Ask HN: "Basic SEO"?
I see this phrase thrown around like a magic spell from time to time, a la "Put up a site about a profitable niche, do some basic SEO, and bam! presto! Instant $50/month!"
But what is "Basic SEO" to you? Blogging? Adwords? Paying dudes via MechTurk to write articles for you?
Googling "Basic SEO" brings you a few good articles, but a LOT of "buy our ebook" articles too. I just feel that if it were so basic, it'd be more common to find.
So what do you consider to be "Basic SEO", HN?
58 comments
[ 3.6 ms ] story [ 119 ms ] thread1. After that, make sure people link to you with proper anchor text for other keywords. If you want people to search for "fishing tips" and find you, then several people will need to link you something like this:
2. The more authority these links have, the better you will do. If you get a very high PageRank site to link "fishing tips" to you, you might be immediately first for that query in Google. Or, 2-3 medium links might do it.3. The words you want people to search for need to be used several times on your site. You should have the words "fishing tips" on several pages and you should link to your best page on "fishing tips" by putting that phrase in your own links to your own pages.
4. Also make sure you put the keywords you want searches for in the title of your page, and enclose them in H1 tags or other bold/header tags. This won't help very much, but it's probably worth doing.
5. Links don't really help you unless they are from a real domain - a link from someblog.blogspot.com will not help your PageRank much at all. Also, the domain needs to exist for a while to help you - something like a year.
6. Good places to get links are from your college and high school newspapers, local newspapers, and anyone else who has a website that would appropriately cover you.
7. Some sites have way more PageRank than you might expect. www.cmu.edu is an anchor site for the link graph, and a link from this site will do wonders for your PageRank.
8. Here are some excellent pages on SEO:
You should also check out Mahalo.com. That site is SEOed within an inch of its existence on Google, so take some tips from them but tread carefully.9. In the end, it really boils down to having authoritative links with the right anchor text linking to you. The rest matters very little.
http://www.seomoz.org/knowledge/duplicate-content
The reason there's so much mystery around SEO is because you've got a black box called Google which may or may not do what you want when you want. You can't run any real scientific method on it because, just like how you can't cross the same river twice, you can't evaluate the same Google twice (since they're constantly running experiments on you, the surfer, in real time), and obviously you're competing with a gigantic graph of web sites, which you also can't evaluate twice, since it changes every time a new page goes online, goes offline, or gets republished with some tiny edit.
Most SEO people don't think this way, but because I'm a programmer, I think you should start learning SEO by understanding how Google works. For instance (blatant self-promotion), you could check out this video where I give you Google PageRank in five lines of Ruby:
http://gilesbowkett.blogspot.com/2010/03/google-pagerank-in-...
There's also a free SEO book which is supposed to be great. I haven't had time to read it but I hear great things about it.
http://www.seofaststart.com/
1 Speed matters. Google is on record about this. They want your site to load quickly. This factor is part of the 'caffeine' update that Google is in the process of rolling out http://searchengineland.com/site-speed-googles-next-ranking-...
2 Don't stop with H1s. Think about the structure of your page. Subheaders should be H2, and so on.
3 Use Google Webmaster Tools to submit a sitemap. They'll regularly pull it down from your server so as you add pages to your site, Google will know to index them. They'll tell you how many pages you have submitted, and how many are indexed.
While you're in Webmaster Tools, check out if your site has crawl errors. Check out the HTML suggestions. Google will point out the pages which are missing title tags, which have duplicate tags, short descriptions, etc. Lots of other information in there, including page load times.
It can give you a few days lead over the cloners and that's sometimes all you need.
Setting up a sitemap is usually just a placebo for finicky webmasters, and it's a waste of time to boot. The biggest sin at a start-up is wasting time, and this easily qualifies.
The only time you want a sitemap is when you don't have a text front-end, and even then, you are way better off mimicking your JavaScript with a text layer for the search engines that implementing some sitemap.
Even submitting a page in a sitemap doesn't ensure it gets crawled, and it certainly won't get crawled if you have a large site and don't set up your sitemap just right.
That's cool with me.
> Have you actually ever had someone clone your site or tried to protect against it this way?
Yes, otherwise I wouldn't have written that, now wouldn't I?
> I am all ears for your story... I have never heard of anything like that.
That you haven't heard of something is not a reason to assume that it doesn't exist, there are actually more things that you probably haven't heard about that aren't bullshit either.
In fact, there are probably lots of those things.
