As long as Twitter (and Facebook even more so) don't allow Google to actually index them fully, I don't think that a service like the one outlined here would work very well.
And even once these services open up more: At what point would you set the threshold for inclusion of $social_network in the list? I certainly would not want a list of various social network batches under every search result.
Only list the ones the user is active in? Sure. But how do you determine "active"? Networks in which the user posts a lot? In which he reads a lot of other users content (how would Google find out about that?)?
The integration as it is currently made for G+ basically only works because Google has all that G+ profile data and for it to work as well with other networks, I'd say, Google needs specialized access to it.
Or to be more cynical: So twitter wants to have their accounts integrated into the search results of a search engine? Easy: Just create a search engine too and have all of Google's user switch over.
And let's not forget that Twitter and FB both have search features (actually I'm not sure about FB, I've never been on it, but it has to, right?), and furthermore, they're both in their ways trying to obviate search by providing what people want in other ways. That's the premise of FB as a Google killer right? That social is undermining search as the way people get information on the internet?
Setting aside how absurd I find that notion to be, it seems to me that the whole G+ project is a response to that threat, and it's a response that says "I can duplicate your functionality a lot easier than you can duplicate mine, and it adds a lot more value to my major service to duplicate your functionality, than it adds to your major service to duplicate mine."
It also seems like a play to get access to Twitter and FB data. Google uses free data on the open web. If you don't allow them to index your data for free, why would you expect them to include them in their search results?
As a user, I want Twitter and FB to give Google all of their data, and I want Google to run all of that data through their algorithms and give me the results that are the best, regardless of what service it's on. From where I sit, it doesn't seem like Google is the player that's in the way of that happening.
> If you don't allow them to index your data for free,
> why would you expect them to include them in their search
> results?
This tool demonstrates that Google is already accessing and indexing the data it says it needs to improve its search results. This isn't rhetoric; it's code.
> This tool demonstrates that Google is already accessing and indexing the data it says it needs to improve its search results. This isn't rhetoric; it's code.
It doesn't really demonstrate that. It demonstrates that Google can show G+ pages related to a topic, and that it can show high-ranked social pages linked to the owners of those G+ pages. But it doesn't show that on a broad, scalable basis Google can dependably link from a topic to a highly relevant 3rd-party social page. Demonstrating it on a couple of easy, celebrity-focused queries isn't proof of anything.
I've noticed Facebook even asks you to login if you're logged out of Facebook and you're reading on a Facebook page. Why should Google have to deal with that when even Facebook themselves don't want their own data to be very public?
Blake why are you guys completely kicking out the legitimate G+ profile links from the search results using the bookmarklet? I thought you were against search result manipulations. I search for people like Guy Kawasaki or Trey Ratcliff who are very active in G+ and yet their profiles are not listed on the first page of the results if I use your bookmarklet.
If Focus on the User really doesn't want to be "evil" and you want to "focus on the user" then the plugin needs to be changed. The edits the plugin makes are better, but it just looks like Facebook/Twitter are just making sure they get their websites mentioned. It appears they are not focused on the user, but focused on how it will help them.
Here are some major suggestions:
1. If someone searches for cars in Google, 95% of the time they are looking to either research cars, buy a car, or the movie Cars. First of all do you honestly feel that Ferrari's facebook or google+ page is more relevant/useful than the websites of CarsDirect, Edmunds, or MotorTrend? The Google+/Facebook pages are more prominent than these websites. That makes zero sense. I don't know if you can do this, but if possible the "People and Pages" section should be removed. If not, see #2.
2. Why would the first result be Ferrari's facebook page in this search over Ferrari's own web site?
3. The Autosuggestion thing should work the same way as above. When I search for Matt Cutts I am not looking for either his Facebook or his Google Plus page, the most relevant result would be mattcutts.com, no? Or just remove the autosuggestion thing period. Assuming that I want to go see Matt Cutts Facebook profile over his Google+ profile is silly.
4. Make it so that you can actually remove Google+ results period. When I search for NFC Championship Game I don't care to see that MG Siegler went to that game. Nor would I want to see that on facebook either. It just ideally would be removed.
As I see it, the code that adds social networking links to people in the right side-bar (which is probably the most prominent feature of the bookmarklet) depends on the G+ profile being there.
