Search Still Sucks
techcrunch.com
techcrunch.com
All evidence indicates that JC Penny was ranked #1 until the NYT provided Google with the evidence that it had collected. Google then acted on the evidence. As a result of those actions, JC Penny's rank dropped from #1 to #71.
It is funny that he should say "Google's algorithms had started to work". Presumably, the algorithm doesn't depend on a New York Times reporter telling Google that they were being gamed :)
Algorithmic changes went into play 1.5 weeks ago and had started to drop JCPenny, then the spam team gave them some cement shoes.
If I'm looking for a local vendor I go to Yelp, if I'm buying something I go to Amazon, and if I'm looking for coding help I go to Stack Overflow. I don't know that we necessarily need one site to replace Google; it seems to me that the era of free text search being the dominant paradigm may just be over.
Could you give an example or two where Stack Overflow gets you an answer more quickly than a Google search?
Here's my example: In gdb, I want to list all of the threads in my running program. A Google search for "gdb list threads" gives me the answer in the first result. The same search on Stack Overflow doesn't seem to give me the answer in the first few results. This is hardly scientific, of course, so I'm interested to hear of coding questions where Stack Overflow does a better job than a Google search.
Your post yesterday got me thinking about why Google is still pretty useful for coding questions but not so for searches of a more commercial nature. I guess it isn't worth so much money to game Google for the former (Experts Exchange notwithstanding).
Perhaps it is the era of a global single instance search engine that is starting to come to an end - maybe we need a sensible way of having topic specific search engines and/or trusted content sites and allowing an easy way of treating these as if they were a global search engine.
Pretty much the same interface but a better implementation.
And yes, I know "meta" search engines have been around for a long time but, as far as I recall, don't try and identify the best places to perform your search - they just search run your search through a pile of generic search engines and aggregate the results.
It's a natural see-saw between specialization and generalizations (there's probably a one-word term for this in evolutionary biology). We see the same in thin-client/thick-client talk in client-server/webapp contexts.
> If Google was good at search, Demand Media wouldn’t exist.
Would Associated Content? Would AOL?
- - -
This is a cheap article that plays with words; it's titled "search still sucks", which is debatable but "truish", but what it really tries to say is that search "sucks more and more", which is provably false.
And it's 100% evidence free, too, as is the rule in the new entertaining noisosphere of blogism.
> So what is the evidence that search still sucks? Well, you know it’s true, just like me.
Mr. Arrington sounds like a priest now; what's the evidence that God exists? Well, you know it's true, don't you?
-- Well, no, I don't. What else do you have?
-- Err, nothing.
Now I feel very naive. It Must be obvious I am new here. I understand no one here want to piss off this guy (like in my other comment). At least no one trying to launch a startup.
Ok now I need a new account.
I think Google results could vastly be improved if they tied together their normal rankings with a "how many people have bookmarked this URL (and to some extent, URLs from this domain)" metric. I've worked closely with Delicious' set of data, and it's nothing short of incredible. Billions of instances of people categorizing sites and vouching for their quality are going unused in the search space.
History of search:
1. New company releases new search engine, better than the competition because it pays attention to a previously ignored signal.
2. Seos work out how to game the signal.
3. New search engine is now as crappy as the search engines it replaced.
Really, internet search is like macroeconomics: soon after you understand {how to prevent recessions, how to find the best webpages}, the problem vanishes and is replaced by something even more complex and incomprehensible. Furthermore, this has already happened.
It's almost like a negative network effect. The more people that use a search engine, the more people try to game it, and the less useful it becomes for finding what you want. This seems like it'd make things much easier for new competitors like duck-duck-go to bring a superior product... at least until they gained enough market share to be worth SEO time gaming.
Judging by the votes, apparently it is more valuable to point out why a certain idea might not work, even if all of the alternatives have the same problem, than it is to suggest something with exciting possibilities. Nothing against you, personally -- your comment is quite right and I agree gaming is a major concern.
If I recall correctly, Yahoo tested using this data a few years ago and found the signals not to be as useful as others they used at the time.
Of course, Bing and Google have been working to include more social signals in rankings: http://searchengineland.com/what-social-signals-do-google-bi...
The way I see it, search engines are entirely inclusive, while bookmarking sites are selectively inclusive. For certain ultra-specific queries, I'd much prefer the search engine approach -- I want as much breadth as possible. For other not-so-specific queries, I'd much prefer the result sites to be vouched for by people. Looking one step further, probably for a decent chunked size of my search queries, I'm OK if the number of sites it is searching over is only about 10,000,000 (my estimate for how many sites on Delicious have been bookmarked by more than 20 users).
What I find exciting is that the two can be effectively combined. Take all the normal results that a search engine would show, but give it a boost based on the log of how many times it's been bookmarked. It really is that simple.
This is how Google is finally toppled. Not by a "google killer" or another giant like Bing, but by a thousand tiny niche services that get "X" search right.
I usually always find what I am looking for when using Google and haven't noticed any major changes recently when it comes to the search results. Maybe because I am taking some of my searches straight to respective sites like Wikipedia, IMDB, Discogs but still.