I recently did a fairly large project, well documented here on HN and in the press where I did just that. Some less than honorable characters were cloning the data as fast as it hit the server and turning it into 'mfa' fodder, so at some point I shut that down and submitted the site to google which then crawled it at it's leisure.
Even today that crawl isn't 100% complete yet, and the only way you can reach those pages right now is by going through the google index.
Long term I expect plenty of those pages to be linked again, both internally by linking pages that are related as well as externally by people that link to their content.
I get about 50 emails daily confirming that the strategy worked, those people have lost their content and find it again through google and this is the only copy of it on the web, in spite of active attempts at making clones.
The ip blocklist for that site is close to 10,000 IPs.
> Even submitting a page in a sitemap doesn't ensure it gets crawled, and it certainly won't get crawled if you have a large site and don't set up your sitemap just right.
No, if you don't set it up right then of course, it won't work, but that goes for most things in the technology world.
Right now there are 197,000 pages indexed on that site according to google so I really can't complain, it seems to have worked very well.
Still, it seems really odd to be using a sitemap like that. That's certainly not the intended purpose, and if that were a reasonable defense, what's to stop the spammers from employing it as an offense?
As far as I can tell, you use a sitemap if your site doesn't map in the normal way. Even then, it's not nearly as effective as a text site with any in-links - so much so that I wouldn't hinge any SEO strategy on a sitemap, and I wouldn't recommend setting one up to any new start-up founder. Even if your observations on this are correct, then you are a very special case.
It's also worth pointing out that it seems your site map was no defense until you took other measures, which wasn't clear from your original post:
"so at some point I shut that down and submitted the site to google"
Responding to some of your points:
"I get about 50 emails daily confirming that the strategy worked..."
* This confirms that your site is indexed, but it doesn't confirm your strategy beat the spammers. It may be more of Google's algorithms deciding you are best, no? I can't say for sure obviously, but you might have fared better if you hadn't done this at all.
"Even today that crawl isn't 100% complete yet"
* Part of the reason for this is because your pages have a low-priority to be crawled, because they are submitted via a site map.
no problem.
> Still, it seems really odd to be using a sitemap like that. That's certainly not the intended purpose,
Agreed, but that's what we're hackers for right ?
> and if that were a reasonable defense, what's to stop the spammers from employing it as an offense?
That they didn't have the URL, but the google bot did.
> As far as I can tell, you use a sitemap if your site doesn't map in the normal way.
Or if your site has crappy navigation, or if you want to comply with some countries' accessibility rules.
> Even then, it's not nearly as effective as a text site with any in-links - so much so that I wouldn't hinge any SEO strategy on a sitemap, and I wouldn't recommend setting one up to any new start-up founder.
Agreed, but we're not all just start-up founders, and even if we are we're not all above using the occasional trick to get an edge.
> Even if your observations on this are correct, then you are a very special case.
That's possible.
> It's also worth pointing out that it seems your site map was no defense until you took other measures, which wasn't clear from your original post:
"so at some point I shut that down and submitted the site to google"
The thing I shut down was the publicly accessible version of the map, so only google had access to the real thing.
> Responding to some of your points:
"I get about 50 emails daily confirming that the strategy worked..."
* This confirms that your site is indexed, but it doesn't confirm your strategy beat the spammers. It may be more of Google's algorithms deciding you are best, no? I can't say for sure obviously, but you might have fared better if you hadn't done this at all.
Google keeps something called a 'quad' list if I'm not mistaken, which contains a series of word ids in sets of four for every page that exists in their index. If certain sets of 'quads' are unique to your page then those are used to judge your page as original for that bit of content. I gather that quite a few of the 'random keywords spam pages' are built on that predicate. I'm not sure if that is outdated or even plainly wrong but it would explain the pattern in clone sites sometimes ranking higher than the originals simply because they got crawled earlier.
> "Even today that crawl isn't 100% complete yet"
* Part of the reason for this is because your pages have a low-priority to be crawled, because they are submitted via a site map.
That's quite possible, but I'm not in a hurry. The project was huge and if it takes a year to get it indexed I'm perfectly content with that.
The 197,000 pages result in about 7000 unique visitors daily, the total number of pages is in the millions.
I'd add that SEO is not just a ranking exercise-- it's a conversion exercise. On top of all of the factors that you need to consider for ranking, you should consider that the first ~60 characters of you <title> tag and the first ~150 of your meta-description are what drives people to click through to your site.
Also, all other things being equal, the first position on SERPs gets ~42% of clicks... So it's a winner take all game unless you are talking about a content business.
