The following analysis investigates selected properties of the language of the tweets used in the interaction with consumers on English and Polish brand profiles. The analysis examines the structure of tweets, word frequency, as well as the frequency of informal and non-standard language items, language mistakes, the use of emoticons and hashtags. The study contrasts the language used by English and Polish representatives and reveals a number of similarities and differences between the corpora. What the corpora share is a high frequency of conventional politeness acts and language structures reflecting a customer-oriented tone of the interaction. Differences are observed in the frequency and use of informal and non-standard structures, emoticons and hashtags, as well as in the structure and complexity of the tweets. The study indicates a lower formulaicity and a greater individualization of the interaction on the Polish profiles.