Mark Glaser recently queried me about improving the SEO of PBS's MediaShift. The tips and advice I gave him apply to most blogging and media websites. The piece was well balanced, with information from Poynter, and he mentioned Joost's great article on Newspaper SEO.
As Google clearly states (with their actions), bartering for links is fine as long as money is not part of the exchange (or if there is editorial discrimination and relevancy when it is). What is the guiding principal for bartering links? Massa said:
When do other people WANT to accept your link request, publish your article, run your press release or accept your submission?
When it does something for them.
Either it makes them money, saves them time, provides added value to their visitors or they believe it makes them look good or smart or benevolent to their visitors, their peers, their friends, their relatives, to the search engines, award sites or just about anyone that can make them a buck or stroke their ego.
So, the absolute best chance you have of getting that link is to cover as many of those bases as possible at the same time. When you can satisfy some need, want or desire of the webmaster, the visitor to the hosting site and it makes the search engine look smart, BINGO. You just hit the SERP buster hat trick!
If you are going to sell links, you can cloak them as being AdSense ads to lower your risk profile, because if it's Google its got to be Good! Or you can require the link purchase be wrapped in a guest article or some other format that does not look like a link purchase. :)
One of the reasons I was so motivated to change the tagline of this site recently was because the new site design contained the site's logo as a background image. The logo link was a regular static link, but it had no anchor text, only a link title to describe the link. If you do not look at the source code, the link title attribute can seem like an image alt tag when you scroll over it, but to a search engine they do not look that same. A link title is not weighted anywhere near as aggressively as an image alt tag is.
The old link title on the header link for this site was search engine optimization book. While this site ranks #6 and #8 for that query in Google, neither of the ranking pages are the homepage (the tools page and sales letter rank). That shows that Google currently places negligible, if any, weight on link titles.
I have ranked other sites for more competitive queries based exclusively on internal links (without thousands of links from other sites, like those pointing at the SEO Book home page).
Does the image alt text carry more weight? In a word, yes. Here is now I proved that to myself through yet another site error. :)
One of my hobby sites has a fairly flat file structure, and some of the internal pages are somewhat linkworthy. The site was not marketed aggressively and the only sitewide link to the homepage was the logo, which I forgot to put an image alt tag on. Google ranked 2 pages on the site well for the core keyword, but neither of those pages were the homepage. I noticed the lacking image alt tag, fixed it, and within a week my homepage was outranking the other pages.
If the only link to your homepage is a logo check the source code to verify you are using descriptive image alt text.
Would this be considered a paid link? Where is the line drawn?
Nice business model too. Are those arrows pointing at the ads legit? How about the request to support our sponsors?
Get a few hundred million in VC funding, buy some old domains. Fill the domain with user generated content. Place a thin layer of link schemes and please click ads on the top. Web 3.0...Wow. Exciting times.
The SearchGuild domain auction at Sedo just ended. $8,655 for a known brand, an old trusted site with references from sites like Wired.com, and 28,000+ inbound links. Even without getting the content that is still cheap. Some of my leading affiliates make that per year just from recommending my book, let alone all their other affiliate income from recommending other products and services.
If you had to try to build those same links Search Guild has in today's market they could cost $100,000 and/or a couple years to build. Even after the site changes ownership, most of the inbound links will stick too. In fact, if you move a domain to a new URL and ask people to redirect the links, over 90% of them will not.
Just like a strong domain name offers defensible traffic, so do those old links pointing at an old trusted site. If I had not already bought, ran, and closed Threadwatch, I might have considered buying SearchGuild, but I already have too many SEO sites to spend another $8,000 on a domain name that I would then have to develop. I hope they do something cool with it.
In the past many Google penalties were blatantly obvious. You either got traffic or you did not. But as time has passed penalties are getting blurrier, meaning your site can be penalized and still get traffic from Google. Some traffic reductions are due to competitive market forces, some are due to algorithm changes, some are due to automated filters, and some are due to penalties. If you are new to the market (and in some cases, even if you are experienced) it is hard to know which problems, if any, are holding back your ranking potential.
A friend just told me about how his Google traffic went way up after he spoke with a Google engineer, but he didn't want to talk about it publicly. I wonder how many other people are just like him, but don't speak about it or don't know they are penalized? And then I think back to the ban of the official AdSense blog, Brian Clark's PageRank hit, and Sugar Rae's ranking woes, and have come to the conclusion that spam fighting has become more of a shoot first and ask questions later game. They do not make a lot of mistakes, but when your site is just a number, it hurts pretty bad.
