Jim Boykin recently claimed to have kicked the paid links habit:

So, the best thing for my company to do, if we want to stay out of the fire, is to make sure that We Build Pages adheres to the Google Guidelines, and that means we won’t be getting any more paid links for manipulating search engines.

When I first got on the SEO scene and quickly started buying links, one of the sites I kept running into was WeBuildPages. One of my friends jokingly called me "the original link spammer" but Jim Boykin started buying links before I did and was doing it with more scale than I did. To see Jim dismiss link buying outright seems like it is either over-reaction or link buying is nearing its death.

Is Link Buying Nearing its Death?

When search click distributions may end up similar to the below graph how can one not want to push the limits?

For some keywords (and some entire business models) one or two rankings difference can be the difference between a profitable business model and a money loser. Yes real businesses should not be so reliant on Google that Google can chose to kill them, but there are a lot more people doing business with me too offers than there are creative and original people offering significant value added services from a unique approach.

Most business models are arbitrage, and Google wants to claw away as much of the easy value as they can, forcing you to spend on brand building.

The Cost of Branding

Most traditional businesses are lucky to have a 10% or 20% profit margin. When one company controls 70% of the search market (closer to 90% in some niches and some geographic regions) it is easy for them to exert enough influence on a business (through quality scores, hand edits, threats) to move it from having 10% profit margins to losing money.

Many regional offline brands are dying because their cost structure does not work on a network of infinite competition.

Many online brands are money losers or break even at best, with some losing hundreds of millions of dollars before coming profitable. Some of the more savvy online companies (like Monster.com, Expedia, and BankRate) may break even on the brand and leverage the brand to build out profitable networks of thin websites that allow them to double or triple dip in the organic search results.

Death Grip Growing Stronger

Google's death grip on the web is only growing stronger. While the web and search are making some bulky business models (like that of the NYT) irrelevant, in response the New York Times publishes articles about how Google Seduces With Utility:

“The most powerful form of advertising is to be exceptional,” said Ranjit Mathoda, an investor and technologist who blogs at Mathoda.com. “Google has created an ecosystem that perpetuates itself by being useful.”
...
“We do have a philosophy that our products should speak for themselves. We tend not to make a lot of noise,” said Jeff Huber, senior vice president for engineering at Google.

Google is the front door to the web. And while Google is getting credited for "not making noise" and "being exceptional" they use their ad platform to give themselves free distribution in any vertical they want to compete.

Part of Google rising to such dominance was their aggressive bundling of their toolbar on computers through deals with OEMs and other software companies. Now that Google has a browser they want to take it one step further by doing Chrome distribution deals:

Sundar Pichai, Google Vice President, Product Management, revealed that Chrome will be ready to come out of “beta” testing by January, and that the search giant was looking at ways to make Chrome the browser of choice for the everyday user.

“We will probably do distribution deals,” he said, adding, “we could work with an OEM (Original Equipment Manufacturer) and have them ship computers with Chrome pre-installed.”

Chrome replaces the address bar with a search box. More search volume for Google.

Do You Still Buy Links? Do Your Friends?

Knowing how good Google is at marketing and that they are still gaining marketshare, do you still buy links? How has your link building and link buying strategy changed over the past year or two?

Changes in Search

In recent years personalization, localization, universal search, search suggestion, and specialized algorithms like query deserves freshness have altered the landscape of search. But even outside of these add-ons, Google's core relevancy algorithms are (at least to some degree) query dependent.

Competitive Keywords

When there are many matching search results for a given search query, Google places a lot of weight on core domain age & authority and on external signals of quality like link quality, link diversity, link anchor text and perhaps other signals of quality like usage data and a LocalRank boost. For competitive queries where there are many matches on page optimization is not given as much weight.

Long Tail Low Competition Keywords

For search relevancy algorithms where there are fewer matches and fewer external signals of quality available, Google must put more weight on the content of individual pages. Where there is no community to rely upon Google must trust publishers. And while each longtail ranking might have little value the nickels and quarters add up. Their limited search volume and value leads many competitors to skip over them as they do not appear in most keyword research tools.

In a recent blog post the Google AdWords team asked "Did you know that 20% of the queries Google receives each day are ones we haven’t seen in at least 90 days, if at all?"