OH-- also do some keyword analysis. In Andrew's example-- how many people search for "fishing tips" versus "fishing advice" or "how to fish" or "intro to fishing"? Keyword analysis is about finding out what people are searching for and then digging in to see how competitive those keywords are. If you want to dethrone the #1 result for "fishing tips" you'll want to analyze their SEO. If they are a 10 year old page-rank-9 domain with 87,500 inbound links, you might consider gunning for a less sought-after search term.
2. Make sure those keywords are on your page and maybe urls also.
3. Get links. More the better. Not from spammers.
1: Making sure your site is visible to search engines. Make sure there are meta-tags, that navigation, headers and text is text, not images or flash. Have a sitemap. Make sure links to pages contain a good description of what's on the page.
2: Making sure someone's linking to your site. Pitch to blogs to get them to talk about and link to your site. Comment on blogs, putting your website in the "website" field. Have good content that people will want to come to and share - the best way to be visible is for get other people to do your promotion for you.
The reason it should be a niche, is that when there isn't too many people writing a website, it's easier to rank high for a search term. If you picked generic technology news, you'd be hard pressed to break page 10 on Google. If make a good site about how to grow your own organic hamster food, you could make page one with little effort. Picking the niche is as much basic SEO as the technical stuff for this "trick".
Finally, TANSTAAFL.
1. SEO is not keyword frequency. Although the page you are optimizing needs to contain the search term you are trying to rank for it shouldn't be saturated with that term. The first four words of the <title> tag, the <h1> tag and maybe an <img> alt attribute are a few good places to put a good keyword.
2. More links isn't necessarily better. A couple good links from one or two good pages can really boost your rankings more than a lot of links from unimportant websites.
3. Meta tags won't help a page rank higher. The <meta name="keywords"/> and <meta name="description"/> tags aren't used as ranking factors but it's good to include them because they do signal "about-ness" to an engine.
Some things that might hurt your SEO are frequent server downtime and site inaccessibility, cloaking by user agent, links from spammy domains and acquiring links from known "link sellers".
Source: http://www.seomoz.org/article/search-ranking-factors
In order to find what terms you should associate with your site, do some basic research on search volume (google has some tools for this). Once you have the terms you want to rank for, it's a matter of making them more prominent in your site. Header tags, the title tag, the first 100 words of your body copy (p tags), and your URL structure (readable & hierarchical) have the strongest impact on this, as does the anchor text of both internal and external links. Meta description doesn't impact your ranking, but it does impact click-through in the search engine listings because it's the default text the search engine will display under the title.
The last part is making sure that content is available in text form on your page. Banners that are in flash or image files should either be replaced with images, or you can do graceful degradation by having the text be replaced with images via javascript (this is consider OK by google so long as the replacing image contains the same text). If you have videos, include transcripts or at least summaries. If you have lots of images, use the alt attribute. Don't make things Ajax unless you really need to.
The last recommendation is don't spam. You'll get penalized. Spamming is putting more SEO-bait than content on your page, and/or placing content that search engines will see but most users won't (putting the same content in different forms is ok). Most tricks you can think of, someone else has thought of first, and you will get burned for it. Done properly, white-hat SEO and accessibility overlap very well, since search engine spiders are basically low-capability browsers.
- Research your keywords, make a list, especially paying attention to which ones are the most popular. The results may surprise you.
https://adwords.google.com/select/KeywordToolExternal
- Optimize your title tags (different on each page, mention your keywords, first 4 words of the title count the most)
- Put your keywords into high-value elements (h1-h4, img alt text, ul's, dl's)
- Get links from other websites (best case scenario: link's anchor text is your keyword). Higher the PageRank of the linking site the better. NoFollow's don't count.
To get links:
- Don't use subdomains (i.e., put your blog at mydomain.com/blog/ instead of blog.mydomain.com)That will give you a good start. From there, start a blog with interesting content, and submit it around. The idea with the blog is that the posts are more interesting than your site's copy, so people will actually link to it.
Side note: sometimes this is desired. For instance, one domain can only occupy 2 slots in a SERP, but each subdomain can take 2 more slots. So in theory, you could take up an entire results page with your main domain and 4 subdomains.
First decide how much an hour of your time is worth. Then decide if an investment in this website is worthwhile.
0. Find a new, non discovered niche. This niche should be gold, this mean the Adsense Pay Per Click is high. Register a related domain.
