2 ms·

The problem, IMO, might be the monoculture we have around search. Because Google is soo big, it's enough for spammers to target it and they have the vast majori

The problem, IMO, might be the monoculture we have around search. Because Google is soo big, it's enough for spammers to target it and they have the vast majority of the search visibility. If we had better, more diverse competition, that might manifest as a tradeoff, presumably, they would have competing and diverse criteria so you would probably not be the top result on _all_ dominant search engines. SEO spam needs upkeep and attention to latest algos, else it decays. Competing algos would yeld better results for everyone. Maybe Google is just ripe for a shakeup.

jefftk 5y ago
Doesn't your model predict that Bing would have substantially less SEO-gamed results? (Disclosure: I work at Google, but not on search)
- RealityVoid 5y ago
  Well... Yes, it should. But, no, it seems it does not. I thought about this when typing it but did it anyway, maybe because I thought there is still something worthwhile there. I still think the model could work if the algorithm is sufficiently different than Google's. Ideally, people would go "I did not find anything I cared about on Google, I know, I'll use Bing!" - but nobody does this, because the results are consistently worse. Don't get me wrong, I like G as a company, I think they do worthwhile things! But they have left things slip and need competition into this field, I mean real competition, then maybe they would actually address issues. Maybe the issue is also on the incentive level as well. I mean more searches means more eyeballs and more money for Google. If someone searches one thing and they are done that is less interaction! I hope they don't work like this, but it's possible. And another possible problem is the opposite. Maybe Google is optimizing search for what it thinks people want, but it uses the wrong metric. Or it gives people what they want but not what they need.

jefftk 5y ago

Doesn't your model predict that Bing would have substantially less SEO-gamed results? (Disclosure: I work at Google, but not on search)

RealityVoid 5y ago
Well... Yes, it should. But, no, it seems it does not. I thought about this when typing it but did it anyway, maybe because I thought there is still something worthwhile there. I still think the model could work if the algorithm is sufficiently different than Google's. Ideally, people would go "I did not find anything I cared about on Google, I know, I'll use Bing!" - but nobody does this, because the results are consistently worse. Don't get me wrong, I like G as a company, I think they do worthwhile things! But they have left things slip and need competition into this field, I mean real competition, then maybe they would actually address issues. Maybe the issue is also on the incentive level as well. I mean more searches means more eyeballs and more money for Google. If someone searches one thing and they are done that is less interaction! I hope they don't work like this, but it's possible. And another possible problem is the opposite. Maybe Google is optimizing search for what it thinks people want, but it uses the wrong metric. Or it gives people what they want but not what they need.