So Google has their own algorithms to, say, connect Yehuda Katz with a search query for "ruby". This means that the plugin only finds ruby people who also have a G+ profile and then adds links to other services too.
Now for this to be really useful, it should find people even if they don't have a G+ profile and that is what I wonder whether it's even possible by just crawling the web to the extent allowed by the various services.
Adding other Social Networking links to already determined Google+ profiles is not interesting. Finding people on Twitter related to a search term regardless of whether they have a G+ profile or not is the interesting part.
AFAIK Google only has access to the public information that it can gather by scraping those websites (which isn't an easy feat at all, given the frequency of updates) or using the public APIs, that only give you a limited amount of information.
Until last Summer, Twitter had a contract with Google to give the latter its firehose stream, with all its tweets in real time. AFAIK, it was Twitter that decided not to renew that contract, evidently because they thought they could monetize that information better without Google. But now, apparently, they want their cake and eat it too...
> Google has access to the same information about other social profiles that it does about its own profiles, as the tool on the site demonstrates.
Does it demonstrate that? It's more likely that those profiles just show up because of the normal page rank algorithm (eg Jamie Oliver's twitter account ranks highly because so many people link to it). It is a reasonable proxy for popularity for popular people, at least, though there's probably too much noise before you get too far down the long tail. But it isn't the same thing that the OP was talking about.
I actually think it would be cool if google took the more Open Social approach by taking the sites I list in my profile as mine, looked at my friends and frequency of posting where available (more twitter than facebook), and then biased my results to favor my preferred networks. I don't know if it would end up with great results, but it could be an interesting separate tool, perhaps.
normally i feel like i'm a lot more relaxed about privacy than most people on this site, but holy crap. the top voted comment is currently suggesting that facebook open up their entire site to google indexers?
google can do personal search on g+ because they maintain the permissions database for g+ content and won't show private stuff to the wrong people. how do they do that for facebook?
Comments
As long as Twitter (and Facebook even more so) don't allow Google to actually index them fully, I don't think that a service like the one outlined here would work very well.
And even once these services open up more: At what point would you set the threshold for inclusion of $social_network in the list? I certainly would not want a list of various social network batches under every search result.
Only list the ones the user is active in? Sure. But how do you determine "active"? Networks in which the user posts a lot? In which he reads a lot of other users content (how would Google find out about that?)?
The integration as it is currently made for G+ basically only works because Google has all that G+ profile data and for it to work as well with other networks, I'd say, Google needs specialized access to it.
Or to be more cynical: So twitter wants to have their accounts integrated into the search results of a search engine? Easy: Just create a search engine too and have all of Google's user switch over.
And let's not forget that Twitter and FB both have search features (actually I'm not sure about FB, I've never been on it, but it has to, right?), and furthermore, they're both in their ways trying to obviate search by providing what people want in other ways. That's the premise of FB as a Google killer right? That social is undermining search as the way people get information on the internet?
Setting aside how absurd I find that notion to be, it seems to me that the whole G+ project is a response to that threat, and it's a response that says "I can duplicate your functionality a lot easier than you can duplicate mine, and it adds a lot more value to my major service to duplicate your functionality, than it adds to your major service to duplicate mine."
It also seems like a play to get access to Twitter and FB data. Google uses free data on the open web. If you don't allow them to index your data for free, why would you expect them to include them in their search results?
As a user, I want Twitter and FB to give Google all of their data, and I want Google to run all of that data through their algorithms and give me the results that are the best, regardless of what service it's on. From where I sit, it doesn't seem like Google is the player that's in the way of that happening.
> If you don't allow them to index your data for free,
> why would you expect them to include them in their search
> results?
This tool demonstrates that Google is already accessing and indexing the data it says it needs to improve its search results. This isn't rhetoric; it's code.
> This tool demonstrates that Google is already accessing and indexing the data it says it needs to improve its search results. This isn't rhetoric; it's code.
It doesn't really demonstrate that. It demonstrates that Google can show G+ pages related to a topic, and that it can show high-ranked social pages linked to the owners of those G+ pages. But it doesn't show that on a broad, scalable basis Google can dependably link from a topic to a highly relevant 3rd-party social page. Demonstrating it on a couple of easy, celebrity-focused queries isn't proof of anything.
The tool isn't limited to celebrity queries. Install it and use it on whatever queries you wish.
I realize the tool isn't limited, but it only works with the (relatively) short list of terms that Google is using to display that Google+ pane.
