Hacker News new | comments | show | ask | jobs | submit login
It’s OK to make an extra $2k per month if you’re a programmer. Here’s how. (markmaunder.com)
325 points by mmaunder on May 27, 2011 | hide | past | web | favorite | 173 comments

So, basically, create a demand-media site but without even paying writers the $3/hour to create crap articles? Just scrape them from the government? :-)

I'm all for the lifestyle business thing, but there are more fulfilling ways to make $2k/month, imho... As the cost of creating businesses drops, the number of viable niches literally explodes.

This isn't "scraping government websites", this is using your skills as a programmer to manipulate data, to make graphs and visualizations of the data.

This is what I got from the article. data.gov has LOADS of datasets for the USA, as well as hundreds of other APIs from different departments that can be used to retrieve that data in real time.

The problem comes just after you click "download" and realize you have a 600MB XML file listing the Ph levels of groundwater sites around the country for the past 25 years. How in the hell are you going to make that data an interesting draw that users want to read or learn about?

I have no skill for frontend design, but I've recently begun to learn how to display data in an intelligent manner. If anyone is interested, these books have helped me:

Information Dashboard Design: http://oreilly.com/catalog/9780596100162

Visualizing Data: http://www.amazon.com/Visualizing-Data-Explaining-Processing...

Beautiful Data: http://www.amazon.com/Beautiful-Data-Stories-Elegant-Solutio...

> How in the hell are you going to make that data an interesting draw that users want to read or learn about?

People are probably looking for that data. Find out exactly what they're searching for, and tailor your content/site presentation around those keywords. You can use Google Keyword Tool to find this information:


Right, but you still need to present it in a way that is easy for people to understand. XML is a fantastic way to deliver a large amount of data, but it is hardly parse-able to anyone who doesn't know what a markup language is. JSON as well. My point was that "take data, throw adsense on it" is hardly a simple plan.

Developers need to know how to display data in an easy to understand, easy to follow, and easy to consume manner in order to get people to the site. Keywords and SEO only go so far. You still need readable content for humans.

Right, you need to curate the data as you import it. I assumed that was assumed. You need to make it readable for humans and for search engines (SEO).

>As the cost...drops, the number of viable niches literally explodes

My God, thank you for the warning. I will be alert for exploding numbers!

Obligatory XKCD: http://xkcd.com/725/

This reply is quite hilarious. I'm bursting with laughter.

Thanks, I had plenty of upvotes until you guys started replying.

My first time being downvoted, I think. Humor falls flat == bad. Got it. Sorry for dragging you down with me.

I believe that most people would be willing to set up an "unfulfilling" side project to generate an extra $2,000 USD/month.

I wouldn't. I'd rather put that time into my fulfilling business. After a year of such extra effort, it'll probably make more than $2000/month difference in my bottom line; though that assumes I have a real business to invest my time in, and that I can afford to work for a year before seeing that $2k (or more) show up.

I have my doubts that yet another generated content spam site would be worth $2k/month, though I'm sure there are a few...I kinda suspect it'd take months of experimentation to actually find one that pays, and keeps paying over a long term. If you only get one $2k month, and it drops like a rock soon after (either after competitors sweep in and rejiggle the same data, or after the data becomes stale and needs more human intervention to update).

Also, ew. I'd feel gross if I were creating spam for a living. I've come upon this kind of site quite a lot when looking at government auctions (which admittedly have horrible sites, when provided by the government), and the level of spam in the industry is just disgusting. Everybody fighting for the same keywords, clogging up google results for the real data, always out of date, often pointing to no-longer existent goods, mingling many types and branches of government (each branch has different requirements for buying from a government auction; even if you qualify to buy at one, you may not qualify to buy at several others) to beef up their listings, etc. When done poorly, it's evil, plain and simple. A low-grade evil, but evil, nonetheless. I don't want that kind of thing on my conscience, and I don't see how anyone could do it well on a few hours a week or month.

Ah, so you already have an existing business. I believe that most programmers do not.

You have to work to make money from your business (most likely). With an authority site providing data that people want and are linking each other to, you're making money passively.

$2,000/month is only $66/day. That's not a big number. That's a single sale of certain products on Clickbank[1]. If a given product averages one sale for every 1,500 visitors, your goal is 1,500 unique visitors per day. That's not unattainable, either. Making money online is not a complicated process.

I don't see how the OP's idea is anything near a spam site. It's taking data that people could be interested in, making it available, and helping people find it. If people happen to click some ads while they're on the site, great! A spam site IMO would be like those fake torrent sites that come up when you search for a product name.

[1] http://www.clickbank.com/marketplace.htm

Google Search has an insecurity complex where it will show you a ton of low quality content instead of telling you it did not find anything good.

The problem isn't that a bunch of folks have made lame sites; the problem is that no one has made a good site. Make one, and Google will happily rocket you to the top of the results page.

"Also, ew." Made me LOL.

But not when it takes 6 months to get it going. For a 6 month project I would aim higher. How about building a useful app or web based service in 6 months and making some real money from it.

additionally if you're going to be pulling down public data sets there are probably more interesting and useful ways that you can restructure/reorganize that data. There's a lot of useful data out there that is a real mess that I'm sure some niche would actually pay decently to presented in a more usable fashion

But would they pay $2,000 per month?

