I do not know if Google (for example) uses the average time spent on pages, etc. for the classification, but if a website receives many visits, especially if people spend some time, read several pages, the likelihood of those people being linked to it. The site (sharing links, etc.) is higher (since that behavior usually means good content), and that definitely helps to look for classifications. It all comes down to: good quality content (and "user experience") = better ranking.
In terms of SEO, in addition to good quality content (text, images, user interface design, easy navigation, load times of decent pages, etc.), the most you can do is follow the google guides to collect and correctly mark the content that the search engines must know. what it is more easily (and more accurately).
An example (exemption?):
My website is a niche of information that provides one. People search on Google what they want to know, get a page from my site as a superior result, go to it, read it and go out. If a visitor does not click on any other page of the site, Google has no way to measure the time spent on a page (there is a way to avoid this for everything I know, but I did not bother). That visit is recorded as a bounce. Therefore, my site has a relatively short average time for visitors and a relatively high bounce rate, but it is still very high in searches (at the top for some terms). While the behavior of the "measured" user could hardly be worse. The content, on the other hand, is of good quality, relevant and properly labeled for what it is (the image of a car, for example, is labeled "auto", if it explains that term, if it explains the car's tires, it is labeled "tires in a car", etc. – similar for file names). In addition, people refer to the links when someone requests information: in forums, Facebook, etc.