# Missing data in regression

I am researching the effect of different marketing mix variables (e.g., price promotion, innovation) on the market share. More specifically, I want to analyze the effect of different marketing mix variables on market share, by taking the position of the manufacturer brand into account (because a leading manufacturer brand may cause a different effect). Hence, I have variables like price_promotion_brand_1, price_promotion_brand_2, price_promotion_brand_3, etc.

Furthermore, I have panel data over multiple supermarkets in different countries. Some of the supermarkets are hard discounters and therefore have mostly private/store labels and one or two manufacturer brands. Accordingly, I have a lot of "missing data" because there exists no manufacturer brand at the hard discounters. Hence, it does not make sense to impute data because there is, in reality, no data.

Does anyone have a solution to cope with the "missing data" (other than listwise deletion)?

P.s. I use multilevel/hierarchical models

If the proportion of missing on $$X$$ is not low (for example, >10%), you can create two variables: $$X_1 = X$$ if $$X$$ is not missing, = 0 if $$X$$ is missing. $$X_2 = 0$$ if $$X$$ is not missing, = 1 if $$X$$ is missing.
Then fit the model with $$X_1$$ and $$X_2$$ as covariates (other covariates have no change and still in the model). The meaning of coefficient of $$X_1$$ is the same, the change of response variable when $$X$$ increase by 1 unit. The meaning of coefficient of $$X_2$$ is average change of response variable comparing $$X$$ is missed and $$X=0$$.
If the proportion of missing on $$X$$ is low, maybe delete them from the analysis is good approach.