df1                              df2
Geogkey    Brand  Week               date       Impressions
TMZ43434   x      6/16/18           6/14/18       798798
KRO36783   y      6/16/18           6/21/18       562314
....                                6/28/18       462534
n

Geogkey    date    Impressions
TMZ43434  6/14/18   798798
TMZ43434  6/21/18   562314
TMZ43434  6/28/18   462534
KRO36783  6/14/18   798798
KRO36783  6/21/18   562314
KRO36783  6/28/18   462534

empty <- data.frame(df2$date, df2$impressions)

#creates a new data frame with unique geogkeys
geogname <- unique(data.frame(df1$GEOGKEY))

#create some function that will index each unique geogkey and make a new
column for df2 with that name (e.g. df2$geogkey <- some function)

new_df <- rbind(empty, df2)
#this should theoretically append all the geogkeys to the dates and
impressions

df <- data.frame()
for(i in unique(df1$GEOKEY){
  for(j in 1:nrow(df2)){

       df <- rbind(df,
                   data.frame('key' = i, 'date' = df2[j,1], 'impressions' =  df2[j,2]))

   }
}