I used to work at a firm that paid gobs of cash for just this sort of data. The trick, of course, is that you need to be well versed in the needs of potential clients before you can even think about staking your claim. But if you know what people need, and if you can collect and effectively present it to them, they will pay for it.

More, if you do it right. Take a look at instantdomainsearch.com. It provides a tiny amount of information, all of which is readily available from other sources. It's still quite profitable.

The article does recommend buying freelance writing for $10 per page, which is more than $3/hr for the low-complexity writing presumably required. For example, he kind of writing on DMV.org, that just puts some chatty flow between steps in a workflow and around the data.

> there are more fulfilling ways to make $2k/month, imho

like what?

This to me is classic linkbait, how it is on the front page is beyond me. The "made for google adwords" sites are the bottom of the barrel when it comes to online income. I am surprised this guy doesn't have a bunch of ads on his blog so he can make $2,000 a month telling other people how to make $2,000 a month.

There is much less work involved in maintenance if you setup sites in a way such that users create the content

Theoretically, perhaps, though the initial uphill battle for traffic is harder, and if you're not careful there's always the chance you're going to get stuck policing and removing your users' content.

It strikes me that there must be huge and quiet money in some forums. One forum in particular I use a lot is Better Bidding, where people discuss how to get cheap rates on hotwire and priceline and share information on what the unnamed mystery hotels are (so far I've had a 100% success rate in finding out the name of a hotwire hotel before booking it). But every time the words "priceline" or "hotwire" show up on the site they're converted into a (presumably) affiliate link, so the owners of the site must be making one metric shirtload from those. If you can find a way to combine user-generated content with affiliate links to stuff that the users were about to buy anyway, that's gotta be some great passive income.

Oh well. If this story did nothing else it at least inspired me to dust off a blog project I've had in the back of my mind for a while (http://projectomniscience.blogspot.com) and experiment with adsense. I doubt it'll be a big money-spinner, but at least now I know how to use adsense (and besides, I got to learn interesting facts about zebras, pants, and governors of North Dakota).

I agree with what you have said, here. There is definitely money in forums, and this is surely why I have seen motorcycle.com buy up lots of forums over the passed couple years.

You describe my biggest hit; people searching for something, finding my site and writing a post or a reply that helps generate more search traffic. I don't have to police much since the sites are very focused (not forums).

How to make an extra $10,000 per month:

Find 12 people who conduct significant business online, tell them this story, and set your bill rate accordingly.

Thomas is, if anything, undershooting what this is worth if you're capable of pulling it off with a data set adjacent to a liquid affiliate market.

I think this specific article sort of underestimates the amount of effort involved ("Now that you have a site, just add marketing"), but if you're totally skeptical about it ever working, a) take a look at BCC's sales and b) think of what they would be in a niche with $50 CPA affiliate payouts which are 10x easier to achieve than purchases. BCC levels of traffic are not an unachievable goal for a programmer's first web project, for obvious reasons.

Unfortunately, every time you write a reply, I keep hearing "Hello ladies..." in my head.

What does BCC stand for?

Bingo Card Creator his website. See http://www.bingocardcreator.com/

Patio is SEO incarnate. I can't even say he "struck again" with this organic link, because he doesn't strike, he just exudes SEO with every fiber of being and press on his keyboard.

patio11 also has a blog ( http://www.kalzumeus.com/ ) about how he ran this sort of business.

How do you find liquid affiliate markets?

This kind of idea might be new to a lot of HN readers, but it's a tired old game that doesn't work all that well anymore. Working with the limited data that Google provides on keywords/search volume/CPCs, you'll be crowded into the same niches that are already hyper competitive with hundreds of other people playing the same game.

Also, this is advocating the creation of classic webspam. Creating thousands of autogenerated pages using data thats already publicly available doesn't "help" anyone.

Puzzles me when I think back to the righteous outcries on HN a couple of months ago about the declining quality of Google's search results.

I agree and disagree.

There is a lot of information out there that is terribly hard for your common searcher to find still, and if that information can be extracted from government data farms and displayed and organized in such a way that it is easy to read and navigate, this could be a really great thing.

That being said, this article does come across as: Here's how to create some Google spam for profit.

This is nothing new and Google is well aware of this tactic.

Years ago I use to buy pre-populated databases (such as all the veterinarians in the US) and create SEO-perfect websites with one piece of information on each page. It was highly crawlable and did well. I threw adsense on it and made good money for a while. A lot of other people figured it out and Google started dinging sites like this. I don't think this would work anymore.

I'd say you're doing exactly what Google wants and needs by putting the world's information online and making it indexable by them. It's a symbiotic relationship and searchers are grateful to both you and Google for helping them find what they need (that vet database you bought for example) online.

Not only that but Google profits directly through ads that appear in search results showing your data. And indirectly keeping 32% of revenue if you're running adsense.

Now if it's covered in ads that destroy the content or if the site uses cloaking or tries to trick you into signing up or thinking you have to sign up the way stackoverflow's competitor did, then that's worth a penalty.

Information that is unique and useful, presented in an honest and timely fashion is exactly what Google wants and desperately needs more of.

Don't confuse what I'm describing here or in the blog entry with duplicate content or content that has been scraped and remixed to look unique.

As a searcher, no I am not grateful to content farms, "directories", and other pseudo-content. In fact I find it quite frustrating that searches for common maintenance problems, housekeeping, medical topics, etc., now turn up page after page of superficial, ad-ridden content. I have the sense that the actual act of finding information on the internet has become quite a bit harder since, say, 2004.