1. Purchase unrelated domain names, any domains really, get ".info" domains since they are cheap.
2. Buy a SPAM tool like Xrumer and SPAM related/unrelated blogs with unrelated keywords. (Did you wonder the thousands SPAM comments you get with xfdfrg in anchor text and wonder why the spammer didn't found better than those keywords?).
3. SPAM a couple thousands blogs, get a couple hundred back links. You need many links, your anchor is not targeted, so quantity need to beat quality.
4. After you get the back links listed in all your domains, make a 301 Redirect, this will bring all the Google juice (back links) to your domain.
5. Your domain rank high, it brings traffic, optimize it for Adsense.
Why all this mess?
A. It helps stay under Google SPAM radar since you are using unrelated keywords, in fact keywords that related to nothing.
B. Other black hat or white hat web masters won't discover your Gold niche while you are building back links.
As I said, this is black hat and involves spamming. I never tried it, but it seems to bring some good cash it you choose well the niche.
1. Leverage your competition, by finding your top five closest competitors and spend your first round of SEO getting links from the top 100 sites that link to them. Once you filter out the ones that you won't be able to contact or you know can't edit their content to add another link, you'll be down to 20% of those 500 links. Now, you can do this in a day. Create an email, send it to each one of those guys asking for a backlink, telling them you're in the space and explaining why they should link to you. Don't pay for links, because it's not worth the risk of the google "police" finding out.
2. Long Tail. Don't focus on ranking for the super-competitive keyword such as "flowers", rank for medium-tail keywords that you can make some ground on with less effort. "flowers going away party", "flowers evil mother-in-law", etc. Related to #1, when you ask for a link, feel free to suggest anchor text in the link that helps you get what you want from that link.
3. Kickstart your on-site content strategy. Make a list of 100 articles that would help you drive traffic. Take the top 10 ideas and hire someone to write them. For each article, when you publish it, find 100 more people to ask for links to the article. Send them specific text and a specific page you want them to link to.
4. Authorities. Make a list of 10 authorities on your subject who blog or create content. Write an article that you think would be interesting to them, something specific that is in their space. Make the content really good. Contact them personally (phone call, even) and ask for links to the article. In the flowers example, academics focused on environmental issues related to flower farming, for example.
5. Directories. Build resource articles that cover your space in a way that focuses on your strengths. Find directories and ask them to link to your article. In the flowers example, an article on determining the proper type of flower for all special occasions or religious events.
6. Build links.
7. Build links.
8-10. Build links.
Some article writing services do pretty much the same thing, being essentially middlemen rather than having talent inhouse, which is why I don't recommend them, but YMMV
Got a contact who takes a keyword or phrase, does research to extrapolate, writes content pertaining to it, and then delivers it. Can also ghost-write ebooks and the like. Tried three different, cheap writers and her work was better than the other two.
My tips (once you understand the really basic): http://www.gabrielweinberg.com/blog/2009/04/search-engine-op...
HN comments on my tips: http://news.ycombinator.com/item?id=553613
http://www.blackhatworld.com/blackhat-seo/adsense/42980-how-...
Which was discussed on HN here: http://news.ycombinator.com/item?id=511935
1. Content is king, obviously, but get inbound links through guest posts
2. The rest is details
The details: 1. Use alt img text to describe images
2. Use hyperlinks with solid naming conventions: http://www.localseoguide.com/yelp-seo-analysis-part-one/
and not
http://www.localseoguide.com/p=188270
3. Use rel=nofollow for links that you don't want the search engine to pick up.
For instance, if you're an amazon affiliate, your links should look like:
<a href="http://amazon.com/product/affiliateid=blahblah rel="nofollow">name of amazon product</a>
4. Pick a title tag that you think people will search (e.g. "How to Lose Weight," not your post title, "Amazing thing that I did last month")
6. Add certain pages to your robot.txt file that you don't want to get picked up (i.e. your privacy policy or terms).
7. Ensure that there are no broken links within your site (internal broken links)
8. Have keywords and a metadescription on each page
My advice: take all of the time and money and, instead of SEO, focus on making your product great or creating great site content.
If you have a site worth visiting, the quality linking and SEO will eventually take care of itself. People will blog about your site and link to it without you having to do anything other than being the best at what you do.
SEO effort is a multiplier. You can have outstanding content and make no SEO efforts and yes, you'll probably do okay. You can have crappy content and be brilliant at SEO and you'll do great, too. The best scenario is to have great content AND great SEO.
