To be specific, the data shows Google search interest - not number of protest events.

Protests are not necessarily “newsworthy”; however, that search interest would track media stirring up news before elections seems likely.

Further, (and I say this as a Minneapolitan) the significant protests of this summer were very clearly and obviously precipitated by real world events, not astroturfed protest crowds.

One question is why and how certain specific real-world events go viral and others get memory-holed.

I would surmise that a lot of the protests this summer were, emotionally, a consequence of the stress, economic uncertainty, and way-too-online cabin fever of the early stages of the COVID epidemic.

You wouldnt believe how sophisticated the firms are that the dem party hires. They have their own versions of Cambridge Analytica which are more than capable of amplifying messages.

I'd believe it. It would certainly explain a lot.