Amazon's A9 Drops Retained Data Methods

eldavojohn writes "The recent update to Amazon's A9 service has removed its ability to record searches. A9 (which now uses Windows Live & Alexa) used to tout the ability to save every single search the user made, which required a login. Now, they no longer require you to log in and have dropped the recording of searches from their toolbar. What they added was aesthetic changes to the search site. What they dropped was the A9 Instant Reward, the A9 Toolbar, the A9 Yellow Pages, the A9 Maps (including Block View), the user diary, bookmarks, and history. Although they claim that A9 is merely 'shifting its priorities to areas where it can provide the greatest benefit for customers,' this smacks of a move to avoid the ethical controversies and pressures that come with retaining your user data. What does the rest of Slashdot think about retaining search data? Is it a liability or an asset?"

Depends...

I suppose that depends on what data is being stored, and why...in the case of Amazon, I personally could care less if someone else could look at which CD's I'm interested in. I think the decision should be left to the end user.

Re:And If the Government Asks?

The government> I accuse Mr. White in the Conservatory with the CD!

A search is not a purchase. For purchases, there are already records. It doesn't matter if you're searching for that new-fangled satanic rock and roll. If it matters at all, it will be if you purchase (or download) the CD.

Instant Reward dropped?

So are they saying that you no longer get the A9 discount at Amazon if you use the search page? That's the only reason I was using it! Without that discount, I, and I'm sure others, will no longer use their search.

Re:It just goes to show..

Hey, when you have a family to support, every penny counts! And besides, I hardly ditched Google, I just sent a search through there once or twice a week just to keep the discount active, then I usually ended up putting the same search into Google anyway :P

Re:It just goes to show..

When A9 first debuted, it used google results, so it didn't really matter. Now it uses msn's search, it isn't nearly as good, but once or twice a week instead of typing in a url, I'll just search A9 for the title of the website.
1.5% may not be much, but it takes so little effort that it is worth it.
I just checked and the pi/2 discount is still in effect, so I'm not sure if that actually is going away or if they just aren't promoting it any more.

Asset, under certain circumstances.

Let's assume I'm not searching for too outlandish things. Let's pretend I do not look for 70's pr0n featuring dogs, or if I do, that my fiance is aware of it and condones it. Being fairly normal and open means I have nothing to hide. They can retain all the data on me they like. It's no skin off my back, if you will.

The only but in that statement is that I don't want them to retain search- or private data under two circumstances:

1) If the search data affects the answers given in future searches
2) If it results in spam or theft of any kind

Now, providing it's securely stored, doesn't affect answers to future queries and helps them enhance their services, I really don't mind.

They'll still keep it, just not make it public.

Amazon would be downright stupid not to save the search data; they're just not visibly saving it anymore (thus, less questions), that's all.

By this same argument, one might say "oh, look, google isn't saving search data either; if I can't see it, it's not there!". Yeah, riiight.

Cost cutting

Sounds to me like they can't keep up with the big boys on this one. Not too many companies can compete in the all-facets-of-the-internet game like Google, Yahoo, and Microsoft. Amazon is just sticking to what it does best.

On a side note, I haven't been keeping up with Amazon in the news for the past few years... has it turned a profit yet?

What if....

The big issue on recording searches is that a third party is storing (and possibly selling/revealing) information on you. So what if the system were changed to store your searches locally, run the analysis locally, and report back only links between data, not the data, or personally identifiable information.

For example, you search for David Bowie, then you search for 'The Cure'. The analysis process determines that you have linked David Bowie and The Cure and reports that link back to the search engine. Even if the search engine's database is leaked, all the public would know is that some number of users have found a link between David Bowie and The Cure. The big question mark in such a system is how the analysis process could determine that link.

-Rick

It should be the consumer's choice

Web sites that do collect "non-essential" data should give the customer the opt in or opt out, depending on the nature of the site. Sites like banks that must create audit trails for certain transactions need to have full disclosure.

Slashdot gives me the option of posting and submitting stories under my user-id or anonymously. Everyone knows that I, davidwr, posted this message at the time indicated above. Any paying subscriber can look up "davidwr" any time in the future and see all the messages I posted. If I wanted to be private, I could be "Anonymous Coward."

I avoid newspaper sites tha require a login because I want to eliminate the possibility they will tie what I read today with what I read tomorrow.

More data is always better than less, however...

..you then need the right policies, procedures, and security to deal with it. Data Cleansing or Data Scrubbing is not a trivial process and unless you have the right models and expertise, you start to get buried. My guess is that this decision was one half poor planning and the other half was fear from all the bad press. Adding true relevancy to searching will take some type of historical data, but you need to how to use it correctly and protect it wisely...

