Total Budget Optimizing Result has incorrect initial total response

facebookexperimental / Robyn

Robyn is an experimental, AI/ML-powered and open sourced Marketing Mix Modeling (MMM) package from Meta Marketing Science. Our mission is to democratise modeling knowledge, inspire the industry through innovation, reduce human bias in the modeling process & build a strong open source marketing science community.

https://facebookexperimental.github.io/Robyn/

MIT License

1.07k stars 322 forks source link

Total Budget Optimizing Result has incorrect initial total response #919

Open rohan1997vk opened 4 months ago

rohan1997vk commented 4 months ago

Project Robyn

Describe issue

The total response for the last 4 weeks is ~750K. However, after a 2000ierations with 5 trails and selecting a model, the budget allocator shows that the total response is ~350K. The actual spend is 1.06M and the actual revenue is ~750K. R2 is around 0.90

Provide reproducible example

Environment & Robyn version

Make sure you're using the latest Robyn version before you post an issue.

packageVersion("Robyn") [1] ‘3.10.3’ sessionInfo() R version 4.3.0 (2023-04-21 ucrt) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows 11 x64 (build 22631)

gufengzhou commented 4 months ago

The total response refers to media only response, or revenue contribution from media, NOT the total revenue. The budget allocator only maximises the media contribution and it can't move anything for trend/season/others that have no spend. Makes sense?

apoorvabc commented 4 months ago

Just curious, Does this mean the bounded response output also contains the spend and revenue of media channels alone and doesnt include the trend/season/competitors and other factors? Do we have to add these values up incase we are looking for the total revenue earned during the time period?

gufengzhou commented 4 months ago

Yes and that'd imply forecast of trend/season etc that is quite a topic of itself. But yes you're right.