The article en:Hard disk drive gives en:Seagate Technology as a suggestion, but not any of its competitors. This is undesirable.
Topic on Talk:Reading/Web/Projects/Related pages/Flow
@Ruud Koot if it's based on calculated similarity rather than editor bias, is it still an NPOV violation? It is not favoritism...if these were authors, I don't think we would feel the same way.
Yes, because everything about the presentation of these results tries to make the picks look like an objective selection instead of a number of best-effort search results (most notably them being displayed without a request from the reader, close to other hand-picked navigational content). This would be very different if these results were just the top results on several pages of search results, displayed after a user-initiated query. Context, presentation, and user expectation matters.
When the algorithms selects, say, classic authors this may not always be harmful, but when it starts picking commercial companies or political parties it certainly is. See also Search engine manipulation effect.
Okay, thanks for the additional context.