A reference application to demonstrate simple Media Mix Modeling (MMM).
JavaScript CSS HTML Python Shell
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
Docs
POCs
SampleDataSet
analysisEngines
conversionHub
database
mediaHub
reportInterface
LICENSE
README.md

README.md

Media Mix Modeling (mediamixmodeling)

Media Mix Modeling(MMM) is an econometric technique to measure effectiveness of media in the marketing initiatives. It is the most scientific way, that marketers use, to measure Return on their Marketing Investment(ROMI). Please refer to Wikipedia to know more about MMM - https://en.wikipedia.org/wiki/Marketing_mix_modeling

Brand spends billions of $$ every year worldwide on advertising through different media but only few brand measure their media spending. The recent survey by https://cmosurvey.org suggests that less than 20% brand use econometric modeling to measure marketing ROI and rest simply believe that their media spending works but never measure. This survey also reveals that only 2.31% of total marketing budget is being spent on measuring marketing effectiveness. The primary reason why marketers do not measure their media spending are: a. Complex : Econometric Model requires complex statistical analysis. It require lots of volumetric data from different media sources that is challenging to collect, streamline and process. b. Expensive : Most of the marketers believe that measurement techniques are expensive that requires specialized knowledge and proprietary tools.

These reasons are valid but the industry need to find solution that can enable marketers to count every penny they spend on media with transparency. The industry can use technology to overcome challenges that prevent marketers to measure media spending. Today, technology is quiet capable to manage data collection from different sources without much hustles. The complexity of statistical analysis can be easily resolved by using various open source statistical analysis APIs and tools.

The cost of media measurement must be affordable to brand. The industry need more professionals with specialized knowledge and technical ability to innovate new generation of tools that can not only enable marketers to manage their media spending, in affordable ways, but also empower them to connect with emerging earned media.

The new entrants, specially technologists, in the field of media measurement need a simple tool to understand and explore how basic Media Mix Modeling works. This tool can be used as sandbox to innovate new possibilities in the emerging media that can help brand to shift to human centric marketing.

I propose to build a tool for basic MMM through this repository with following features:

  1. Web based responsive application built on top of latest technology stack a. Frontend - Reactive framework - TBD b. NoSQL DB - Mongo or any other similar DB c. Server Side - Node.js based any framework d. Statistical Analysis library - R or Python
  2. Ability to upload Volumetric data in CSV format for Media Mix Modeling
  3. Ability to plot Area and line chart to show relationship between media and sales data
  4. Ability to edit uploaded volumetric data.
  5. NoOps and micro services principles will be followed in system design and Architecture.

Insyallation Instructions - https://github.com/rkthakur/mediamixmodeling/wiki/Installation