Also, ever since day one a bit of 'common sense' helps determining which results are actually useful (and I am all right with that).
There has been a lot "google search sucks" submission from Techcrunch lately. This was starting to feel a little weird. So I googled "Michael Arrington Microsoft" and got this :
http://techrights.org/2010/02/07/michael-arrington-microsoft...
Are people aware of this ? is it old news ? Is HN being spammed ?
That’s ethically questionable behavior but certainly not bribery. It has also no relation to the submitted article.
(I like how the author of the article you are linking to describes an interview with Ballmer as “spoke to Ballmer 4 months ago” which sounds much more suspect, especially considering the context. He certainly has a way with words.)
When a content farm pushes first pages results this causes the real manufacturers and businesses to use Google's advertising platform to reach their customers.
When you're looking for a kitchen appliance It's better for Google to return a landing page filled with Google Adsense ads instead of the official website of one of these appliance manufactures. Again for the same reason mentioned above.
It's all about the revenues.
- Reza Sardeha
I was trying to Google references to Greek Playwright Aristophanes' use of fart jokes (yes, seriously I was looking for that) and I came across this linkspam page:
http://wn.com/2934_Aristophanes (seems to have dropped in Google ranking)
For one thing, this site doesn't offend my eyes. It doesn't immediately insult my intellect, either. After the first second, though, it still proves to be just as useless.
In a way, what Cuil was trying to do was to apply enough AI to make an actually very useful scraper site. What would the world be like if such sites actually existed?
(Seems like the site is just trawling through my cookies to make guesses about what I'd like.)
Google /= search, it's just one engine. No need to throw all your eggs in one basket.
The article does make some interesting points, and I am a bit confused at the supposed sequence of events with the JC Penny/NY Times story. Plus it does raise the question of why a company who previously has had some SEO issues was - apparently - able to game Google's rankings in such a major way (using such spammy methods) without Google picking up on it for months.
That's a conclusion point from the article which I think is a well founded one.
As for the other points in the article's conclusion though:
>> "But all the evidence suggests otherwise. Demand Media is worth $1.6 billion, and their entire business is based on pushing cheap, useless content into Google to get a few stray links. If Google was good at search, Demand Media wouldn't exist."
"wouldn't exist" is subjective. If Demand Media weren't pumping out content, what would rank in its place? Part of Demand Media's success is that they are 'filling in the gaps' and covering content surrounding keywords which don't really warrant much competition. Sure this isn't always the case - and Demand Media's sites do sometimes rank highly for some rubbish content with other alternatives available - but in many cases it does seem to be at least part of the reason for DM's success.
Do I like some of Demand Media's content? No.
Do I like that Demand Media seem happy to have some pages with more adverts than content? No.
Do I think that Google et al should not even index some of their pages (some of which can at current rank fairly well)? Yes.
Do I think that there's many good alternatives to some of DM's content? NO. Unfortunately.
>> "And Bing wouldn't be making solid gains in search market share."
I might have different ideas about what "solid gains" means, although to me a percentage point here or there doesn't amount to much.
If - say - a Presidential candidate seen his/her opinion polls increase by 1 or 2%, you wouldn't see a bunch of stories saying how the candidate is making "solid gains".
In-fact, if a mainstream media source run an article on (say) a political party who seen a 2% rise in their opinion polls - and thus concluded that said party is making "solid gains" - I'd feel intellectually insulted and suspect the mainstream media source of bias.
So I definitely feel that this conclusion point is sensationalism (and inaccurate sensationalism, at that) and nothing more.
So all in all, a decent article which makes some decent points (at times). But it does also feel a bit of a shallow 'staying in the bandwagon' type of article, too.
--
P.S. Google: I'd accept, gratefully, as a stopgap, being able to specify multiple "site:foo" parameters. I don't know what the computational (or financial) impact to you would be.
(So far, I've held off trying browser extensions to accomplish this, but I guess I can start looking for one having a good reputation.)
I tried DDG some months (years?) ago, but it was too inconsistent. Maybe it's ready for prime time, or at least the Late Late Show, now. Jazz hands!
site:http://news.ycombinator.com OR site:stackoverflow.com <query>
I don't know if there's a max # of sites you can add. I tried it with three and it seemed to work great.
http://www.google.com/search?q=site:searchengineland.com+OR+...
(Yes, I did an ego search. Isn't that the best way to know if the results are likely accurate? :)
If there's a set of sites you want to search through often, you can set up a custom search engine for what you describe. For instance, see this page:
I'll have a look at your links, in a minute -- probably should have done that first.
EDIT: Well, your second link sure works. I wonder what I was doing wrong. When I saw someone else('s or s') comments online to the effect that it didn't work, I assumed that was why my own experience failed.
I'm in serious need of some coffee, at the moment. Thanks for correcting my mis-impression; I look forward to straightening it out in my mind and taking advantage of this feature.
(lowercase or just uses "or" as one of the search terms, which I think actually is considered a stop word and is then ignored.)
And I'm always a supporter of the more coffee recommendation.