I have a dataset on companies with distinct id, products they sell, the region they sell their products to, and in which year.

For example:
id product region year
2 87150000 1 2011
9 65069990 2 2011
9 65059010 2 2011
24 94055000 3 2009
24 94015000 3 2010
27 84514000 4 2009
27 95038000 2 2009
27 84211910 5 2010
33 63029900 2 2010
44 39235000 1 2011
44 42021210 6 2010
44 42023200 1 2009
44 95034100 1 2009
44 63053200 1 2009
44 39249000 1 2011
44 83017000 1 2010
44 61099090 8 2010
..........


I want to create a dummy variable that equals to 1 when a company introduces a new product to a region in a year i.e.: when an -id- sells a new -product- to a -region- in year -t- but did not sell that -product- to that-region- in any of the previous years. This is my dependent variable for my regression later on.

I have tried:

Code:
// Create first occurrence:
egen id_prod_region = group(id product region)
egen int introduce =min(year), by(id_prod_region)

// Dummies:
gen byte new = (year==intro)
I'm having trouble with this. Is this the correct way to go around it? Could anyone help? Thanks.