Median Absolute Deviation (MAD) without additional add-ins and with blanks in data

Hi guys, I was just wondering whether you could help me with the following situation.

Basically, I want to be able to compute MAD on a data set of around 2000 companies with scores for various factors, e.g. #of employees, revenue figures etc The problem is that for any given factor, only 70-80% of companies with have that data, the rest are blank. Is there a way to calculate MAD for each factor using only the figures given, ie without substituting blanks for zeroes (as that would affect MAD) and without downloading additional add-ins (as that would require whoever is using the excel file to have that add-in loaded, sharing would become tricky).

Many thanks in advance!