I'm not familiar with the module in question, but currently Perl will interpret your strings as latin-1. If your source file is saved as UTF-8, Perl will see "Mäster Mäster"
If you say use utf8;, your strings will contain the correct character sequence (...assuming you saved the source file in UTF-8. Most *nix editors do that by default today.)
Furthermore, the Excel::Writer::XLSX perldoc contains a "UNICODE IN EXCEL" section which has a few examples on getting your strings to Perl's internal encoding correctly.