Big data search instruments, Google and Yandex and others, offer new research opportunities, statistically very impressive. The questions arise here as to what extent we can trust the statistical data obtained, what kinds of hypotheses can be verified through Google and Yandex searching. The authors try to experimentally verify the hypothesis relating to the word order in the attributive group of the kind healthy nourishing food vs nourishing healthy food using Google and Yandex and in this way test the explanatory potential of the big data sources. The choice of the word order is related to the theory of classes which explains cognitive mechanisms that govern the order of adjectives in a noun group.  The empirical results via Google and Yandex help to verify this hypothesis and identify cognitive grounds for choosing the order of attributes. The authors claim that in addition data search engines, particularly, Russian Yandex and Google corpora, can help identify linguistically and culturally relevant concepts.