Benchmarking Information Quality

Wikipedia ranks #2 for Aaron right now. They also rank for millions of other queries. They don't rank because their information is of great quality, they rank because everything else is so bad. About.com was once considered best of breed, but scaling content costs and profitability is hard. Google doesn't hate thin affiliate sites because they are bad. They only hate them because the same thing already exists elsewhere. Search engines try to benchmark information quality, and create a structure which encourages the creation and open sharing of higher quality content. When you see poor sites at the top of search results view it as a sign of opportunity. Realize that whatever ranks today is probably not what search engines want, but it is what is considered best giving the lack of structure to the web and how poor most websites are.

Cats and Mice: The Shifting Sea of Search Results

Google can never show the most relevant results for everything. No matter what algorithmic loopholes they close they inadvertently open up others. And anything they trust gets abused by marketers. Cat and Mouse.

  • Search engines trusted page titles and meta descriptions. Marketers stuff them full of keywords. So then search had to move more toward trusting page content. Marketers used hidden text and other similar techniques.

  • Search engines trust links. SEOs buy and sell them and create link farms. Search engines only allow some sites to vote, have some sites pass negative votes, make certain votes count more than others.
  • Search engines place weight on anchor text. SEOs abuse it, so they created filters for too much similar anchor text, and offset those by placing more trust on domain names when they exactly match the search query.
  • Search engines place weight on exact match domain names and domainers start developing nearly 100% automated websites.
  • Too many new sites are spammy so they place weight on older sites. SEOs buy old sites and add content to them.
  • Place more weight on global link authority. Spammers find cross site scripting exploits on .edu domains and media sites start posting lead generation advertisement forms on their sites.
  • Bloggers are too easy to get links from and comment links are easy to spam. Search engines introduce nofollow to stop comments from passing PageRank. Then Matt Cutts pushed nofollow to try to get webmasters to use it on advertisements.
  • Too many people are created automated sites, especially affiliates are creating a large number automated sites. Search engines employ human reviewers, get better at duplicate content detection, and require a minimum link authority on a per page level to keep deep pages indexed.
  • Social news sites are providing a sea of easy link opportunities and low quality information. Too many people are doing linkbait. Perhaps Google may eventually only count so many citations in a given amount of time.

When your site changes in rankings it may not be just because of changes you made or changes in your field, it may also be due to Google trying to balance

  • old sites vs new sites

  • old links vs new links
  • big sites vs small sites
  • fresh content vs well linked pages
  • exact match vs rough phrase match
  • etc

Domain Names Will Become Less Relevant (to Search Engines)

Currently domain names carry significant weight in Google's relevancy algorithms if they match the search query, but that is a signal destined to lose value. As more people get into automated and cheap content to turn park pages into automated low cost community driven sites, search engines are going to learn that it doesn't make sense to give a matching domain default status in a category. Frank Schilling, a well known domainer, recently blogged about how to unseat Google, and is using wiki content to seed community driven sites on generic domains. Two ways Google is fighting off the automated content generation that domainers will explore is through duplicate content detection and minimum PageRank requirements on a per URL basis.

Domain names are trusted because they are an expensive commodity largely controlled by a few people, and most of those people are not developing the names. Once those people start turning parking pages into content sites I think search engines are going to need to look for other relevancy signals. Many VCs and domainers have been looking at automated or low cost development ever since Paul Sloan's 2005 article Masters of their Domains, but a search engine full of sites like eHow, WeHow, and WikiHow is not a search engine that is useful to searchers.

How Global Link Authority Will Lose Relevancy to Subscription Data

Search engines, especially Google, place a lot of trust on Global domain authority when computing their relevancy scores. If a page is on a trusted site it automatically gets a boost in relevancy scores. Premium content providers know that Google needs them so they take advantage of their market position. Lets look at a few examples of authoritative leverage in action:

How will Google minimize the media's ability to sell advertisements that rank in Google as though they had editorial value? They are going to have to start breaking sites and search results into pieces. If they do, it will not only minimize the downside potential of advertisements on popular sites, but that would also kill the .edu spam.

Personalization will lower the profitability of leveraging off topic broad nationwide ads by lowering their exposure by biasing search results to locally trusted sources. Matt Cutts has talked about personalization recently, not once but twice.

Given that some domainers control millions of daily page views, and there are plenty of cheap traffic sources I don't think raw usage data will be that strong of a signal of quality. Like links, usage data is easy to manipulate.

As the web gets polluted with low cost links, other low cost marketing opportunities, and recycled junk the key to Google's success is tracking attention data. They have the most popular feed reader, and their new homepage allows you to create topical tabs that show related feeds.

In tracking how people consume and share information they can understand local topical networks, while reinforcing the views of the local topical editors. It sorta creates a bunch of micro-Digglike communities, but more free-flowing, with less focus on free votes. Time is the one thing we are all limited on, and if you are paying attention to something then there is a lot of value in trusting that attention.

Balancing Website Investments

Short term you can do well with major holes in a website, but what do you need to have a successful long-term website? I think every site should have at least all of the following criteria:

  • credible sounding memorable domain name

  • credible trustworthy site design
  • credible content that people pay attention to
  • link equity & marketing strategy
  • profit strategy

Lets go through those one at a time:

Domain Name: Can you do well with a bad domain name? Of course you can short term, but long-term if your business name is hard to remember you are going to lose market share to people who have a more memorable domain name. I am not suggesting that everyone runs out and spends many thousands on their first domain name, but if your domain name is hard to remember, or people will not accept donations from your site because your domain name looks spammy, it is time for a change.