What do you think of sites like DMV.org that wrap 50 godawful state websites and non-websites with nationally uniform easy to navigate content, plus ads mixed in?

It adds a lot of value to me, and Google rewards with huber page rank than government websites, which one assumes get a huge authoritativeness bonus from Google.

A lit of folks get indignant that someone is "just" exposing true information for money. That is 99% of what Google does and why everyone loves them. If you cut the money out, the ride stops. Organized information is more valuable and more costly than scattered hidden information.

Here's the question... would you pay for better information? Perhaps one with an encyclopedic knowledge that was constantly expanding, say an eHow with depth.

> Perhaps one with an encyclopedic knowledge that was constantly expanding, say an eHow with depth.

We already have one of those: it's called the World Wide Web.

If only bottom-feeders could stop getting in the way of navigating it, we could get back to answering useful questions with a quick Web search the way we used to.

People used to buy books and magazines or go to the library.

The great thing about the internet was that it replaced the hassle and expense of doing that. It works pretty well even with the nearly-content-free sites clogging up the system.

Incidentally, for Q&A, IBM might have something that will beat google search:


I think the biggest issue with the way I was operating the database-driven sites was that the content was paper thin. One piece of information on each page is not what Google wants. They want rich information from authorities on the subject. All I was doing was automatically building out sites from a large databases of information, it wasn't high quality reviews or reports.

I made sites that list British post offices and post boxes:

http://edwardbetts.com/postoffices/ and http://edwardbetts.com/postboxes/

I think both these sites went live in 2008.

Here are the AdSense figures for this month so far (May 1st to May 27th):

  * $52.94 earnings
  * 21,708 Page views
  * 147 Clicks
  * 0.68% Page CTR
  * $0.36 CPC
  * $2.44 Page RPM

You do realize the Adsense tos bars you from disclosing this data, right? Not that I care (I think the TOS is b.s. on that point), I'm just saying...

Thanks, I had no idea.

Um, yeah, I would edit those URLs out before the edit window expires. Google will kill your account.

I guess the real money comes when you help marketing people find spend-happy people's mailboxes ;)

Or connect conmen with heiresses, vacuum cleaner salesmen with Persian cat owners, horny teenagers with lonely centerfold models... I can come up with these all day, now get busy!

How did you get those sites known?

Either this program (Market Samurai) is shitting the bed or he has 291,557 backlinks to his domain. That's a pretty serious number of backlinks for a personal site.

Post a link to a site that you created following this that is making $2k a month.

No one will do this because their sites will be mirrored verbatim within 24 hours. It takes a couple minutes to write a scraper using Nokogiri. :P

A copy is just that, a copy in the eyes of Google. The copy will be penalized in rankings. Your community won't suddenly leave your website to go to the copycat. There are a million sites that copied Facebook but in the end everybody is on Facebook.

Go take a look at webmasterworld if you believe that to be the case. Panda has severe issues with determining original source.

That is true, but if the penalties were severe, you wouldn't have Monster, Trovit, Juju, Indeed, SimplyHired, etc. all reposting the EXACT same job and all ranking nice and high.

Autoblogs work entirely on copied content and do REALLY well for how little work they require to setup.

That's true but it still doesn't hurt your own site IMHO. Copying site would be way below you in rankings (especially since Panda update it seems). Whether it's possible to make an autoblog work is a different matter.

The title of the post confused me at first. I don't think I've ever seen "OK" used as a measure of difficulty.

The "Here's how" gave it away, but I originally thought the article was an argument against people feeling bad that they do development work on the side.

I've clearly been spending too much time reading English.SE...

It's not using "OK" as a measure of difficulty. It's saying "don't feel like you're wasting your time on a project just because it isn't going to make you Ferrari money". I guess it's a restatement of the programmer's adage "Don't let the perfect be the enemy of the good.".

Had the same thought. I thought the post was about encouraging programmers to start side projects that make money and not feel bad about it.

I love that quote about "temporarily embarrassed millionaires" because it's so true.

Sometimes I see people who refuse to use coupons in stores because they think it's beneath them (even if the coupon is right in front of them in a display for what they are buying).

The people I've known who were eligible to but refused to use food stamps did not feel it was "beneath them"- I'm pretty sure they felt that they didn't truly need them, and that there was someone worse off who needed them more than they did.

Actually, the amount of paperwork and hassle to apply for foodstamps is crazy. It can be embarrassing because you have to provide so many details about your life and give them legal permission to dig into all kinds of records.

I realize the need to prevent fraud but there has to be some kind of balance to allowing someone in a bad situation to keep a little dignity.

Then there is the craziness that you can buy things like candy with foodstamps but not vitamins.

I realize the need to prevent fraud but there has to be some kind of balance to allowing someone in a bad situation to keep a little dignity.

On the contrary, that would be counterproductive.

At low levels of income, there is virtually no relationship between earnings and consumption. So if a person's utility function is proportional to consumption, with a slight disutility for working, they will choose not to work.

This happens in the US up to the $30k/year level - i.e., a person earning $0k/year and a person earning $30k/year will consume roughly the same amount of goods and services.


However, if you make receiving food stamps/welfare/etc unpleasant (i.e., more unpleasant than actually having a job), you would encourage some people to work for a living instead.

