Demand Media (and other ad-focused sites, but most remarkably their operation) understand this deep in their bones, which is why their content is to writing what chicken McNuggets are to chicken.
The replies to this comment seem to have missed one important point: You or I might not want to eat a Chicken McNugget, but millions do.
You and I might consider Demand Media-style content to be shit, but there's a market for it. We might want big, meaty content written with clever words and insightful conclusions, but the mass of the market wants little snackable boneless pieces that they don't have to work too hard to digest.
Demand Media understand this. They understand that the Internet is now predominantly used by people who read at an eighth-grade level, who don't know how to use browser tabs or ctrl-f, who will go straight for the back button if presented with big blocks of text. Demand Media seek to produce huge amounts of content that is just good enough to be the most relevant content for any given long-tail search term. They do that better than anyone.
Google aren't fools, they aren't being gamed, Demand Media are just creating content that is the most relevant content for very specific search terms. It's not of great quality, but it's the best that's available. If you know how to form a search query then you can find better content most of the time, but most people have neither the ability nor the inclination. Cheap, convenient, omnipresent. It's the stuff empires are made of.
Like 99.99% of products in the world, it's not bad, it's just not for you. If you are strongly opposed to what other people consume, you have three proper grown-up options. Either take it up with them, satisfy that demand better, or ignore it. Railing against the people who fulfil that demand is going to get you nowhere.
I agree with the substantive thrust of this comment, but disagree with elitism being a necessary prerequisite for it being true.
For one thing, you may be under the impression that you -- the broad "you", HN users in general -- read on the Internet. You don't. You are a paragraph-skimming button-clicking satisficing ADHD squirrel like everyone else is. Your behavior is trivially influenced by A/B tests, and you lack sufficient insight to notice it. You click on shiny buttons. You do what calls to action tell you to do, and are exquisitely sensitive to their wording. You respond to advertising. You probably don't eat chicken McNuggets, because you think only poor people eat chicken McNuggets, and you don't even remember why you think that.
I agree with you to an extent - we certainly fall for most of the same psychological tricks.
We are also clearly exceptional as a demographic. We are dramatically better educated and more literate than the mean. We're mostly male, whereas the internet is now a female majority population[1]. There is evidence to show that we behave very differently when presented with long-form content[2].
While I'm perfectly prepared to accept that our actual preferences are far more homogenous than our expressed preferences, I think the predominant error made by "our kind" is forgetting just how strange we are as a group and as individuals, at least in terms of our abilities and our expressed preferences.
As regards 'elitism', I've seen an awful lot of startups fail because they overestimated their users, but never once seen one that underestimates them. Amongst the HN community, people's first reaction to Bingo Card Creator seemed to be "who the hell pays money for that?". If it takes a bit of arrogance and disdain for us to grok the level of the average user, then so be it - no offence to HNers, but we're not exactly overflowing with humility are we?
I'm going to need to revert to Mahalo as an example of content mills because it's easy for me illustrate. Demand is largely the same, just bigger and better then Mahalo.
Jason Calacanis (Mahalo's CEO) describes the site as a "trifecta of search" that includes: SERP's, answers and community. He's really on to something with this. While Mahalo has taken the low road to amass low quality SERP's and recently killed the community it built, the answers building block is truly the most important part of the content mill puzzle.
"We might want big, meaty content written with clever words and insightful conclusions, but the mass of the market wants little snackable boneless pieces that they don't have to work too hard to digest."
No one want's big, meaty content. Everyone wants to know the answer to the question they're looking for. Sometimes, that answer can be uncovered or explained with big meaty content, but the best content will answer the question that the user is searching for with as little "meat" as possible.
It's going to take content mills a while to get there, and a drastic user interface change will occur, but eventually I think that small pieces of information will exist. Similar to what happens when you type, "What time is it in Toronto" into Google, the search engine will simply return the exact answer that you're looking for. The answer might exist on Demand, it might exist on a small personal blog, however, the SE will have identified the answer to your question regardless of if you type in "How To Shave" or "Shaving Tips" or "Shaving 101".
