ajkohn

My problem is that the filter bubble is increasingly used as a retort of fear, uncertainty and doubt when discussing personalization, marketing and privacy. It’s become a proxy to end discussions about how our personal data can, will and should be used as technology advances. Because despite the dire warnings about the dangers of the filter bubble, I believe there’s potentially more to gain than to lose.

It's interesting. Sometimes the argument is that people are more savvy than we give them credit for and then it's that they don't really understand.

I'm not saying you're doing this now (not at all), but I do see people trying to have it both ways when talking about this issue.

The thing is, I'm old so I worked offline before the Internet. The personal information that is accessible offline is ... incredible, even to this day. It's what has powered the direct marketing revolution for decades.

So while I don't think people understand the details of how things work online I think they have an intuitive sense that the same way they get targeted offline is what happens to them online.

Facebook is actually letting people marry the offline and online information which I find super intriguing yet goes relatively unnoticed.

And I'm still shocked that we willingly present ID when we buy something with a credit card since we're not really required to do so.

Going slow wouldn't be a bad thing but I actually think we've been doing that already.

Well, using both the links above (though I botched the first link to ad preferences) you can opt-out of the the remarketing/display cookie and the search cookie.

For those that get to the Ad Preferences page the stats are pretty interesting.

http://www.mattcutts.com/blog/clarifying-valleywag-comments/

The general public when presented with this doesn't seem to have such a drastic reaction.

Point taken. But you don't get to speak to those who think Google is evil without getting them to read it first.

For me the ends justified the means. Of course you may disagree and simply find it to be an annoying type of pandering.

FWIW - the evil debate we're having here has once again sort of derailed us all from the main thesis which is ... Google wins when people use the Internet more.

So while some laugh at the self driving car initiative and see it as indulgent I see a company who is trying to unlock countless hours of time each day that is currently 'non-usable' from an Internet perspective.

Oddly, I agree. That's one of the reasons I wrote this piece. Though clearly I've used the evil word to tease people into reading it.

I agree. If there really is this desire, then Google needs to honor it.

The Ads Preference manager is fairly straight-forward (www.google.com/ads/preferences) and more people who go there change settings rather than opt-out.

The full opt-out instructions (http://support.google.com/accounts/bin/answer.py?hl=en&a...) are a bit tougher. Should they make it easier to do this? Maybe.

But if you make it too easy, you get false positives and then people may become upset that Google isn't as useful as it once was not fully understanding that it was their own preference to do so, right?

When you ask someone whether they want their search results or ads to reflect their browse and search history they say no but when presented with the two options, users nearly always select the personalized version.

Remarketing works. It's not what people say, it's what they do.

And why shouldn't it be that way. I go to a local Peet's. They know my order now. Is that bad? Should I tell the barista to wipe my order history out of her memory?

I'm not saying there isn't a issue of privacy but the line is not where most think it is IMO. People are very willing to give out personal information in exchange for a chance to win a car in the local mall.

Taken in another direction: product. Users often don't express their real needs or can't identify them. One of the more interesting cases was the introduction of the mini-van. It tested miserably when they asked users. But Ford (I think it was Ford) decided to go ahead anyway and ... users wound up LOVING it.

I covered Google+ in the piece. It's definitely a bit of an outlier. I see Google+ as two things:

1) A way for Google to get primary and active signals about user preferences to better personalize search results. (aka - make search better).

2) A way to break the monopoly Facebook had on social attention. The strategy only works if users are in places where Google has a presence. That's almost everywhere - except Facebook. So ensuring that attention doesn't pool and persist in a dead zone was important - and I think they've succeeded.

Yes, the 'Don't be Evil' motto sets Google up for being a target for every little thing that someone thinks is wrong, whether or not it really is or not.

What I think is meant by that motto is that they want to be a positive force, to use their technology and smarts for good. But of course, there's an ends justify the means argument and not everyone will have the same definition of good.

The censorship stuff is ... sticky. The data collection stuff is largely fine though a full opt-out is difficult. The problem there is that Google knows that personalized experiences are better for users whether they know it or not.

While I didn't expand upon it in the piece I believe many apply the 'evil' brush to a host of normal, though sometimes ruthless, business activities.

People seem consumed with those 'wrongs' and wind up missing the larger picture entirely.

Yes, Facebook is still the leader here. It's taken a good 6 years for Twitter to decide it wanted to be a real business and there was a lot of drama at the executive level.

But in the last year Twitter has a clear direction and they're moving on it fast. So I see 2012 as the build year and 2013 as the conquer year.

The funny thing about that Facebook Australia announcement is that they identify mobile as an achilles heel they have to solve. Facebook let mobile pass them by and now they're playing catch up.

I too am eager to see Google experiment with AuthorRank. Not only should it be much harder to scam, break and abuse, I believe Google is thinking about those things in the creation phase, something I'm not sure occurred with PageRank.

I'd be interested to hear if others agree with my assessment or not.

Thanks for putting this up on HN. Let me know if anyone has suggestions on how to improve the guide.

The definition of 'low-quality' sites can be rather subjective, can't it? I personally get no value out of Mahalo or Squidoo but I know others actually do.

While I might not love eHow's process and tactics, I'll admit that I've benefited from one or two of their easy-to-read articles. (Though I'll often second source that information.)

I think Google's in an interesting pickle to decide whether results should be: fast food or fine dining.

Google could serve up 'fast food' in results (e.g. - farmed content) and it would likely be 'good enough' for MOST people. Plenty of folks will eat fast food even if there is a better alternative down the road.

I'd like to see more fine dining results, but I'm not sure I'm in the majority there. Quality for the early-adopter might look different than quality for the late-majority.

If the algorithm is leveraging user behavior as a signal, doesn't it follow that popular and familiar sites and brands may gain more trust and authority?

Is Google thinking at all about search in this way?

Here's some preliminary data that shows that Google Instant may have changed the distribution of clicks by rank, with more emphasis being put on results above the fold.

http://bit.ly/dAzvX3

It's not about the query behavior, but about assessment behavior. As such, your mileage may vary based on your particular rank for terms.

Outside of this analysis, the keyword length could remain the same, but a site may receive less traffic if the auto complete algorithm has changed for critical terms. Again, this will vary site by site - and where one loses, another will gain.

Well, the impact on search query length may be negligible but there are other ways in which Instant may be changing search behavior. I have a small slice of data I hope to post in the very near future.