Hello @Anne-lim-te,

Thank you for sharing a detailed question and including relevant research you have done for this!

To answer your question on setting channel-wise budget constraints, yes this can be done in Meridian. The documentation for that is in fact available on the same page you have shared. You can see that the optimize method accepts parameters spend_constraint_lower and spend_constraint_upper. Each of these can either be a single value (which sets the same limit on all the channels) or a list of size n_paid_channels. Check the following quoted from the documentation:

spend_constraint_lower
Numeric list of size n_paid_channels or float (same constraint for all channels) indicating the lower bound of media-level spend. If given as a channel-indexed array, the order must match (InputData.media + InputData.reach). The lower bound of media-level spend is (1 - spend_constraint_lower) * budget * allocation). The value must be between 0-1. Defaults to 0.3 for fixed budget and 1 for flexible. Consider using InputData.get_paid_channels_argument_builder() to construct this argument.

spend_constraint_upper
Numeric list of size n_paid_channels or float (same constraint for all channels) indicating the upper bound of media-level spend. If given as a channel-indexed array, the order must match (InputData.media + InputData.reach). The upper bound of media-level spend is (1 + spend_constraint_upper) * budget * allocation). Defaults to 0.3 for fixed budget and 1 for flexible. Consider using InputData.get_paid_channels_argument_builder() to construct this argument.

Regarding the comment on GitHub Discussion 378. That is,

Apply stronger optimization constraints on these channels to prevent them from receiving too much budget allocation under optimization.

This particular suggestion was in response to the user’s initial query regarding how to prevent the budget optimizer from rounding up the budget while allocating it. They were also facing issues with incorrect response curves and over estimated incremental outcomes. The context of this suggestion is different, however, it still suggests customizing optimization constraints which is what you are also trying to do. And the implementation of the same can be done as suggested in the first paragraph.

Do reach out if you have any further questions or suggestions regarding Meridian!

Google Meridian Support Team