A shame to lose A9 Maps

What they dropped was the A9 Instant Reward, the A9 Toolbar, the A9 Yellow Pages, the A9 Maps (including Block View), the user diary, bookmarks, and history.

It's a shame to lose A9 maps, the block view was a very useful feature. Google Map's integration of satellite photos is impressive but lets face it, most of us view the world from the ground not the from the sky, so the block view is helpful for seeing what your destination looks like. Unfortunately, they only got around to photographing the downtown core of a handful of major US cities.

Certainly an Asset

The problem with the whole setup is that it's seen as an asset for the business or service provider, A9 in this case. which is wrong. it shouldn't be. It is an asset though, for the users using that service. You and me could get great benefit from having our past search data, clikthroughs and surfing habits analysed and used to improve our web experience, however I don't trust any company (even ones that "do no evil") to do that without skewing things in their favour and/or violating my privacy, whether now or in response to some government request.

What I don't get is, with these toolbars, desktop searchers and whatnot, why does this even need to be stored at the service provider? Why not simply store it locally in an open format and then start publishing tools for users to make use of this lovely data goodness? Surely a browser extension that uses your usage data locally, from across the web, will completely outperform something based on only one site, and avoid the privacy and commercial skewing issues.

Seems to me like A9 is dodging a bullet, but missing the boat.

There's saving and then there's retaining

Removing the ability for users to save their own searches is not exactly the same as retaining user searches.

I betcha the latter is still very much happening. If I were Amazon, I'd retain the data just to have it available for analysis. That sort of stuff has tremendous business value.

Safer Not To Save Searches

Most of the time the stuff I look for isn't anything special as near as I can tell. But I'd ultimately prefer that my search data is not saved. What's innocuous today, could get one landed in jail, embarassed, or worse years later. When I browse a real bookstore, no one is following me around noting every book I pick up or cover I glance at. I'd like the same consideration given to my online searches. I have nothing to hide (right now), but I still don't want to be spied on, unless it's by attractive women with amorous intent! ;)

Both liability and asset

It's a liability if others have access to my searches, it's an asset if I do.

The problem with me having access to my searches, is that it's possible for others to get access as well (cracking and court orders come to mind). Encrypted on my own hard drive is good, but still subject to court order (since it doesn't matter how "personal" your hard drive is, it isn't subject to the 5th ammendment - which only applies in the USA anyway). Encrypted in escrow is probably better, but really tough to get right.

Maybe I should only search for things I don't care about.

Slippery slope

It always starts with, "why would I care about my CD searches... " it's down hill from there. The most unfortunate part is there is a huge level of end user responsibility here. End users must understand the choices and be able to make their own choices. Maybe this is a part of Web2, browsers where on can make the choices on a global level. Now matter what, it comes down to who cares and who cares more Uncle Sam, "The Amazons of the net" or John Q. Public. I can guess who is last in this list.

Both

Retaining search data is both an asset and a liability to a company. The question isn't which is it, but whether it's value as an asset outweighs it's cost as a liability or not. If it doesn't, then it doesn't matter how much of an asset it is you get rid of it.

One thing that has gotten lost in the noise...

is that Block View is getting dropped. This was one great feature that A9 had that really is totally unrelated to any sort of ethical controversies. It was a great supplement to Google Maps if you're going to an unfamiliar place. Now its dead. RIP.

Slashdot blinders

What does the rest of Slashdot think about retaining search data? Is it a liability or an asset?

This is a classic case of Slashdot blinders, thinking the entire world is black or white. Here, let me help you:

Retaining data is obviously a liability. It's invasive of my privacy. It gives companies data on me that they can sell to other companies without my permission, and those other companies send me annoying pre-screened offers of credit. My info will eventually end up in a big database in the basement of the FBI, where they'll try to link me to terrorism by playing a big connect-the-dots game. 'He searched for the Anarchist Cookbook and Catcher in the Rye. Arrest his ass.'

On the other hand, retaining data is obviously an asset. It allows companies to cater their services to my personal needs. It lets them suggest products that I may not have even known existed. My information builds company worth, which stimulates the economy and provides employees with jobs. It facilitates retailer/customer trust and loyalty, which benefits both parties.

I'm sure I could go on and on with more arguments for both sides, but I hope you've gotten the point by now: the world isn't black and white, business decisions are hard, and the rest of the world already knows these things.

Good for them, but...