The same post highlighted that "broad match currently accounts for over 1/3 of all clicks and conversions for advertisers, worldwide" and that Google "recently improved the search query report to provide more granular detail on which queries are triggering ads for your broad match keywords."

A Comparison

This graphic makes no attempt to be 100% correct for any given query, but was made to show an illustrative difference between competitive keywords and non-competitive keywords.

If you are starting a new site and have built little to no offsite signals of quality you can expect to rank for longtail phrases first. As your site builds authority you can compete for some of the head keywords.

As part of their 10th birthday celebrations, Google recently released a 2001 index, to show us how much things have changed.

It is fascinating to look into the past, especially from an SEO point of view. Has the nature of spam changed since 2001? How has Google changed in order to nullify the affects of spam?

When Google filed their registration statement prior to IPO, Google identified a number of risk factors.

One of these risks was:

We are susceptible to index spammers who could harm the integrity of our web search results

There is an ongoing and increasing effort by “index spammers” to develop ways to manipulate our web search results. For example, because our web search technology ranks a web page’s relevance based in part on the importance of the web sites that link to it, people have attempted to link a group of web sites together to manipulate web search results. We take this problem very seriously because providing relevant information to users is critical to our success. If our efforts to combat these and other types of index spamming are unsuccessful, our reputation for delivering relevant information could be diminished. This could result in a decline in user traffic, which would damage our business."

Curious how Google conflates spamming with relevance, eh. While it could be true that manipulating rank could lead to lower relevance, that isn't a given. The manipulation could, after all, produce relevant results. "Relevant" being a subjective judgement made by the user.

I digress...

What Google are really getting at is the type of manipulation that leads to less relevant results, commonly referred to as search engine spam. In this respect, what has changed since 2001?

Has Search Spam Been Defeated?

Or, to put another way, what changes have Google made to reduce the business risk of non-relevant search results?

Compare the following examples with the results we see today:

Buy Viagra
Viagra

Now try searching on those two phrases in today's index. How many differences can you spot? How have the result sets changes? Are they less "spammy"?

Here are a few aspects I noticed:

1. The search results are much tighter and much more well policed. You wouldn't find the penis-envy.com site's link exchange page ranking in Google's 2008 search results for Paxil search queries.
2. Google used to match keyword strings a lot more than it does today. This is the reason why a lot of on-page optimization techniques have become redundant, and the reason why effective on page optimization in 2008 is more about diversity than repeating words.
3. Blogs have came from an obscure force to category leaders in many markets.
4. If you happen to be searching outside the US, Google now incorporates, and boosts, regional results.
5. Google now incorporates YouTube, news, and other related informational sources, thus forcing results from smaller sites further down the page
6. There used to be a lot more hyphenated domain names showing up top ten. Not so much these days.
7. Wikipedia, then called Nupedia, had only just started in 2001, so wasn't yet appearing in every single search result ;)

When Google first emerged, algorithmic search was in real danger of becoming unusable. Engines like Alta Vista were losing the war against spammers, and result sets were becoming increasingly irrelevant. Sergey Brin once declared that it wasn't possible to spam Google. When Google came along, they had defeated spam forever using a clever link analysis algorithm. No more spam!

Well, not really.

Spam hasn't gone away. But it is fair to say that Google is doing a pretty good job of maintaining relevance, and in many cases, eliminating the worst forms of spam. For example, it is now uncommon to see the type of deceptive redirects that were common in 1997, whereby if you clicked on a link, you were led you to a site that was unrelated to the link text.

We've seen the rise of the authoritative domain, and the relegation of the influence of many smaller sites. Pages hosted on authoritative domains are more likely to rank higher than pages on sites that haven't established authority. This has, in turn, led to a different type of spam. People hack into authoritative sites in order to place their links, or entire pages, on these domains. Wikipedia has an ongoing battle to keep their pages free from "commercial imperatives".

The target has, in many ways, shifted down a level.

Big Changes

Since 2001, Google has incorporated verticals.

In this article, Danny Sullivan outlined the use of "invisible tags" in the delivery of search results.

"The solution I see coming is something I call "invisible tabs." Quietly, behind the scenes, search engines will automatically push the correct tab for your query and retrieve specialized search results. This should ultimately prove an improvement over the situation now, where you're handed 10 or 20 matching web pages."