What you are saying (and that link) sounds exactly like a person that has never been poor without a support network of family and friends. It's just a mental exercise with no basis in reality and no "hands-on" experience.

The modern welfare system does not allow non-disabled people to draw an infinite amount of assistance - it's very restricted and for a finite amount of time. They must constantly show real proof they are looking for work and even with part-time jobs their assistance is immediately scaled back.

Assistance shouldn't be a punishment, it should be care and kindness while someone gets back on their feet. You aren't going to make assistance less attractive than a minimum wage job, it's impossible.

Could you tell me the flaw in my analysis? Or are you merely asserting that because I lack a certain experience, I am therefore incapable of analyzing the problem using statistics and logical inference?

They must constantly show real proof they are looking for work and even with part-time jobs their assistance is immediately scaled back.

This is simply false. The vast majority of the poor are not looking for work, nor are they working. Nevertheless, they consume an average of $30k/year.

Did you even bother to read my link?

You aren't going to make assistance less attractive than a minimum wage job, it's impossible.

It's hardly impossible, but I agree that a lack of dignity is probably insufficient to push more than a few people over the threshold. That doesn't mean we should skew the incentives even more toward couch surfing and away from work than we already do.

Unfortunately for all involved, the poor are people too and to decide to make their lives as uncomfortable as possible in order to force them to get a job is maybe logical but certainly not a kindhearted thing to do.

Furthermore, it is likely that in many cases the very poor don't work because they are unable to -- mental illness is a commonly cited factor, but these days it's hard for anyone to get work -- particularly those lacking high school degrees, for instance. If someone cannot work for such a reason, then by attempting to punish them for not working you simply reduce their already-slim ability to contribute to society.

Edit: removed rudeness

The flaw in your analysis, and your linked article, is following the fallacy of Homo Economicus, which is "an arbitrary definition of man, as a being who inevitably does that by which he may obtain the greatest amount of necessaries, conveniences, and luxuries, with the smallest quantity of labour and physical self-denial with which they can be obtained." http://en.wikipedia.org/wiki/Homo_economicus

In reality, humans rich and poor alike have many considerations of equal or greater importance than their utility function. For example, people not work full-time because they are in school, because they are practicing an art or craft, because they are taking care of family, because they have a wealthy benefactor, because they are employed in an illegal trade like drugs or prostitution, or because they are trying to break into a difficult industry like film.

Please go learn what a utility function is before claiming it doesn't apply. Everything you described is simply a part of a person's utility function.

If you wish to claim that the poor disproportionately prefer to gain other things than wealth (e.g., they prefer hobbies/long shots over work which might earn them $0-30k), you are agreeing with me.

You've implied that you believe the number of poor people not working is so high because we skew their motivation toward "couch surfing" through aid programs. I am arguing that many are in fact "working" on something besides a full-time job, and would not be convinced to adopt another lifestyle by minor disincentives in welfare. Aid to the poor is a social program, not an economic widget, and surveying the number of people in the program who are actively seeking self-sufficiency is an utterly worthless metric of its efficacy.

We seem to be in agreement. The poor don't work because they don't need to work in order to consume and instead prefer to engage in other (non-work) activities.

We also seem to agree that "a lack of dignity [or other minor disincentive] is probably insufficient to push more than a few people over the threshold [towards work]" (as I said a few posts up).

I'm always glad when I can convince others (including others with whom I share different values) of the correctness of my pet hypothesis. It suggests I'm on the right track.

That link makes the unfounded assumption that the poor choose to be poor. If you assume that the poor choose to be poor, then it is easy to see that there are situations in which it's rational to choose to try to live off welfare. From that perspective it makes perfect sense to want to make the lives of poor people even more miserable so they'll choose to stop being so poor...

Also, don't forget that Clinton ended welfare as we know it, resulting in fixed periods of welfare anyway. It also resulted in some evils, like single mothers working 6 hours a day (with their kids in a daycare) making less money and spending less time raising their children.

I think the final line (heavily edited) "it is rational not to work as long as work opportunities pay less than $30k/year" is what we should focus on: let's try to make their work opportunities better through extra services (maybe mandatory without a waiver of disability or insanity) that give training and education.

That is an interesting link; but it will probably catch a lot of flack.

I believe that instead of making poverty "unpleasant", we should structure the reward system so that it drives the poor towards self-reliance using positive rewards. For example: right now, if a mother is not working, she gets support for childcare. But the moment she starts working, her support dries up, so it becomes an economic disincentive to find work. So the rewards should be graded in such a manner that they don't disappear; it should be a continuous function, whose gradient points to self-reliance. So for the same single mother in my example, for every dollar she earns, her childcare benefit would go down by, say, 10 cents. She would still get 90% of the benefit of her labor.

I'm sure if we put our minds to it, we can come up with a system that works.

Unfortunately, a lot of the current stakeholders (in addressing these issues) have a lot to gain by maintaining the status quo; they'd be out of a job if there was no poverty....

On the other hand, I had a roommate who spent his time criticizing "government handouts" while at the same time using food stamps himself.

Sounds similar to the 65+ year-olds out in the rural US where I grew up that complain about big government while defending their Medicare payments, sometimes in adjacent sentences.

are the food stamps really limited in number??

Yup, there is a limited budget and if things get bad enough in a state they will even limit them to only people who have disabilities. It's rare but has happened.

Then there is the craziness that you are immediately ineligible for foodstamps the very day you get a job, regardless if you do not get a paycheck for a couple more weeks.

