1/ It's fascinating to see how query patterns have changed as a result of COVID-19 and the resulting social distancing and shelter-in-place orders. #search #seo #google
2/ Query volume in some verticals has simply evaporated. In other verticals, terms that were once thinly searched are now suddenly rising to head term status.
3/ With a few weeks of data it also seems like time of day and day of week curves have changed. More so the latter with the difference between weekday and weekend not nearly as pronounced.
4/ Today I began to think about how this punctuated equilibrium in query patterns might be handled by Google. The intent behind many queries has changed as has the type of content demanded by users.
5/ Do learning models based on long clicks adapt in the short term or not? If they do, does the model stay in place when query patterns revert to 'normal'? Does how you perform now lead to continuing success in the future?
6/ Those that 'lean in' during this strange time could have a considerable advantage. If your brand is suddenly used more often with important terms and modifiers, that could have a long-term impact in search.
7/ From an 'academic' standpoint I'm interested to know how Google approaches such a change in behavior. This isn't just a QDF identification problem or understanding emerging behavior around say, the Olympics or an election.
8/ How does Google think about adapting search quality based on a radical change in search behavior due to so many people being less mobile and working from home? cc: @haahr