They retain FOREVER what books or other products you've purchased, and they retain all your amazon searches. They use this to 'recommend' other titles or products you might be interested in. Their internal data tells them exactly what percentage of people will "bite' on this method. If it gets any more sophisticated you won't have to order anything anymore. Just check the box that says, "Send me what I was going to order." This is not transparent. They tell you what they are doing, and you can even go to the trouble of modifying how the tracking system works. So, for example, if I buy a toy for my kid I can uncheck that item so I won't be innundated with "If you liked this toy you might like this other one, too."

So, if you decide to buy "The Anarchist's Cookbook," you better pay cash at a local physical bookstore and hope you don't have to sign for it, like you do with ammunition.

Let the Browser Do It

Most web browsers can already remember what you've previously entered on forms. The only risk of a privacy breach is if someone visits your house and starts to type "britain" and sees that it wants to prefill "brittany spears nude." Sure, you can't log in from another machine and scroll to that one time you searched for "why does it hurt when I pee," but isn't it just as easy to type that again?

I heard that the average Google search is about one and a half words long. And if the search went well, you found what you needed already. How often do you need to search again?

Why do users EVER want to save their old searches?

I have never understood why anywone would want to save their previous search queries. The search strings themselves are usually not very long so it is not to save laborious typing. What is it for? I understand why the search engines want to save all search strings, but why would I, a user, ever want to review or re-search my previous queries? I do not care what I searched for a week, a month or a year ago, so would someome who cares please explain it to me? Thanks.

No more BocKView??

Damn, that was very useful for Manhattan. Any alternatives out there?

I like saving my searches

I enjoy the fact that I can easily see what and when I searched, using Google's search history. I use it and enjoy it. I have nothing against servers recording my personal information as long as I can benefit from it. If they record what I do and provide me no benefit, that makes me cranky. And no, better targeted advertising is not a 'benefit'.

Bah I liked Instant Reward

I got my half-pi discount on Amazon on a lot of stuff, and I wanna keep using it!

You mean I have to give this up because of all you privacy nuts?

No More Block View?!?!

Oh hell! I must have processed thousands of those block view records via Amazon's Mechanical Turk program.

All those countless hours and all that work DOWN THE DRAIN!!!

Well, almost. I did get a good $4.23* out of it.

* = All kidding aside, I did manage to get about $60 into my account after a fair amount of work, but still...that sucks. :(

To all those that "don't care"

anything they keep, the government can subpoena, anything the government can subpoena can be used against you in court. And yes, what books you read, and what music you listen to and what movies you watch CAN be used against you. Has it been so long that we have fogotten the fascists FBI "commie" hunts of the 50's. Where they were snooping on anyone for anything they dreamed up, library card records, anything. Just hope you remember your apathy when you are the target of one of these soon to be reinstated for citizens "terrorists" hunts. Does it only really take a single generation to "forget" the civial rights breaches of the Hoover and McCarthy years.

dropping the discount makes sense

It was a gimmick to get people to try the search engine, and hopefully like it. However, once people tried it for a week or so, they either liked it or moved back to their favorite engine, just sending enough searches A9's way to keep the discount up. So it was still costing revenue but not really generating new active users out of those signed up for the discount. A case could be made, therefore, that the cost per user continued to rise over time.

Are there OSS alternatives to the A9 toolbar?

Personally I really liked the A9 toolbar. Using multiple machines I used the bookmarks very often and the history was very helpful a more than a handful of times. I'll miss it but at the same time giving Amazon access to all that data on me started to bug me so I'm glad it's gone so I won't be tempted to use it anymore. I just hope they trash all that past data on everyone.

Are there maybe open source alternatives to the bookmarks and history in the toolbar that maybe I could serve from my own box?

Reminds me of Sherlock

Looking at the www.A9.com, its feels very much like Sherlock [wikipedia.org] on the Mac. That is, instead of being a general search engine, like Google, it is indexing specific major sites and allowing you to search within them. The one thing that Sherlock had going for it was the ability to add in other sites, that weren't orginally provided by Apple. I suppose Amazon decided that they weren't winning any ground in trying to by just another search engine, that they decided to make the engine different to what is already out there, therefore adding an added values to make them worth considering.

A9 dropping features

Initially I was devastated that my favorite start page was dropped without any advance notice. Funny how we get used to something even though the limitations drive us nuts. Oh well...

Re:1-Click Ordering, but a 3-step sign-out procedu

Don't save your cookies and you'll get signed out every time.

Re:1-Click Ordering, but a 3-step sign-out procedu

When you click "Sign out" you're logged out. That's one step. Because you saved your login in Firefox, your username and password are prepopulated with your locally-saved values. Most web sites, including Slashdot, send you to a login page when you log out; if you saved your user/pass in Firefox they'll be prepopulated there too.

Disclaimer: I work for Amazon. This post does not represent the official opinion of my employer.