> SEO-wise (and in no other way), I think we should have more sympathy for Google. They are just… losing at the cat-and-mouse game.
I don't think they are; they have realised (quite accurately, IMO) that users would still use them even if they boosted their customers' rankings in the results.
They could, right now, switch to a model that penalises pages for each ad. They don't. They could, right now, penalise highly monetised "content" like courses and crap. They don't do that either.[1]
If Kagi can get better results with a fraction of the resources, there is no argument to be made that Google is playing a losing game.
--------------------------------------
[1] All the SEO stuff is damn easy to pick out; any page that is heavily monetised (by ads, or similar commercial offering) is very very easy to bin. A simple "don't show courses unless search query contains the word courses" type of rule is nowhere near computationally expensive. Recording the number of ads on a page when crawling is equally cheap.
> If Kagi can get better results with a fraction of the resources, there is no argument to be made that Google is playing a losing game.
Google's algorithm is the target for every SEO firm in the world. No one is targeting Kagi. Therefore, Kagi can use techniques that would not work at Google.
I don't think they are; they have realised (quite accurately, IMO) that users would still use them even if they boosted their customers' rankings in the results.
They could, right now, switch to a model that penalises pages for each ad. They don't. They could, right now, penalise highly monetised "content" like courses and crap. They don't do that either.[1]
If Kagi can get better results with a fraction of the resources, there is no argument to be made that Google is playing a losing game.
--------------------------------------
[1] All the SEO stuff is damn easy to pick out; any page that is heavily monetised (by ads, or similar commercial offering) is very very easy to bin. A simple "don't show courses unless search query contains the word courses" type of rule is nowhere near computationally expensive. Recording the number of ads on a page when crawling is equally cheap.