You get so much per month - based on your income and fixed expenses. In my state, income must be below $1500/mo period, and then there is a secret formula that says f(income, expenses) = food stamp amount. Every six months they ask if your situation has changed.

The money has to come from somewhere. Food stamps purchase real goods, and the government pays the stores for them. So, like the money they are a mostly direct proxy for, they are limited in number.

I do not believe the stamps themselves are, but government resources are certainly finite.

On 150,000 pageviews a month I make about $50.00 from adsense. "1.6% CTR with 0.85c CPC" seems pretty unrealistic to me.

I've seen client sites that get 10-30% CTR w/ $4 CPC and I've seen 1% at $0.20-$0.80 CPC.

It really depends on a number of factors including the audience, their intent, how they find your site, the site layout, etc.

It may not be common but I wouldn't say it's unrealistic.

PS: If you're making $50 from 150,000 page views per month, are you open to selling the site?

You're right, unrealistic is not the right way to describe it. I don't doubt that quite a few people are able to make a couple thousand off adsense and 150,000 pageviews or less per month.

However, the article is saying it is incredibly easy to do this. On that point I disagree. I think if you spend a lot of time working on finding the right niche and developing the site that you can definitely pull it off, but I don't think it's easy. And, as someone else pointed out there are a lot more interesting ways to make $2000/mo building a niche web app.

Just for fun, I'll throw in my own stats: I make $650/mo from 4,000,000 pageviews. However, the low rate is easily understood because I run a forum.

I've seen 2-4% CTR across my sites. Positioning of your ads matters a LOT.

If you use Wordpress, consider using something like http://www.ctrtheme.com or ProSense.

I have about 5,000 page views per month and make about $50 from adsense.

In April, I had a 1.78% CTR.

The shitty thing is, it is an AJAX app, so there is only the frontpage, and each lookup does not refresh the Adsense ad. The average time on the site is about 12 minutes.

This is also not a very high volume search term. So the upside for me is pretty limited. Even if I ranked #1, ahead of the .gov site, I do not think I would come close to $2,000.

> The shitty thing is, it is an AJAX app, so there is only the frontpage, and each lookup does not refresh the Adsense ad.

It should be pretty simply to reload the contents of your ad block in the "success" callback of your Ajax library... you're probably leaving a ridiculous amount of money on the table.

I make ~$100-$200/month from < 10,000 pageviews (although not with AdSense).

It's pretty much impossible to claim any CTR or CPC rate is unrealistic because they vary so massively depending on the site, the market and even things like the time of year.

I get about 1% CTR with $1.50 CPC. Not unreachable in the financial space, if you have a site with relevant information. I make about the same as you do, on many fewer pageviews a month.

I'd agree, I recently tested out an AdWords campaign and the CTR has averaged about 0.05% (the maximum was 0.22%).

It's figures like this that make me think I'm not doing very well with my articles site.

0.02% CTR on 50k page views a month; revenue under $1. I guess that's the compromise you make when your niche is involved articles for a technical audience, but is there a way to improve that?

I get about 3,500 pageviews a month, ~ 2.6% CTR with a CPC of €0.10

That's a lot of work for $2,000, and the hard part is the blogging and marketing, which has nothing to do with programming.

The ??? part of his profit formula—the blogging and marketing—is a part-time job on its own.

Regardless, 150,000 hits per month is unrealistic.

Exactly. A website like this won't just start getting a ton of search traffic overnight. It takes a lot of effort building links before the site will see any traffic. This is definitely not a part-time gig.

This is true. I run a community/blogging site for fun, not profit called yakkstr.com. There are about 50 regular users and it currently has around 5k posts. I mention it because it's a good counter example. I've done no link building, and I get very little search traffic. Maybe 50 visits per day. Occasionally someone will post about a product, instead of their relationship drama, and one girl posted about "mio liquid". That post gets about 2 hits a day from google. We can extrapolate from this, 1000 pages like this would get around 2000 hits a day and from my expeience that would earn about $1.50 a day from ads.

> I've done no link building

You have 1,211 backlinks.

You have no meta descriptions on any of your pages. The links to your content pages (e.g., http://yakkstr.com/posts/3635-Cute-Friday) aren't inside a H1/H2 tag on your index pages. You are letting Google crawl/index your tag pages (pointless/bad). Your robots.txt isn't pointing to a sitemap. Are you using Google webmaster tools?

Get rid of the /posts/N-title and just roll with /title, if possible.

Set up a RSS feed and add OpenSearch.

You have a lot of on-site SEO work to do. :)

Most of those back links are from ubuntu forums, my profile, which seems to show up on a ton of pages. They are nofollow. I think I have 2 back links that are follow, both on friends blogroll, which I believe google discounts.

How much of a difference do you think those changes will make? I'm willing to do them as an experiment, except for the /title because that has caused me a lot of problems in past experiences.

It's difficult to say exactly since SEO isn't really an exact science. Some of them will definitely be an improvement. For example, it's not really valuable for someone to land on one of your tag pages.

Putting all your links to your content pages inside H2 tags tells search engines that they're more important that other links on the page.

The /title one is big. Google likes short URLs without a lot of extra cruft. You're using rails from what I can see, so the change is trivial. :) Add a route, throw an index on your slug field, and have requests to the old URLs do a 301 redirect to the short version.