This applies to HN-types just as well as any others, by the way. Everyone just wants fast, simple, and concise answers to the question they have. Look at Duck Duck Go and WolframAlpha -- both are based around this idea of presenting the information you want as quickly and unobtrusively as possible.
Google is failing its users. Google is a machine, its does not have human intelligence. Google falls for simple SEO techniques, so that the first results for a query represent those site owners who best play the SEO game.
This is not a dig at Google, indeed its much better than it used to be, its just that its a machine with defined rules and will therefore fall foul of tricks (even if the tricks keep changing).
This is why Goole need to be scared of Facebook, as Facebook looks likely to be the first business to successfully implement a human recommendation level into search.
jdietrich is saying that DM's content is most relevant for particular long-tail search terms; if you edit your query to an apparently equivalent search term, you can find more relevant content. Presumably, as Google's query re-writing gets better, DM's niche (queries that are failing to return good content) will get squeezed out.
I had a number of conversations about this exact sentiment when I was writing for Mahalo.
I came to understand that there is a balance between writing enough information that users found an article helpful but not so much information that they would not want to click on ads or related pages.
The problem of course is that when one is encouraged to write almost completely with profit and clicks in mind that eventually, their body of work becomes devalued by the audience.
Search for "recover notepad files" and the first Google result is "How to Recover an Unsaved Notepad File | eHow.com "
and then the whole article is about how you can't recover the file...with 4 adsense links at the end with most having click worthy titles like "recover deleted files"
so Google users get screwed, Google advertisers get screwed, and eHow and Google both make a few extra bucks. I'd bet that all these content farms would have been long banned by Google, if they weren't using adsense.
To be fair, I'm very happy with the performance of my ads on their pages. (For better or worse, they rank well for e.g. "How to make a bingo card" and my ads on that page have 400% ROI.)
I don't think users get anything by seeing their site before my landing page, though (aside from "frustrated").
I tend to think that AdWords augments search for generic inquiries and cannibalizes it for brand queries, and I've got data kicking around somewhere here for both of those being true for BCC. (I still bid on brand queries, because I still see positive ROI on them, just not nearly as positive as AdWords thinks. Long story, ask me some day when I'm not due at a conference in 20 minutes.)
I seem to have broken the Google AdWords interface and can't get exact data at the moment, but my CPC at content farms ranges from an average of 4 cents to an average of 8 cents. They send some of the highest quality traffic which is not direct from competitor's websites. (I wouldn't have predicted that but, hey, data knows more than I do.)
I think that scum might be a little harsh. Google indexes and makes money (through adsense) on a lot worse things then content farms.
However, just like a Torrent site will rarely show up on the Google's first page for a TV show name, their algorithm needs to be adjusted to account for mills that value profit over content.
Nah, I uninstall crashy apps. And I, for one, never plan on buying anything they produce: how can you not notice 10 days of your application failing to work, and only notice because you checked ad revenue? Sorta implies where their focus is, neh?
edit: Maybe this is part of the reason. Ratings from the iStore for the application in question:
*****: 0
**** : 0
*** : 1
** : 5
* : 9
Several other applications have similarly-weighted ratings, reports of crashing, etc, though a couple are higher. Color me disinterested.
I agree with your comment overall; however, I wanted to touch upon the following quote.
> how can you ... only notice [your app is failing] because you checked ad revenue? Sorta implies where their focus is, neh?
In what business is revenue _not_ the primary focus? While I understand the importance of product quality and customer service, as well as the correlation between these factors and revenue, a company that's paying more attention to their app store reviews than their revenue is doing it wrong.
With that being said, the rest of your comment is perfectly valid but I do not feel that a business should be bashed because they watch their revenue closely.
