For what it's worth, David "Waggish" Auerbach thinks that it's not specifically the content farm/spamware changes in Pandas that did it but favoring "premium content" (sites that keep people reading longer and more pages after they arrive from a search).
The answer to how long they'll be able to keep the site alive is "probably for a long time", since they're going to be in the black after the layoffs and as a result of this Matt Haughey has added the option for NPR-style subscriptions. But it's crummy that he had to let three people go in order to get to that point.
1.1: wait, sites that do that which are free? That is some shady business, google!
Does Google like or penalize "premium content"? The former makes sense, but the latter is what sounds like is happening if MeFi is getting punished.
I don't understand the internet outrage over this. I've used ask mefi, and I'd say that I get a useful answer about 20% of the time. Not that 20% of the answers are useful, but that there's a useful answer at all, among all of the answers, 20% of the time.
Many answerers clearly haven't read the question. Worse yet, with most answers, you can't tell for sure whether or not they've read the question because they give an answer that's completely irrelevant.
If I'm not asking the question, and it comes up in a search result, there's a good chance that the question itself has become irrelevant with time, or that the question is so specific that it's not relevant to what I'm looking for.
So, I might care about the metafilter answer to some question 5% of the time, or maybe 10% at most. That's not quite yahoo answers territory, but it's bad enough that I don't mind that metafilter isn't among the top search results anymore.
It's sad when a business has to lay people off, and it's weird that metafilter is getting penalized for having inorganic links (if that's what's actually happening), but there's a lot of outrage that ask metafilter doesn't show up on the first page of search results very often (not in the blog post, but in others), and I don't see why that's outrageous.
This is from my usual position of total ignorance (not a MeFi person, know nothing about Google's processes) but I think the outrage is from a perception that Google, which feels like a totally transparent, objective, utility, is specifically biased against a site people like for obscure and possibly ill-founded reasons. The outrage is about fairness, I think.
Seconding LB's 3: it makes no sense that Google should favour sites that keep people reading, and thus not being exposed to more ads with every click.
Either way, it is a horrible illustration of the way that we are all sharecroppers now, and that "the internet" is 99% of the time whatever Google chooses to cache and show us.
My view is that Google in general, effectively is a utility now, and can and should be regulated like one, if it remains in private hands at all. Of course the odds of this happening are vanishingly small.
position of total ignorance...Google, which feels like a totally transparent, objective, utility
That's about right! I shouldn't pick on you. But Google is notorious for being secretive and unresponsive about its algorithms and their effects on search results.
This may be a bias because I'm a programmer, but I don't see how any major search engine could hope to be transparent. SEO folks would immediately flood the top results with spam if they knew what the rules were.
Even DDG, which is as open source as a search engine can be, doesn't open source its core ranking algorithms.
Not trolling; I'd imagine most people here agree, except maybe the ones who actually work at Google. Of course "regulated like a utility" leaves open a lot of room for what the specific regulations should be.
4: Your experience and mine (both as an asker and an answerer) are, if not diametrically opposed, pretty damn close. And one of the great things about Ask is that things like this happen.
I'm not sure I understand how 7 would work.
You can figure out the details when I make you minister for expropriated Google property.
Luckily for you, I'm not one to turn down arbitrary power just because there's no good end for it.
It would be straightforward to make the workings of the search engine transparent if people found guilty of SEO were killed.
And one of the great things about Ask is that ... this happen[s].
That is remarkable, and touching.
12: I suppose it depends on the type of question. If I ask a question like "Can I find an mp3 player with features X, Y, and Z?" I invariably get answers like "I don't know if the ipod mini supports any of those but I love my ipod mini!".
Looking back at my questions, literally every question of the form "Can someone recommend an item or service that meets this criteria?" is filled with useless answers. The other category of questions that's a complete miss for me are finance related questions. It's probably not generally bad for finance in general, but anything obscure or technical seems hopeless.
Questions like "Can someone tell me what book I read? It had a character with brown hair who . . ." seem to get answered pretty well, but I never ask questions like that. That goes for the vast majority of questions that get decent answers.
16: not all of the subs at my place of work deserve this. Couldn't we just have flogging for a first offence?
When I was pregnant and looking for good online info about drinking alcohol, MeFi had the best discussion (with links!) I found.
The general idea as I understand it -- and I don't do SEO, so this might be mistaking thing -- is that Google is for "premium content" over "thin content", which includes a) people clicking a bunch of links and spending time on the site b) social shares c) original content, particularly images and d) lots of text. The idea is to cut down on the ability to make a billion pages offering the same boilerplate ("get an insurance quote in $CITY_NAME!" or thousands of different search terms giving the same items on an ecommerce site). But when you're talking about an answer site like AskMe, you're getting none of these.