I've noticed Facebook even asks you to login if you're logged out of Facebook and you're reading on a Facebook page. Why should Google have to deal with that when even Facebook themselves don't want their own data to be very public?
Yes, it is a technical solution to a "problem" (air quotes denoting that this was a conscious business decision by Google).
It brought to my mind the infamous checklists explaining why technical solutions to spam would fail: http://craphound.com/spamsolutions.txt
Blake why are you guys completely kicking out the legitimate G+ profile links from the search results using the bookmarklet? I thought you were against search result manipulations. I search for people like Guy Kawasaki or Trey Ratcliff who are very active in G+ and yet their profiles are not listed on the first page of the results if I use your bookmarklet.
We are not touching the organic results. The code for the bookmarklet is freely available; anyone can verify that.
You don't have to guess how well such a service would work. It has been built and is available on the site.
Google has access to the same information about other social profiles that it does about its own profiles, as the tool on the site demonstrates.
Blake:
If Focus on the User really doesn't want to be "evil" and you want to "focus on the user" then the plugin needs to be changed. The edits the plugin makes are better, but it just looks like Facebook/Twitter are just making sure they get their websites mentioned. It appears they are not focused on the user, but focused on how it will help them.
Here are some major suggestions:
1. If someone searches for cars in Google, 95% of the time they are looking to either research cars, buy a car, or the movie Cars. First of all do you honestly feel that Ferrari's facebook or google+ page is more relevant/useful than the websites of CarsDirect, Edmunds, or MotorTrend? The Google+/Facebook pages are more prominent than these websites. That makes zero sense. I don't know if you can do this, but if possible the "People and Pages" section should be removed. If not, see #2.
2. Why would the first result be Ferrari's facebook page in this search over Ferrari's own web site?
3. The Autosuggestion thing should work the same way as above. When I search for Matt Cutts I am not looking for either his Facebook or his Google Plus page, the most relevant result would be mattcutts.com, no? Or just remove the autosuggestion thing period. Assuming that I want to go see Matt Cutts Facebook profile over his Google+ profile is silly.
4. Make it so that you can actually remove Google+ results period. When I search for NFC Championship Game I don't care to see that MG Siegler went to that game. Nor would I want to see that on facebook either. It just ideally would be removed.
As I see it, the code that adds social networking links to people in the right side-bar (which is probably the most prominent feature of the bookmarklet) depends on the G+ profile being there.
So Google has their own algorithms to, say, connect Yehuda Katz with a search query for "ruby". This means that the plugin only finds ruby people who also have a G+ profile and then adds links to other services too.
Now for this to be really useful, it should find people even if they don't have a G+ profile and that is what I wonder whether it's even possible by just crawling the web to the extent allowed by the various services.
Adding other Social Networking links to already determined Google+ profiles is not interesting. Finding people on Twitter related to a search term regardless of whether they have a G+ profile or not is the interesting part.
AFAIK Google only has access to the public information that it can gather by scraping those websites (which isn't an easy feat at all, given the frequency of updates) or using the public APIs, that only give you a limited amount of information.
Until last Summer, Twitter had a contract with Google to give the latter its firehose stream, with all its tweets in real time. AFAIK, it was Twitter that decided not to renew that contract, evidently because they thought they could monetize that information better without Google. But now, apparently, they want their cake and eat it too...
> Google has access to the same information about other social profiles that it does about its own profiles, as the tool on the site demonstrates.
Does it demonstrate that? It's more likely that those profiles just show up because of the normal page rank algorithm (eg Jamie Oliver's twitter account ranks highly because so many people link to it). It is a reasonable proxy for popularity for popular people, at least, though there's probably too much noise before you get too far down the long tail. But it isn't the same thing that the OP was talking about.
I actually think it would be cool if google took the more Open Social approach by taking the sites I list in my profile as mine, looked at my friends and frequency of posting where available (more twitter than facebook), and then biased my results to favor my preferred networks. I don't know if it would end up with great results, but it could be an interesting separate tool, perhaps.
normally i feel like i'm a lot more relaxed about privacy than most people on this site, but holy crap. the top voted comment is currently suggesting that facebook open up their entire site to google indexers?
google can do personal search on g+ because they maintain the permissions database for g+ content and won't show private stuff to the wrong people. how do they do that for facebook?