SEO for Software Companies | Hacker News Reader

SEO for Software Companies

kalzumeus.com

79 pointsbjplink14 comments

This was a conference presentation for mostly folks who had never heard of me, so it has substantial overlap with what I've written about SEO. Still, if you haven't seen it it is new to you.

Folks seem to have liked it, so I'm happy.

If you have any questions, feel free to ask and I'll get back after the next presentation is over.

Summarizing: pollute the Internet with cheap content as hard as you possibly can and then "milk" it.

Assume my company has finally nailed the text recognition, I must tell you I'd be disappointed by having the need to employ $10/article English majors to keep my site "up to date".

The issue I'm having with SEO is this: if millions of fake articles/blogs didn't generate so much noise, my hypothetical speech recognition startup would have gotten a few rave reviews organically via handful of legit tech news outlets and that would be enough. But thanks to the current situation this won't be enough (I know from experience) - you need to spam the hell out of everybody to "be in".

I am sad watching Google losing the battle... It is beyond me how Digg pages (that only have a link!) show up higher in search results than the actual content they're pointed to?

Just yesterday I wanted to learn more about SAN storage and no matter what I searched for I could only get "buy! buy! buy!" links on Google's front page, until I retreated to wikipedia. Look at this crap (below), how could it possibly be included in the index, the text-to-link ratio (as well as text-to-dd) is horrendous, why not just ignore these?

Here, I think it was search result #3 for "how fast SAN storage is?", the usefulness of google is fast approaching Altavista circa 1996: http://searchstorage.techtarget.com/news/article/0,289142,si...

I like the advice (find out what your audience is searching for and write that content), though I'm not a big fan of building a ton of throwaway content ala Demand Media.

That aside, it sounds like your teacher connection is working really well!

What type of advice would you give a business with a less niche focus to try and find good keywords to write for?

For example, my salon is called "Bloom Beauty Lounge" and the site is http://bloombeautylounge.com/

99% of our keywords coming in are either 'Bloom Beauty Lounge' or 'Bloom Salon Lounge' or 'Bloom NYC Salon.'

I really feel like our google offering could be stronger, but I'm wondering if writing yet another blog post about "how to layer haircuts" is really the best way to go about it.

Any advice?

Your business has customers coming in every single day and agreeing to be restrained in a chair by an employee holding sharp implements next to their head. If that isn't a wonderful scenario for pumping them for information I don't know what is.

Ask your customers what they don't know about hair. What concerns them about their hair? What concerns them about their children's hair?

Write down exactly the words they use. Answer their questions. You'll probably notice patterns: How do I X for a person with Y hair? With a few exemplars of X and Y you can start doing multiplicative idea generation. (10 values for X and 10 values for Y = 100 articles. Get cranking. Better yet, write up outlines and get the freelancers cranking. Note that if Y is irrelevant for some value of X, and the process is actually the same regardless, your customers do not necessarily know that and they'll include Y in the query anyway.)

You'd also want to get some links to have trust to rank for non-branded pages. There is a newspaper in New York City which you might have heard of. They have an ongoing interest in racial/gender equity and many other things of concern to upper East Side liberals. "Hair is a race issue" is the sort of just-counterintuitive-enough-to-be-novel-but-nonthreatening that their editors eat with a freaking spoon. You can plumb that for many, many ideas. cough What White Moms Don't Know About Their Black Kids' Hair cough (If only there were a prominent US politician who was black, raised by a white woman, and beloved inside the newsroom -- the editors would fight over themselves about what got to cover the article.)

Does that get you started?

That's a great start :) We've definitely been successful getting links and I'll think about racially charged pieces :P

For sure the 'ask your customers and track the results then go for the most common denominator' is a great idea. We'll run with that one.

I think the authors of these SEO blogs should come clean at the end of the article about the real purpose of the article - SEO.

I know a lot of ways of getting a backlink which don't involve flying from Japan to Texas. (I went to learn stuff and socialize with folks who helped me get my business off the ground several years ago. I presented because I'm fairly decent at this topic and helping people gives me warm fuzzies.)

Well, maybe I was wrong about you - I'm cynical. When I watch a late night informercial with some guy claiming he'll sell me his super simple plan to make lots of money, I tend to think I already see his plan.

It seems like content that can be bought for 2.2 cents/word is exactly the kind of content that isn't "competitively defensible".

Any ideas for scaling the creation of higher-quality content? Ghostwriting and editing, perhaps? Interviews with domain experts instead of pieces written by domain experts?

It's not completely defensible, but it's not a lyrics site either because they'd have to write the bingo card generator too.

The point seems to be to relate the content back to something else you own that nobody else has built yet.

Yeah, you're right. By combining generic content with domain expertise (in the form of software), Patrick has created content that is defensible.

What I was wondering, though, is if he had any ideas for creating defensible content at scale w/o a software tie-in--like having articles ghostwritten based on outlines by domain experts.

Pick a niche, any niche (though one that I'd grok well would be better). I'll give you an example or two.

It seems like content that can be bought for 2.2 cents/word is exactly the kind of content that isn't "competitively defensible".

One thing that I've had some success with for a client is "Here's graph N out of the 1,200 that Patrick extracted from government data. Write the story told by the graph. Focus on X, Y, and Z."

In that case the words aren't really the attraction -- the attraction is the data (pulled out of a government CSV file that interested people can't possibly make sense out of) and the presentation of data on the website. The words are just a crutch for Google so that it knows that a bar for X being three times taller than the bar for Y actually has meaning.

Reply on news.ycombinator.com