I thought the linked article was pretty interesting for the light it shed on a) how an ad-dependent internet business works and b) how arbitrary and opaque Google's indexing process is.
The original, eigenvalue-based process that rewarded comment spam and link farms was an elegant piece of programming. It's worth understanding how different current indexing practice is.
Huh. I'd think MeFi would do well on 'time on site'. When I follow a link there, I usually end up reading the page for a couple of minutes. Admittedly, there's usually nothing leading me off that page to another MeFi page, if that's the problem.
Fucking web advertising. I feel like its gotten so much worse lately. I've been looking for Minecraft Mods lately to install on my kid's server. Just about every link to a mod that somebody posts goes through Ad Fly, which is about the shadiest URL shorener I've ever seen. "Here, let me try to get you to install malware while you wait for 5 seconds to reach the URL you want." "Hey, wouldn't you like to buy this DDOS tool?"
I really bugs me because a lot of the audience for this kind of thing is kids and teenagers, who aren't necessarily savy enough to dodge the malware. But the Minecraft community seems to be rife with it.
Seems like advertising ruins everything.
Meanwhile, adsense revenue on my Magic 8 Ball site is half of what it used to be. Google keeps pushing me to use image ads instead of text-only, but I think that would hose the user experience.
I agree about web pages being reduced to a hard-to-spot content banner. really sucks. But watching Mad Men streaming actually has just a few brief pauses for ads.
I still browse MeFi pretty regularly.
Huh. Sounds like I'm in the minority on ask mefi, but boy have do I get a lot of bad answers there.
I gave it another shot a few weeks ago, after more than a year off, and the answers were the same as always. I asked where I can find a local doctor in some specialty without a multiple month wait, and pretty much every reply was something like "My cousin used Dr. Smith. He had a three month wait but it was totally worth it. I recommend Dr. Smith". It's not that the question was intractable -- I managed to find someone by asking around, but basically all of the answers on mefi were for literally the opposite of what I was asking for. Ok, maybe not literally the opposite -- no one recommended a plumber with a three month wait.
Maybe it's better for open ended questions?
You know what that sounds like? You're asking questions that want a giant base of users that don't individually know much -- "Who's a fast local doctor" doesn't need smart people, it needs enough people that someone has happened to use the doctor you want. Same with "Which mp3 player has these features"? Not exactly the same -- there might be an mp3 player hobbyist out there who knows all of them -- but your best bet for an answer is a whole lot of people, one of whom will look in their pocket and say "Mine does!"
My impression of AskMe is that, while it has a decent size readership, it's going to do better on harder questions that involve a little more thought.
10 is certainly the party line, and everything I've dealt with myself (I don't work with search ranking) supports the general idea. Doesn't make me much happier about empowering sharecropping, though.
29: This post was deleted for the following reason: This is pretty much the definition of chatfilter.
28: Yeah, a question like this is in what I would consider AskMe's sweet spot. Ask that question on a dedicated bike site and you're likely to get knowledgeable answers mixed in with a lot of attitude; ask it on Yahoo! Answers and you're likely to get dross. On AskMe, though, you get people who know about the subject matter combined with enough moderation to make them not be dicks about it.
28/29: Alright, I'll give it another shot next time I have a question that's less open and shut, but considering how badly it does on questions requiring no thought, I find it strange that it would be better on questions requiring some thought. I mean, are people who are unable to pay attention to one trivial constraint going to do better when the question is more complex?
Another example bad question is, what are some recipes that use ingredient X but not ingredient Y? Most answers use ingredient Y.
I mean, are people who are unable to pay attention to one trivial constraint going to do better when the question is more complex?
Someone who hangs out at a place like AskMe is going to be someone who is compulsively helpful and a compulsive showoff about how smart they are. A question that very few people will know the answer to (which doctor in a specific location will give me a fast appointment) but which is easy if you know is going to lead a population of those people astray: they really really want to give you a helpful answer, but they can't, so they say something that seems thoughtful but is in fact irrelevant rather than letting the question go unanswered.
A more complex question but one that more people can get some kind of grip on, the actually knowledgable people will show up before compulsive helpfulness incites everyone else to just start saying things.
(I may be wrong, but it's a frame of mind I lean toward myself, and that's what the AskMe threads I've read look like.)
