Re:How is search so bad? A case study
by l0b0 on Sunday January 19, 2020 @03:50PM
(#22093970)
Is it time for paid search engines? Make users vote with their wallets and pay for the eternal arms race. Problem is, whoever is behind something like that would have to start with an already sufficiently superior experience (or massive geek cred) to make people pay from the early days. Maybe going back to manually curated results of a subset of topics would work? Or some Stack Overflow-esque model of user powered metadata generation?
[ Parent
| on HN ]
No. Just because you pay for something doesn't mean you aren't also the product. The other ideas are specifically ones which failed. It didn't even work that well in the 90s. User metadata isn't even a solution in itself but an optimization layer at best.
Simpler idea:
Paid search engine that ranks sites based on how often the users click results from that site (and didn't bounce, of course). The fact that it's paid prevents sybil attacks (or, at least, turns a sybil attack into a roundabout way of buying ads).
Of course, at this point, you are now the product even though you paid. But it's a tactic that worked for WoW for ages.
Google already includes clicks and bounces in its ranking factors.
Re:How is search so bad? A case study
by seshagiric on Sunday January 19, 2020 @05:09PM
(#22094458)
Probably not. The problem with current search engines is that they need some way to rank the pages and hence inherently susceptible to clever SEO (that curates pages to get ranked high for certain type of queries).
So having a paid search engine does not fully solve the SEO problem. Having no Ads does not takeout the SEO problem of boosting pages to top.
If the paid for search engine ignores SEO "optimizations" and actually ranks on content AND these results prove better that attracts actual paying users, then we don't have to wait for SEO to die. Just like SEO rose to prominence to win Google result ranking because that's where the users were, the sites would stop the SEO crap.
Getting this hypothetical site to get users is the real problem. Same thing with getting users to a Facebook alternative.