Horizontal aggregation was more efficient in a world of physical delivery. Vertical aggregation makes more sense in a world of digital delivery. And enabling the customer or user to have some control over the curation is possible in the digital world but hardly is in the physical.Shatzkin sees the future information ecosystem trending towards niches or 'verticals' with global audiences.
He is contrasting this model with traditional bookstores and trade publishers that cover a wide range of subjects. It also seems the opposite of the way that a traditional academic or public library is setup with books spanning a wide range of subjects and positioned to serve a local audience.
old=local and horizontal
new=global and vertical
I would argue that in the academic repository arena, we can already observe the difference between these two approaches.
Institutional repositories aggregate scholarship that crosses a wide range of subject areas only tied together by affiliation with a single academic institution. They might be described as local and horizontal.
Disciplinary repositories like the Social Science Research Network and arxiv.org concentrate content in certain academic disciplines. They might be described and global and vertical.
Which model is more successful, the disciplinary repositories or the institutional ones? If this ranking is right, it is the disciplinary repositories. They have the most momentum and interest behind them.
Generally, I think that digital initiatives in libraries will be most successful if they are able to build on a vertical community. Projects that are too wide in scope end up being about nothing.