Page load time is also a SEO factor now. http://yakkstr.com/tags/7-Politics takes 4.25s to load according to your X-Runtime header. That's really bad. You're not paginating, and you're only loading a few entries, so I think you're missing an index. :) Here's some tips: http://madhatted.com/2009/8/12/faster-sites-done-faster

Check this out: http://www.google.com/search?q=site:yakkstr.com Lots of empty/valueless results. Maybe you can noindex blogs that have no posts in them so they don't pollute the search results?

Your slug generator is generating bad slugs: http://yakkstr.com/posts/1793-Climate-Change--- You should look at using String#to_url from https://github.com/rubypanther/stringex instead of whatever you're using currently.

Google text ads may convert better than the image-based ads since they're usually more relevant to the topic at hand. Try experimenting with them.

You didn't really answer the question. I can do all of that work, how much do you think it will increase organic google traffic? 10%? 20%?

This is an experiment I'm trying to do ;) My hypothesis is that these things won't make a significant difference.

Your question wasn't very good. The changes suggested need to be made one time in order to affect your site's traffic many times. The changes might make the site more attractive in search results based on improved snippets. The H tags change could help google better understand the content and rank it for more appropriate searches. If you want hard numbers, it might take years of data collection and analysis to find them. If the end result is 10%, how do you know it wasn't because your users started writing more popular content 4 months after you made the changes?

First, let me reiterate that my site is for fun, and nothing more.

But let's assume that it were a business, and I'll explain how my question is not only good, but the most important question a you should ask. We have finite resources and time, and we have to decide what to use those resources and time for. How do we do that? I could implement all of these changes in a 2 days let's say, or I could use that same time and money to add some feature my users have been asking for. Which should I do?

To answer that we have to estimate in some way the impact of each. Saying it "might make the site more attractive", "could help google", etc isn't a good answer.

"If you want hard numbers, it might take years of data collection and analysis to find them."

No it wouldn't, and what I'm offering here is to do the changes and measure the change.

I would love to know how you intend to isolate variables during your measurement so that your tests are accurate.

You could never isolate it completely with only a single site, but that doesn't mean you conclude that you can't measure anything, or estimate at all. If I made all the changes in a few days, gave it a few weeks for google to notice the changes, and organic search traffic didn't increase in any noticeable way ... that's strong evidence that it didn't have much of an effect.

Now if someone is going to tell someone else that they should do xyz, what are they basing that on? Their gut? Or did they do these things in the past, and measure an improvement, even when they didn't isolate all the variables as you are demanding of me! Why not demand the same from the person making the suggestion?

> Now if someone is going to tell someone else that they should do xyz, what are they basing that on?

Fair question. I would say these things are current, accepted "best practices". I'd liken it to suggesting you use a version control system instead of emailing code back and forth. Your existing system might work, but many other people who were also emailing code back and forth are seeing gains using this new-fangled VCS thing, so it's almost definitely worth doing.

You might try searching for "SEO before and after" or "SEO case study" to see how on-site changes have worked (or haven't worked!) for other sites. At any rate, many of the current "best practices" are trivial, under the hood changes so it doesn't really cost anything to put them into place. Most of them simply help search engines understand how your content is structured and benefit both you and your users.

yes, I agree with you sort of. I"ve done a lot of SEO in the past, nothing you mentioned was new to me. But here is my experience, one of the startups I did had a brand new domain, which had nothing on it before launch, and didn't hyper optimize the type of SEO you mentioned. However, due to the nature of our site we got a lot of links for big name groups like peta, the nra, etc. Follow links on their site, all within a few months of launching.

We ranked really well really fast, whatever sandbox people think google may or may not have didn't seem to affect us. Like I said, within a few months, on a brand new domain, without an SEO expert optimizing things like H1 tags, we ranked really really well for competitive keywords.

That's the basis of my hypothesis, without good backlinks, it doesn't matter a whole lot, and good backlinks matter a lot more than tags on a page, whether you have /posts/title, etc.

But I do really appreciate the time you took to look at my site and offer suggestions. You did catch a missing index :) I didn't notice because no human goes to tag pages, and it didn't show up in 'new relic' because my other pages are fast and get hit a lot more.

btw, i'm curious if you see my site on this search, http://www.google.com/search?sourceid=chrome&ie=UTF-8...

I rank decently for long tail things like that, if I did your suggestions would that page go up? I see it in 7th position.

Yeah, SEO is really just about having good content and having links from other good sites to it. All the tweaks are just to help the search engines determine exactly what your content comprises and where it lives in relation to your other content (see Microformats/rel attribute on hyperlinks). Without good content, none of that matters.

Yup, I do see your site on the first SERP for that query. Here's a shot of the SEO breakdown for that SERP: http://i.imgur.com/URUr2.png Not bad positioning considering there's zero backlinks to that page on your site. If you had 1-10, you'd probably rank higher.

You probably wouldn't see any rise by just doing tweaks, but they're probably still worth doing overall. I would focus on encouraging your users to spread your content. Give them a "share this link with friends/embed this link on your web page" widget. Digg is dead/useless, get rid of that icon. Get a Tumblr and Stumbleupon share icon on there instead. I'd probably space them a bit more away from that big, ugly, unrelated image ad, too. :)

I did answer it:

> It's difficult to say exactly since SEO isn't really an exact science.