From a marketer perspective this shoot first shift is an important one which requires a few things of online publishers hoping to keep their businesses profitable:
Track your traffic using analytics tools, such that you know if/when something goes wrong, can prove it with hard stats, and can research it more specifically.
Publish at least 2 or 3 sites in different markets to give yourself additional data points on whether the issue is site specific or not.
Use public relations and viral link marketing where you once used link buys. If you are still renting links try to make them covert, and offset them with many natural links.
If possible package your offering as a service, so that you can justify charging recurring, and/or create an affiliate program. These make your income less reliant on search engines.
While on the link buying panel at WebmasterWorld's Pubcon a few people were pushing that you might need to consider how Google will view your current link buys 5 years down the road, and that they may hurt you then for what you do now. Upon hearing that I said something like "less than 5 years ago I bought spammy links and if I did not I probably wouldn't be speaking here right now". That got a cheer from the crowd. Who wants to be worried about what Google thinks or does 5 years from now? That is no way to innovate or take marketshare from current market leaders.
Reviewing Result Quality
When engineers view your site they don't just look at "if you have a few spammy links" but they try to consider the quality of user experience and the ratio of clean links to dirty links. If your site is good and ranks for years then you are going to get many natural links that dwarf any spammy links that were part of your site launch.
Building a Real Business
If your business model is entirely reliant on Google 5 years from now, your user experience is sub-par, and you haven't built up any brand equity after ranking for 5 years then there was not much effort put into building a legitimate business, and it deserves to fail. But the sites that rank get self reinforcing exposure. If SEO is part of your brand building and site building strategy you simply can not sit around waiting for the rankings to come in.
Inferior Sites Ranking #1
It is easy to lack objectivity when talking of the quality of your site, but in some fields I compete in, many of the top ranked competing sites are ran by people buying a slew of spammy links and pointing them at their (quite obviously) English second language sites. Because they rank, those sites get some number of self reinforcing links. If I did nothing but create great content they would still outrank my site. You have to buy marketshare in one way or another (public relations, AdWords, link buying) if you are trying to gain marketshare and your market is competitive.
Who Buys Links & Uses Push Marketing to Buy Marketshare?
That does not mean that I am an advocate of bad user experience or poor quality content, but if you care about SEO and have a new site in an old market, user experience and content quality are not enough unless you do some push marketing at launch.
AOL sent out millions of spammy CDs to market their service.
IAC buys a ton of links and aggressively cross links their sites.
Microsoft has got in trouble for launching new products by bundling them with their old products and steals traffic by sending seobook.om traffic to their live search product.
Monster.com owns a ton of thin lead generation sites.
eBay pays affiliates to spam Google.
One of Google's large ad distribution partners tried setting up a deal with me to rank their ads in Google's search results using aggressive black hat spammy techniques, in which I declined to participate in.
We Don't Write the Algorithms (or Hand Edit Search Results)
As an SEO you simply give the engines what they want. Looking at what they rank and how they market their sites gives you better insights for how to rank than blindly trusting the tips they give you to prevent you from ranking and suggesting you buy their ads. All of the web portals you know and love use push marketing to build their businesses. Why shouldn't you?
The Wikipedia ranks for a lot of competitive keywords because they are cited everywhere while acquiring the back links to be envious of. They also keep most of their link juice by linking to their internal pages and placing no-follows on external links.
I found out recently that they rank for competitive key phrases on Google such as:
Loan - #1 and 2
Mortgage - #3
Insurance - #4 and #5
The chart below is from the RankPulse list of top websites ranking in the top ten results for their 1,000 keywords sample database.
But when I added high traffic classifiers to the phrases above, Wikipedia’s rankings dropped significantly.
Insurance Quotes – Not found in top 1000 Google results
Mortgage Rates – Not found in top 1000 Google results
Loan Consolidation - #36
My explanations for the results are:
Although Wikipedia ranks well for competitive phrases, they don’t belong to the associated topical communities. They rank primarily on site authority.
While they have enough content to rank for said terms, they don't have pages targeting those terms. In many cases the relevant content for the phrase is compressed as part of a broader related page.
Their title tags target core keywords and lacks modifiers needed to rank well for popular terms that Wikipedia did not dedicate unique pages to.
By fixing the above issues, they may very well rank for the remaining 11 keywords.