Of course, part of SEO IS building something worth linking to (and doing a bit of social engineering to get folks to link to you more than they otherwise would). Check out Mint for an outstanding example of this.
Great content/software and SEO are not mutually exclusive and 95% of the SEO tactics you need can be learned in a weekend and implemented as you go along.
So we did a minor fix and within a week Google had indexed twice as many pages as before (& ramped up traffic too).
Sometimes it's the obvious.
1) only pick a few keywords.
2) make sure you have those keywords in the title tags, in an h1 that either copies or restates the title, and in an h2 that is basically a site summary sentence/paragraph. you can use css to hide or image replace the h1/h2.
3) get some other sites to link back to you. old, established, trusted sites are best. avoid link farms. avoid becoming a link farm yourself to obtain links. links with "nofollow" attributes don't count.
4) the url you have helps. whether it is yourkeyword.com, your-key-word.com or whatever.com/your-key-word, it isn't a huge thing to worry about (especially if you can't help it). but it does help to have the keywords in the url in some form.
5) in the end, content trumps most things. search engines are smart enough to recognize most spammy things. if you want to get good results, produce a good webpage with good content.
The magic spell you're referring to is just that; a magic spell. The reality is that these people make hundreds, if not thousands of these "niche sites", and 1 in 100 is marginally successfully. They do it at such volume and with such an efficient system of rolling out template websites with useless filth the internet doesn't need that they can turn a profit.
But their real profit comes from selling their system, using select stories like the ones you've heard, or generating a month or two of profits and then flipping the website to another owner. Unfortunately, these websites rarely maintain said profit.
The real question you have to ask is whether you would rather make an average of $5 per month off of 200 websites, or $1000 per month off of 1 website.
For a technical person, there's no doubt that the 1 website is the "easier" road, and it ain't easy. But 200 websites making $5/month ain't easy either. The people doing so have an extensive and intricate understanding of internet marketing sleazeball tricks, the same way many technically-oriented hackers here know their craft inside and out (the good and the bad).
Still, there ARE some quick things you can do to optimize your pages and create your own “link neighborhood” when you launch a fresh site. Real success comes from really making something HUMAN BEINGS want to visit... but these things will jump-start you once you have that material.
1. Pick a domain name that matches your primary keyword.
2. Get other important keywords into the secondary page URLs using mod rewrite (or a platform that supports it, like Wordpress).
3. Make sure every page has a unique title and H1 tag that matches your primary keyword objectives for that page.
4. Make sure the homepage links to most, if not all, other pages (at least to start).
5. Make sure every page links back to the homepage and many other secondary pages using appropriate anchor text.
6. Register on every social media site that makes sense for you (using this list). Include a link to the site in your profile. It helps if the username you choose is a primary keyword.
7. Link the social media profiles to each other where applicable. Fill them out as fully as possible.
8. Actually use the social networks. More activity will create more pages of content with more links to the profiles, in turn passing more “juice” to your website.
9. Claim your site using Google Webmaster Tools. Submit your sitemap (preferably one that is automatically updated).
10. Do a Google Search for every one of your top keywords. Figure out how to get a link from any site showing in the top 20 results.
11. Do not under any circumstance pay someone for a link. Do not offer or accept offers to trade links.
12. Avoid linking out to shady websites of any kind.
13. Study the keywords your competitors target (if they use meta-keywords like bozos, you can just lift those from the source).
14. Write a blog, or find some other way to continually add new content. This adds to the content you have indexed, but is also another opportunity for links. Long term strategy right here.
15. Build a Twitter client, wordpress theme, or something else that people will link to and use with persistent links to whatever you decide is in the by line.
Follow these tips and you’ll end up with a few hundred links to your site. That should get you indexed and off to a start, but its no replacement for the real work of being worth caring about.
What's the last time you actually managed to follow that point of your advice ?
Google recently published an internal SEO audit. The report is more or less a blueprint of what you should follow on an on-site level with your own site and I would have to imagine that was the intent.
http://www.seositecheckup.com
If the above sounds obvious then you'd be amazed how many sites I come across which blatently ignore this - if you don't put the raw ingredients into the mix - Google aint ever gonna bake you a cake!
Basic SEO (to me) is NOT: - advanced keywords research - advanced off site SEO tactics - probably not keyword/time tracking
Joel
Ps I never really class Adwords as SEO
Pps apologies for any typos - on my BB
Keep in mind that since the content is very specific and does take some time to write (research, ...) knocking out 100 or even 10 sites like this would be quite hard work.