The answer is that there is no definite answer. :)

Well you'd want something more to go on to prioritize a roadmap ;)

Like I said, I'm happy to do this as an experiment and see what impact it has if you're interested. email me at yakkstr at yakkstr.com

What resources do you recommend for learning more about SEO?

I'm by no means an expert, though I've worked with experts and picked up some stuff from them. The rest I learned from Google and various forums:




Basically, SEO comes down to having awesome content, identifying it properly on your site, and getting other people to link to it.

What is proposed in the OP's link is NOT impossible as the various detractors in this thread would have you believe, but it IS a lot of work. The tech stuff is easy, consistently building out your backlinks is the hardest part. If you're lucky, they'll start to grow on their own and you have 100% passive income.

I didn't mean to imply that it's impossible, but that it is a LOT of work. Becoming good at SEO and marketing, then executing a good plan, is a lot of work. My site was intended to be an example of not doing that. Doing that means trading a lot of time that could be spent learning ... I don't know, machine learning or something. Is that worth maybe 2k a month? Depends on the person.

But the OP made it sound easy. It's not easy.

As I stated here [1], It's definitely not easy, though I've found the only difficult part to be getting initial traffic and consistently building high quality backlinks. It's not something that can be automated, but it can be outsourced.

[1] http://news.ycombinator.com/item?id=2593650

ug, do not waste time on sitepoint, wickedfire, or the like. seobook, seomoz, seroundtable.com, distilled.co.uk, wiep.net, wolf-howl.com, rosshudgeons.com, and citationlabs are all more worthy of your time. Read The Art and Science of SEO and attend to Mozcon in June and you'll be on the right track.

I don't know about Sitepoint for SEO, but I did like that one topic I linked to. I actually learned a substantial amount of my XHTML/CSS knowledge from Sitepoint (back in 2004-2005), so it's not all bad. :)

I like seobook and seomoz, and I'll check out the other sites you mentioned. Thanks!

The tech stuff is easy

The tech stuff is easy for us. If us programmers can figure out the non-tech stuff, we can be rich!

We have the advantage that we can automate/rip through most of the tech stuff, but there's people out there doing everything manually that are making more than you and me. :)


Is pure gold (there's tons more on SEOmoz).

For some grayhat stuff, take a look over at wickedfire.com to know what other people are up to.

I am not condoning such methods.

Yup. If you plan to enter the internet-/affiliate-marketing space you'll want to keep tabs on what this crowd of folks is up to. You'll learn that this is a space inhabited by an interesting spectrum of people...from clever, helpful people displaying an inspiring amount of hustle to straight-up criminals.

Also, careful clicking any links in those forums. Lots of disgusting, immature people over there.

Reminds me of a reddit "Ask Me Anything" thread from someone making made-for-AdSense sites: http://www.reddit.com/r/IAmA/comments/a74ba/not_an_adsense_m... (This was pre Google's content farm algorithm changes, however.)

It's an insightful read regarding the amount of work required and it seems the person was actually producing/aggregating useful content.

In principle it can work, maybe not as well as it used to if the site is too generic but it can in theory be done. But there is one black hole in this plan:

"Tell the right people about your site and tell them regularly via great blog entries, insightful tweets, and networking in your site’s category."

That being only a single line item in this plan underplays how difficult it can be to get good targeted traffic.

Didn't Hiro Protagonist do something like this?

And didn't the gravy train stop because there were so many CIC stringers chasing so little good intel?

Indeed he did!

Dudes I lived two years out of this, with the Italian adsense market that is for sure poorer than US market (and has possibly more relative competition, as free-money here is something people pay really a lot attention to...).

I was just using NNTP to fetch news and publish them on the web (now this does not work anymore, but there are tons of other stuff you can do).

So in short, yes, that works very well in my experience, and is ideal to fund your work for your startup without external help.

I don't want to promote the site; it's #88 on quantcast right now and it looks like it does roughly what is described in the article at least with regard to gov data. The parent company, which has a bunch of other similar aggregator sites, not too long ago was hiring programmers for $70 in Poland, a pretty high number over here. It seems they made good money with this.

It's the second time in a week some one gives numbers like that. Is it true you can do four figures with just 150.000 page views?

Two orders of magnitude better than I do!

I've been lurking HN for awhile. This article have finally made me create an account and post here.

I recently made a website that scrapes a popular message board and displays the archived posts. My goal wasn't to make money, but rather scratch an itch for the users of the board myself included (the message board deletes old posts and does not allow users search for posts based on user handles).

I was thinking of putting on Google ads to pay for hosting costs. I haven't bothered to do any SEO (I'm absolutely clueless in this area), but this month I have around 120k page views and 40k uniques. Most of the visits (over 80%) is from Google keywords.

Is the return high enough, like this article implies, to deface my site with Google ads? I heard that Google doesn't like content farms and my site could be considered a content farm, so would I even be allowed to put Google ads on? I would be happy to make $80 dollars a month so I can finally move from my free-tier EC2 micro-instance to a small instance... Any advice would be cool.

I'm not that crazy about the approach the author lays out to building traffic -- regurgitating public datasets -- but it is a fact that good content will probably surprise you with how it can pay off with AdSense. $2 - $7 Page RPM is totally do-able. Don't take the author's advice verbatim. Just think of whether you can build a site that will draw those volumes of page views, preferably a site with good content people want to see that will supported context-targeted ads.

