Abstract The article focuses on a comparison of English-Estonian code-copying in blogs and in vlogs. This paper applies a usage-based approach, combining a cognitive angle with the code-copying framework. The aim is to provide a holistic view on contact-induced language change by applying bottom-up analysis of naturalistic multilingual language use. In contact linguistic literature it has been observed that there is a preference for insertions vs. alternations depending on sociolinguistic setting (community type, generation, proficiency) and structural properties of the languages involved. Research on English-Estonian language contacts has shown that in blogs there is no clear preference for either. From a different point of view, it has been observed that lexical impact (global copying in the terms of code-copying framework) precedes semantic and structural impact (selective copying) but provide no explanation. Comparisons between blogs (750 entries and 275,263 words from 45 bloggers) and vlogs (5,5 hours, approximately 36,854 words from 5 vloggers) shows that global copies and alternations heavily prevail over other types of copying, yet the number of selective copies is somewhat higher in vlogs and of mixed copies in blogs. Selective copies are loan translations rather than structural changes. We assume that (1) prevalence of global copies and alternations depends on genre norms (blogs and vlogs are constructed as non-monolingual, highly individualized genres); (2) as selective copies are mostly loan translations, it implies the role of meaning and cognitive aspects: idioms and fixed expressions are figurative and cognitively prominent; combinational properties and grammatical meanings are abstract; so, the more abstract the meaning is, the more cognitive effort and time is required for entrenchment and conventionalization; (3) copying and alternation is denser in vlogs because the genre is oral and spontaneous vs. written, edited genre of blog.