Result sets have increasingly become query dependent, as if you'd pre-selected a topic tab. For example, if your query is determined to have an informational intent, you're unlikely to receive a commercially oriented result set. It is has become a lot more difficult to get off-topic listings - which in this specific case would be commercial pages - into such result sets.

We've also seen the structure of search results pages change markedly. We see images, videos, news, related searches, sub pages, onebox results boxes, personalized results, desktop results, and Adwords. This leaves less and less room for other types of pages, as the search results orient more heavily around a wider variety of data types.

However, in the end, the SERP is still just a list, that looks much like the old list. What will search, and search spam, look like in another tens years?

The Future

Over $10 billion dollars are chasing paid search each year, and that figure will surely grow as media spend increasingly shifts online. There is still a strong incentive to use all means necessary to get to the top of the list.

Google will, of course, continue to try and counter this threat to their business model. The PageRank has likely been changed considerably to when it was first published. Google is likely to continue to incorporate usage metrics, making it more and more difficult for less relevant pages to gain a foothold.

On the flip side, will search be important as it is now? There appears to be a trend for more information to be pushed our way, rather than going out and finding it ourselves. RSS, recommendation engines (Amazon, YouTube, et al), community models (Facebook), and more. Will our surfing habits be (voluntarily) monitored, and answers provided before we we're even aware of the question? We're already seeing the early stages of this with contextual Adwords in Gmail. These changes will, in turn, give rise to a new breed of spam. While the commercial incentive remains, there will always be a level of spam.

The game of cat and mouse continues...

The Google 2001 Search Index is a Great SEO Tool

Having a glimpse of the past reminds us of how things changes, which might help us think of why they changed and how they may change going forward.

The 2001 index provides for a great tool to show past popular SEO techniques that have become irrelevant, which is useful when the boss uncovers an old spammy strategy that they feel you must follow to succeed. It not only helps us inform employers, but also allows us to talk about and highlight overt forms of spam without the worry of "outing" a page that is currently ranking.

Google recently expanded their ad offering by inserting AdSense ads on maps, putting AdSense image ads & banners on image search results, opening up AdSense for Games, and monetizing Youtube with affiliate ads for Amazon.com and Apple iTunes.

The NYT article on AdSense for Games (linked above) promises a couple more new ad units in the coming weeks, and highlights Google's new ad strategy

For the text and graphic ads (but not video) Google will also look at the context of the game and the page it is on for clues that might indicate whether some of the ads targeted by keyword are appropriate.

Mr. Oestlien indicated one small feature of Google’s program that may represent a significant change in the company’s approach: It is starting to broker deals between game publishers and advertisers to have their products integrated into the actual play of the games. For example, a dog food company could have its latest kibble built into Pet Society, a game on Facebook that now has Google ads.

On the high end for brand advertisers Google is becoming something that looks, smells, walks, and talks like an agency. Take a look at this ad unit.

And on the lead and retail front, Google is looking to become the web's largest affiliate. Everyone in search marketing (and online media) need to take a strong look at the merchant beta test Google conducted

How long until Google goes after other online ad markets that are worth hundreds of millions or billions each? More and more Google searches may end up clicking through to a Google property or a Google navigational aid. If Google can get enough merchants to buy in, any (or all) of these could become affiliate links. If the data can be structured Google can take their tax.

AdWords effectively killed the longtail by recycle brand ads on longtail search queries. Look for that consolidation to continue. If the SERPs hold custom ad units by Google, is your lead value and brand big enough to be able to pay for the leads? If not, how can you deepen your experience to create a citation-worthy service that goes deeper than Google is willing to go?

Update: As John Andrews highlighted, Google aggressively cashes in on branding, so if you own a brand you owe it to them to be liberal with their guidelines.

I was just finishing up our guide to how to optimize for search suggestion, and noticed something worth discussing.

I am not sure if safe harbor covers companies that index content, cache/host content, and suggest searches for downloading pirated works...but if it does, I think the law needs changed. It seems Google could have thought about the torrent related keyword suggestions before launching search suggest as a default.

Part of the reason why I had to change my business model was the need for a more interactive higher value service, but another big part of it was also that I saw this sort of activity coming. It is too hard to create valuable information and sell it in a digital format unless it is broken up into pieces, is time sensitive, and/or has interactive elements added to it.

If you think Google respects copyright you are wrong. All content wants to be free, and, preferably hosted by Google, wrapped in AdSense.