I somewhat do this, but with crossword puzzle data. I'm sure there are more spammy ways to do it, but there is money to be made if you can genuinely provide data people are actively looking for and can't find.

It is much more straightforward to take on a little side work. Set your rate to $100/hr and work an extra 20 hours a month. In today's market this isn't terribly difficult to do.

Where do you find clients willing to pay $100/hr for sidework?

I demo'd some apps I made at a local group and before I knew it they were beating the doors down. Job offers too, but I turn those down.

Are you in a metropolitan area?

Care to describe what type of apps you demo'ed? Language(s) used?

Ruby development

Can be hard with the IP restrictions that go with most employers. The nice thing about something that's pure marketing & writing is that for software developers, it probably does not fall under your employer's umbrella.

One good reason to seek out employers that dont' have all encompassing NDA/IP restrictions as part of their hiring contracts. They're hard to find but they do exist, and they're worth their weight in gold.

It's much easier to get the negotiating leverage to strike out IP restrictions on your contract if you've previously worked at a name-brand company, and name-brand companies are the ones that tend to have the most all-encompassing IP restrictions.

What about arbitrage between Facebook ads and Adsense?

Has anyone successfully sent Facebook traffic via paid ads to their adsense site profitably?

My experience is that it is very difficult to send traffic via Facebook ads to anywhere outside of Facebook.

>> Tell the right people about your site and tell them regularly via great blog entries, insightful tweets, and networking in your site’s category.

This, is the one big important step separating $0 per month and $2k per month.

1. do all of this

2. realize that point 15. is by far the most difficult and time consuming of all steps

3. ...

4. no profit

You are correct in that #15 is the hardest part. The rest is trivial.

Here's a starting point: http://www.google.com/search?q=build+initial+backlinks

I like paid directory submissions, doing guest posts on blogs, building high quality Squidoo lenses (which include a link to your site), and paid reviews.

If you want the passive income badly enough, you will figure it out. If you don't want it badly enough, you don't deserve it.

It's fishy that his blog doesn't even link to his business. The sidebar reads:

  I write code and run a Seattle based Web startup.  I'm interested in ...

This article disgusts me. The internet really doesn't need its brightest minds spending their free time creating spam. How can you even suggest that's a good thing?

don't worry. you don't need to be a bright mind to do this - you only need to know hot to lie to convince people to link to your spam site. the technical difficulties of such a site are somewhere between hello world and an irc bot.

Interesting article. Have you done this successfully?

Yes, starting in 2004 with census.gov data. Then friends repro'd it in 2005-2006. Repro'd again in 2008. About to repro again in 2011. Current site launched 4 months ago on new domain that had never been registered before. Now passing 1000 uniques per day with 300,000 pages indexed and 3 million to go - and expect to easily hit the short-term traffic target in 2 to 3 months.

The site is non-spam, genuinely useful and solves a real problem. The current (soon to be previous) top ranking site for the category is awful. It's filled with ads to make it unusable and functionality sucks. It was acquired a long time ago, the core dev team has probably left and the entire category needs a refresh and I'm happy to help.

I should have added to the blog entry to look for categories where the competition is stale and has been acquired by a large publisher looking to cut costs and aggressively monetize.

What are the addresses for those sites?

Sorry, what does "repro'd" mean?


"Reproduce" is only two more characters than "repro'd." I can see why he was thrown off by that. C'mon now.


This is a completely viable concept-www.freepatentsonline.com makes at least $150k/month through taking patent information and putting it online in a searchable format.

When you take useful information and put it up like this, that also creates a lot of value for people-trying to search the USPTO for patent information is a nightmare when you don't know what you're doing.

Do an extra 20 hours of work per month?

Nice, tell more people how to make total shit sites that really don't make 2k anymore as Google already 'fixed' this bug. How did this get on the frontpage?

My back button doesn't work in FF4 on your site but does work in Chrome.

It must be your browser or some of the extensions. The WP theme works good in any browser since I'm the developer/designer that made it.

OK thanks Kevin. Will check it out.

Works for me in FF 4.0.1 on OSX. Clicked the HN link and then back and it works. I haven't done anything sinister, but will try it on a PC as soon as I can find one.

Doesn't work for me either. FF 4.0.1 on OS X.

I have Pentadactyl installed -- maybe that's causing an issue?

OK repro'd it after installing Pentadactyl. Thanks guys, working on a fix now.

Broken on vanilla Firefox 4.0.1 on ubuntu 11.04 as well

Booting ubuntu vm...

Yanked all scripts off the page and it's solved. Going to go through them one by one when I have a moment to figure out which it was.

Thanks again!!

This is exactly what findthebest is up to

You make it sound so easy, I know it is not and I can do all that you have stated, but that will take sometime. If it was that easy you can keep that process going and make alot.

I know a site that has a good amount of traffic and makes nothing from Google Ads. Not saying its not all possible, but just not as easy as you stated and somethings you left out on doing it all correctly.

Hmm, I always assumed you could earn way more with spam.

If you aren't putting in the requisite time to make a quality content-based website, you can probably generate many more websites of lower quality. Perhaps the profits from having more websites (albeit of lower quality) exceed that of fewer (higher quality) websites.

Before making a opinion about getting rich by making money online, coders need to think about the differences between real world and the software world first. How hard is to access information of building software? How much money do you pay for your production tools? Is there any production means that we should share? Think about the wars, also.

"500 Internal Server Error"

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact