Automated headline generation for online news articles is not a trivial task\n- machine generated titles need to be grammatically correct, informative,\ncapture attention and generate search traffic without being "click baits" or\n"fake news". In this paper we showcase how a pre-trained language model can be\nleveraged to create an abstractive news headline generator for German language.\nWe incorporate state of the art fine-tuning techniques for abstractive text\nsummarization, i.e. we use different optimizers for the encoder and decoder\nwhere the former is pre-trained and the latter is trained from scratch. We\nmodify the headline generation to incorporate frequently sought keywords\nrelevant for search engine optimization. We conduct experiments on a German\nnews data set and achieve a ROUGE-L-gram F-score of 40.02. Furthermore, we\naddress the limitations of ROUGE for measuring the quality of text\nsummarization by introducing a sentence similarity metric and human evaluation.\n