My remark wasn't meant to imply that revenue is inherently bad. Simply that undue focus on revenue often creates sub-par products. There are plenty of small software shops out there which need revenue to exist, but create things for people instead of for profit.
My gripe is that apparently checking ad revenue every other week is of greater importance than even checking that the software works. Improving people's lives by providing something useful is clearly further down the importance list.
And some don't even need revenue - a la Linux / Rails / many open source projects.
App developer here. Is an application which turns in a small profit, so I don't look too much at it. On the other hand, Postfix wasn't running on that machine (I had been doing some testing three weeks ago and forgot to enable it again), so the emails generated by the cronjobs didn't get forwarded to my main email account.
I can see that, truly. And I do find the results interesting, though not really unexpected, if only because of the sheer number of people who clicked the ad(s).
May I recommend something to mitigate this sort of thing in the future? Set up a script to launch an emulator for each of your apps which rely on others' services, unit test their core functionality, and save a screenshot of the key screens. Run it when you feel like it, come back in a few minutes, and peek at the pictures / pass / fail.
Though, that said, I have not used the iPhone emulator. Can it be easily (enough) scripted to do something like that? I would certainly hope so, but I've been disappointed in development tools before.
Forgot to add something about those reviews. I guess you looked at the reviews for US App Store which (definitely, those aren't from Spain - there 800 review scores for this app in there), obviously, are going to be terrible for an app which doesn't show information about gas stations in the US.
This highlights a major frustration (one of many) I have with the iTunes customer review system.
I can only buy things from the UK store, so I only see aggregate reviews and ratings from UK people. If I want to see how an app/song/movie/tvshow/etc is being received by, say, the US, I have to scroll down and change stores to the US one in iTunes.
This works in the PC version of iTunes, but isn't possible for iOS devices as far as I can tell.
Even then, after trying to view a broader aggregate set of reviews/ratings, I still need to go back to the 'local' version of the store to purchase the content.
This is why there are many websites that split their 20 paragraph articles over 10 pages, the lower the actual content on a page the bigger the chance of a click on an ad, and the bigger the total ad inventory.
Conversely, if your content is too compelling nobody will ever click on an ad.
Which is when I like to see referral links for Amazon / etc. I'll happily buy something I was already going to buy, and get them some money as well. I'd be willing to bet that the amount of use of those is directly proportional to how compelling / quality the content is.
edit: though I have no clue how successful they are / have been. Anyone know what a relatively-popular blog makes from them?
This clearly increases impressions, and perhaps the variety of ads a user sees in a single session, but does it actually increase the likelihood they will click?
From the article: Scrolling beats paging because it's easier for users to simply keep going down the page than it is to decide whether or not to click through for the next page
I find 'all on one page' far more usable than having to click 'next' and wait for a whole new page to load. Many comments I see generally indicate that this feeling is common.
A former employer (and this is one of the many reasons I don't work there any more) was well aware of this and took advantage of it all the time. Just to give one example, their 404 error was a page of ads...
It could also be that the user thought, since there was no content but the ad, that the app had switched to one of those "click the ad before seeing the story" type things and clicked the only thing available to get the working data.
Also, what kind of programmer doesn't have checks in data acquisition code to make sure it is valid and immediately alert someone when it is not?!?
I think this phenomenon is mainly about UX and IxD. If there's only one route presented to the user .. many of the users are likely to choose the single route presented to them.
Perhaps this is a trite point, but I'd argue that it's a good example of how presentation of advertising should be included as a key part of any UX and IxD strategy.
Isn't that just the same as all the spammy grabbed domains? Get a good domain name, put ads on it. I've heard some people make good money with such sites.
Sorry about that. It should work fine (but a bit slowly) now. I had to stop all the other services on that machine to keep up with the traffic. My blog is in a virtual machine inside a low powered HTPC and this link is #1 here on HN and top10 in reddit frontpage so I'm getting a lot of hits.
