replacing manual collection and preparation.
What problem did it solve?
Analysts were spending a significant amount of time manually handling data collection and preparation, with inconsistent workflows, complex interfaces and a feedback loop too slow to respond to a live campaign.
How does it relate to the Global Resource Allocator?
They answer two halves of the same question. The Global Resource Allocator decides where marketing budget should go across more than 100 brands in over 50 countries. The MROI Modeller Portal works out what past spend actually taught the business.
What is it built with?
Python for backend data processing and ReactJS for the dynamic interface. Wow Labz provided UI/UX design and full-stack development.
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.
