Comment by jellicle | Hacker News Reader

Comment by jellicle | Hacker News Reader

It's a cute trick, and undoubtedly will attract interest. But the interest is as likely to be negative as positive - people think it's creepy. Has Google focus-group tested these changes? Has Google asked focus groups of random people (not Google employees) what they actually want in search? Or is the decision making process basically something like, "This is cool!" (from Google programmers) and "This will help us make money!" (from Google executives)?

Because I think it's the latter.

A search engine's value is in bringing new information to people. Matt Cutts knew that he was playing Werewolf with Brian Fitzpatrick. That photo of Matt Cutts returned to Matt Cutts' search was duplicate information. It's a dupe! Cutts already knew that! It might be new to anyone else in the world, but not Matt Cutts!

Google is busy attaching a tube from our anus back to our mouth to facilitate information consumption. (You can call it an echo chamber or living in a bubble if you like.) This will not end well.

Using focus group testing in the way you're describing is like the French revolutionary who says "There go my people. I must find out where they are going so I can lead them."

Google is notoriously one of the most (if not the most) data-driven companies in the history of the world. That doesn't mean they never make bad decisions, but I'm sure they've taken user behavior into account.

The duplication you mention isn't a bug. Just because I've seen something before, doesn't mean I never want to see it again, especially if it's relevant. Searches are often re-searches for information you've seen before. The most important thing is relevance, and your social interactions are relevant.

I think what many here are missing is that the exposure of social in search results is only the first-order effect. But the social signals themselves probably affect your personal relevancy algorithm, too. I imagine that's a valuable addition, and it will only get better with more social data.

Yes, they took user behavior into account and realized that it didn't matter because they have to get into the social space. Remember, social media success is tied into all Google employee bonuses. They are willing to make search worse in order to make G+ better.

Yes, relevance is everything. When I search for something in a search engine 95 times out of 100 I am searching for something new. Occasionally I am searching for something that I already found when searching before. I am never looking for photos of myself. I mean please.

This is corporate spin and you are eating it up! :)

Yes, they took user behavior into account and realized that it didn't matter because they have to get into the social space.

I think you have the causality wrong. They realized they had to get into the social space and integrate it into their product because of user data.

They are willing to make search worse in order to make G+ better.

I think the point of this announcement is that they make no such distinction. Search and Plus are the same product.

When I search for something in a search engine 95 times out of 100 I am searching for something new.

Thank you for your data point. I'm sure Google has many others. Not only that, but how much of that is because you haven't even thought of using search for information retrieval purposes yet? How often do you access your photos? How often do you access your friends' photos? What's your primary way to access them? What percentage of that access would be made better using search?

I am never looking for photos of myself. I mean please.

I bet you're the exception.

"""I think you have the causality wrong. They realized they had to get into the social space and integrate it into their product because of user data."""

They did it because of "user data" alright, but not in the way he means it. User's haven't told them: "wow, it would be great if you had Plus integrated into search".

Their data was more like: "hey, this social thing is lucrative, numbers show we can milk our users more if we get a stronghold in this area".

"""I am never looking for photos of myself. I mean please.

I bet you're the exception."""

And I bet you that he is not.

At least non in Google searches --people weren't looking for photos of themselves there.

> I am never looking for photos of myself. I mean please

What if your friend 'Fred' e-mailed you and asked "Can you send me that picture of us from the rafting trip last year?"

Which would you rather do - dig through your photo albums until you find it, or type in a search for it and have it show up in the results?

"""What if your friend 'Fred' e-mailed you and asked "Can you send me that picture of us from the rafting trip last year?" Which would you rather do - dig through your photo albums until you find it, or type in a search for it and have it show up in the results?"""

You do realize that this edge case (and tons of similar) are not representative of 99% of the queries we do, right?

(Percentage pulled out of my ass, but still more accurate than the edge case mentioned).

And even if it did, how about doing a special: "Search Social", box instead of polluting the main search results?

Well sure - the exact example I used is an edge case :-)

But I think the general case of searching for something that I either own or that was shared with me is not really that uncommon. I think it's only uncommon on Google today, because - until now at least - Google could only search the public web for you.

The way I think of it is this: When the web started out, the amount of stuff was small and it was possible to keep track of the site you needed to know about. Then the web grew and people tried to manually organize it (ala early Yahoo). Then the web grew more and search became the best way to find something.

The same kind of thing happened with personal computers. In the early days there was limited storage and a single folder worked fine. Then storage got cheaper and hierarchical directories were added and that worked for a while. Now there tends to be so much stuff on your computer that it is often easier and faster to just search for it.

I think the same thing is happening on the web. As the amount of stuff we have on the web grows - especially stuff that is private or shared with a select group of people - the need to effectively search it will grow.

> A search engine's value is in bringing new information to people.

I'd rearrange that sentence: its value is in bringing information that is new to the user.

Cutts knew that he played Werewolf, but did he know that the picture existed, or where it was?

It is creepy, though. Reminds me of this story: https://www.privacyinternational.org/blog/how-facebooks-targ...

I think that is a very narrow definition of the value of a search engine.

I would say that a search engine's value is in helping me find what I'm looking for.

Often - as you point out - what I'm looking for is new information.

But there are plenty of other times, when I know exactly what I'm looking for, but I either forgot where it is, or it is simply faster to search for it than to go digging through folders, apps, etc to get to it.

In the case of Matt's story, maybe he remembered that someone posted a picture of him playing werewolf, but he couldn't remember who posted it or where they posted it to. In this case searching for it was very useful and gave him exactly the information he was searching for.

Or maybe I know that I have a picture from some event last year with a friend, but I can't remember - or I'm too lazy :-) - to dig through my folders to find it. Being able to search for it is awesome.

It's analogous to the way search works on my phone, or on Windows 7 or mac spotlight.

If I want to find a contact on my phone, I can dig through the contacts app, or I can just type the name into the search box. If I want to find a program or a document on my computer, it's often faster to just search for it rather than navigate through menus and/or folders.

This new search feature is simply bringing that same kind of functionality to the data you have in the cloud.

It's true, people may think it's creepy. But there are other values provided by a search engine. A great majority of less IT-savy people types yahoo.com (or yahoo) into google. Or more IT-savy people use search engines to search for pages they have seen before.

We need to see beyond how a search engine is predominantly used by us more-technical folks and look at how real people use it.

Google has an extensive user research team, and make extensive use of quantitative and qualitative feedback. They have scores of everyday users physically come in to the office or remotely participate in the studies every week.

Well they must have missed a week.

Search for cars while you are opted-out of personal search. You will see the Google+ Ferrari page on the right. Is that really what people want to see? The normal Ferrari website isn't anywhere in the top 100 organic results, but yet Google's extensive user research team decided that is what people wanted to see on the right hand side. Or did they put that there because it benefits them.

Well, they obviously didn't skip a week, and making snarky comments like that misses the point. Whatever testing they did, and the way they interpreted it, meant they were happy to go ahead with the implementation you see.

Reply on news.ycombinator.com