4 ms·
IF Google were to bias in favor of "FLOWER" while disregarding content relevance, it would be an abuse of their dominant market position. But from the outside l
by oinksoft 14y ago
IF Google were to bias in favor of "FLOWER" while disregarding content relevance, it would be an abuse of their dominant market position. But from the outside looking in, it seems like Google strives for nothing but providing the most relevant search results for their users. Controversies tend to arise when Google tries too hard, it seems, localizing based on fragile heuristics and such.
If this is the scam, what's the alternative? An open search engine, perhaps, where development, ranking algorithms, etc. are completely out in the open. I'm not sure this is realistic any time soon though given the cost of hosting a search engine.
- cbeach 14y agoA search engine that ranks mainly on social-graph recommendation might be one solution, because the social graph is a simple data structure that transcends "ownership" by Facebook or Google+. The engine itself is a non-trivial system to implement, mind you.
- Silhouette 14y agoThe trouble with this idea is that Google can't magically see private links/recommendations on social networks any more than anyone else can. For example, one site I help with has just a few inbound links from low PR sites as far as Google is concerned, and gets a page rank of zero accordingly. In reality, it probably gets more inbound links via social networking in a single day on most days, and unsurprisingly those personal recommendations often drive more inbound traffic than any search engine. Google's basic premise of ranking the importance of a site based on how many links it gets and where from is wildly inaccurate in the era of social networking. Normal people mostly don't run their own public web sites or blogs any more, they share stuff on sites like Facebook and often only with their "friends". Personally, I think this is a good thing, because both as a visitor and as someone running sites I would rather rely on legitimate personal recommendations from people who are genuinely interested in a site than on some arbitrary algorithm running in a data centre somewhere on another continent. I don't think it's healthy for any one intermediary to have a dominant effect on whether interested people can find interesting content on the Web, and sites like search engines and social networks are merely intermediaries.
- bones6 14y agoIn your example, I see it like this: Google sees those links and gives low page rank but quantity helps too. So if you're getting a bunch of links on social media, it follows that eventually people will blog about it (crawlable content) and link back to the site. Sure it may be a crappy blogger blog but it's legit and Google does count that. Over time a ton of non-spam PR0 backlinks can and will help rank for search terms. Of course, getting some big PageRank links will help a lot more and if you have any say in the matter, can help with what keyword you rank for.
- Silhouette 14y agoI guess my point is that, talking generally now rather than about any one specific site, we just don't need to put as much weight as perhaps we used to on what Google think of our pages this week. More and more of our traffic comes from personal mentions, passed on by people who are actually interested in what we have to their friends with similar interests, and people who find us that way are far more likely to be genuinely interested themselves and to enjoy our material, increase our income, or otherwise match whatever we created the site for. People use buzzwords like "going viral", but really this is just the same old exponential growth of word-of-mouth recommendations for good products, speeded up thanks to modern communication technologies and the social networks they support. Obviously we still make our sites accessible to search engines. It's not like it's difficult to do that if you've got real content and a sensible information architecture, and after all, a bit of extra traffic never hurt anyone. Maybe, if an enthusiastic person finds us that way, it could even seed a new network of people finding us. But I see this becoming less and less of a priority and dedicated search engines becoming less and less relevant for a lot of sites as people learn to use social networks to spread the word. In particular, I don't see cbeach's original idea of a "search engine that ranks mainly on social-graph recommendation" adding much value in that scenario, simply because the social networks will probably offer similar functionality without needing a third party anyway.
- pseut 14y ago> Google's basic premise of ranking the importance of a site based on how many links it gets and where from is wildly inaccurate in the era of social networking. Normal people mostly don't run their own public web sites or blogs any more, they share stuff on sites like Facebook and often only with their "friends". I'm not sure that I understand this point. I thought that Google wants to index Twitter (for example) but is prohibited by Twitter. So it's hard to see that their premise is wrong: personal recommendations on social networks are just inbound links from specific sites, just not sites that Google can index.
- duskwuff 14y agoYou're assuming that social-graph recommendations are somehow not gameable. If anything, they're even easier to game than Google search results on current social networks -- there are tons of sites out there that'll offer to "sell" you Likes on Facebook for a page of your choosing.