Ok, that's totally plausible, and it would mean that I've been going the wrong direction by asking more and more trivial questions each time I get a screen full of bogus answers to a question.
34 gets it exactly right. (Well, except that the "compulsively helpful" part is completely outweighed by the "compulsive showoff" part. Read AskMe for any length of time and you'll start to recognize the repeat offenders.)
I shudder to think what would happen if you asked AskMe for food or bike advice.
37: It's like you didn't even click the link in 32.
is going to lead a population of those people astray: they really really want to give you a helpful answer, but they can't, so they say something that seems thoughtful but is in fact irrelevant rather than letting the question go unanswered.
This is indeed the very essence of the Wrong Askme Answer, and it can be amazingly irritating.
I did the 34/39 thing just yesterday, because I misread the guy's question. (But my answer was better than an answer to the question he actually asked!)
I almost never read MeFi, but when I was first diagnosed with gestational diabetes (and I wasn't even telling anyone here I was pregnant because I was old and sick), I was worriedly googling for info about some specific issue with waking glucose levels. None of the usual health sites were helpful, but then I found an AskMe thread that seemed close. So I am reading along in the responses and hey! Here is an answer that is actually really useful to me! And . . . it was written by redfox! Talking about snark having a similar thing! So it was like my friends understood me, man, and were helpful to me without even knowing. So there.
36: Fortunately, there are other places one can comment where people don't compulsively show off.
Wait, Metafilter is a *company*? I thought it was just someone's blog?
45: it's a floorwax and a dessert topping.
It was somebody's blog, then got big enough a decade or longer ago to start hiring professional mediators and backroom staff, but it's still basically a hobby site run basically to keep Matt Haughey in bike parts.
Weirdly, if I understand correctly, this isn't happening because they disappeared entirely from organic search results. This is just the difference between being ranked the fifth or sixth most relevant result instead of the second or third. The click-through rate drops off that quickly. And the business model is that sensitive to Google's opinion of you as merely "very good" rather than "excellent" (or, really, merely "excellent" rather than "superb").
This is just the difference between being ranked the fifth or sixth most relevant result instead of the second or third.
Yes, and there is a chasm between being #10 and #11.
Its also increasingly the case that even the #1 slot is way down the page, as the top section gets filled with paid ads. There are a lot of people on teh internets who don't really recognize the difference.
Weirdly, if I understand correctly, this isn't happening because they disappeared entirely from organic search results
David Auerbach's article suggests that for many searches for things that are obviously relevant they're now several pages deep in the results for google. While being in the top one or two on bing.
46 displays exactly the kind of attitude that Matt's railed against multiple times: the notion that somehow self-sustaining businesses that don't have aspirations of taking over the world are worthless. MeFi's been a good enough business to provide multiple people with fulltime employment with full benefits; it's pretty fucking far from a "hobby".
Blogging is his profession. Rallying against the attitude that self-sustaining businesses that don't have aspirations of taking over the world are worthless is his hobby.
Here is an answer that is actually really useful to me! And . . . it was written by redfox! Talking about snark having a similar thing!
Wait, Snarkout had gestational diabetes? Either names were swapped, or I have have been confused for a long time.
52: or I have have been confused for a long time.
Psst... helpy-chalk didn't know men can have babies too, pass it on.
I would have thought that Martin, being an active user of MeFi, would be aware that calling it a hobby is bizarre.
52 - Nope, just regular-type.
53 - Sssshhh! Don't let him know!
52,53,55 unfortunately remind me of the time I asked my pregnant coworker with Type 1 diabetes if she was at extra high risk of gestational diabetes.
I thought Type 1 and gestational were considered different conditions.
They are, I think, but it's not like they need to tell somebody with Type I diabetes that they really need to control their blood sugar lest they get diabetes.
Also I think the "giant baby" outcome is the same with both; this same woman just gave birth to her second child, a happy, healthy 10lb 14oz baby born at 36 weeks.
Ah, right - Type 1 is the chronic version. Although it does look like it works the other way - gestational increases risk of Type 1.
Apparently there is a study linking gestational diabetes to late on set Type 1 diabetes. But I think maybe you meant to say Type 2 because that's really a much bigger concern as far as new incidence in women of age for having teh babies. Type 2 isn't the same as Type 1 or gestational diabetes.
Chance of Type 1, or of Type II?
There's more late-onset Type 1 diabetes than people thought, apparently.
I meant Type 1. It was late (adult) onset, but well before she got pregnant.
It's sometimes called diabetes type 1.5 or Shel D.B. Silver.