Credible Site Design: The logo on SEOBook.com cost $100 and the design on SEOBook.com cost $1,500. Both are probably worth at least 10 times that much. I started off with a default design and then bought the $100 logo, then eventually bought a nice site design. In many markets the market leader is determined by people outside the topical community. A beautiful site design makes it easy to vote for you and trust you.

Credible Content: If you are afraid of a manual review of your site content by a search engine then your site is not a sustainable site that will stay relevant. If nobody talks about your site you are losing marketshare.

Link Equity & Marketing Strategy: Editorial links come from conversation, which is often driven by marketing. What makes your site remarkable? If you are mainly ranking due to site age and old links, what can you do to keep your site relevant to the current market? Are the linkarati interested in your site?

Profit Strategy: Some sites are directly monetized through advertising, services, or product sales. Some sites are tools to build credibility and indirect revenue streams. Some sites have other nepotistic values such as a clean link source or a connection to the media.

Every Business Has Holes:
Many of the best sites start off as a hobby and move into biz model. Every market worth being in changes quickly, and every business person starts off with an imperfect business model, and then uses market feedback to move toward being more efficient. Google originally was designed as a project to measure the most authoritative links on the web and then they transitioned into being a search engine.

Filling in the Gaps:
Missing any of these pieces is just like monetizing via off topic pop ups, or other strategies that have moved into irrelevance. Google made the link an ad unit because people won't be able to ignore it.

A site can be successful with only one piece of the value chain done well, but if you fill in the gaps in the other parts of your site profile the profits will be much greater. If one is spread too thin working on too many projects it is easy to lose marketshare, but if you work hard on a few sites you are passionate about and listen to market feedback it is easy to quickly become a market leader.

What other things did I miss? What do you think is an important element in successful sustainable websites?

Large Site Retail Competitive Research: Letting Amazon.com and eBay.com Work for You

Selling your stuff on either of those networks might be a good way to leverage their domain related trust to rank in the search results, but you can also use both of them to research what is worth selling or writing about, even if you do not sell on their networks. Since their research data is conversion oriented, in many verticals it is worth more than keywords from other sources. eBay offers a Marketplace research service that costs $2 to $25 a month.

Amazon shows which products are the best for last year, current top selling products, movers and shakers, and hot new releases.

In addition to giving away this free conversion information, Amazon lets you see what customers think of the product via review, allows you to bid on keywords related to products via ClickRiver. If you sell your stuff on Amazon.com, you can buy a better together listing with a popular product.

If you are in the info-product market Clickbank's Marketplace is a goldmine.

Once you find interesting products you can refine your keyword research using general keyword tools, then further refine that data by using a tool like KeyCompete to see what keywords competitors are bidding on.

Blog Comment Spam to Search Spam to Email Spam

A friend of mine found out about a product that was first seen by them from a spam comment someone left on my site. The product helped automate content generation and syndication, aiming to help new webmasters distribute slightly unique spam across the web. My friend gave that person his email address and they email spam him non-stop. Marketing offers are no better than the marketing method they use to find you.

Extending the Reach / Circulation of a Web Based Content Site & Ad Network

Once you have a trusted market position advertisers will buy in even if you hurt their business model or call their products crap:

To my enormous surprise, the company that threatened to call the feds on me made another offer to advertise on CamcorderInfo.com. Because of our popularity - we have the No. 1 search listing on Google (Charts) under "camcorder reviews" - it couldn't afford to avoid us anymore. I okayed the deal, comfortable that the electronics firm was aware of our ethics policies. It has since become one of our largest clients. When I asked the executive if he ever actually called the FBI on me, he insists he did. He said the agency found me guilty of nothing except a passion to build a reliable company.

Every Ad Network is Flawed:

Once an ad network has buy in they look for ways to leverage that market position. And there is fraud involved nearly every time

  • newspapers puff up their circulation numbers

  • magazines run ads and give themselves away to boost circulation
  • a keyword that normally costs me $4 a month on Yahoo! cost me over $100 this month due to syndicated click fraud
  • look at how much of the web spam Google funds

How to Extend Your Reach Without Getting Too Dirty:

Once you establish a trusted market leading niche site you can extend your circulation by buying traffic from cheap sources. Some people will just send fake traffic generators over their own site for impression fraud, but there are many cheap sources of traffic that may provide ongoing value and look much more natural than blatantly committing circulation fraud.

  • Keep your feed rather clean so it is easy to subscribe to. The people subscribing to it will likely be the people who link to your site. Maybe put one exclusive sponsored by ad unit in your feed.

  • Create an email newsletter that recycles your best online content. Maybe run one or two ads in it, and use it to drive traffic back to your site.
  • Buy interstitial ads from AdBrite on related websites. If AdBrite is underpricing the inventory of a strong site try to buy out all their ad slots.
  • If AdSense is under-priced in your market consider buying AdSense ads on related sites to drive longtail traffic back to your site and further position yourself as the market leader. You see this sort of activity all the time with web directories, where a couple of them have ads on other web directories.
  • If you are really motivated you may want to resell ad inventory of other smaller players in your field.
  • Buy traffic from StumbleUpon.
  • If you write what is essentially a news site try to get syndicated in Google News and other related news sites.
  • If your site has a high authority score, write a few posts here and there that use analogies or other ways to pull in related traffic without looking like you are straying too far off topic. For example, one of the leading search queries for this site is Google Auctions.

  • Interview industry experts that have greater reach than you do.
  • If you have been ignoring SEO at least give the topic a cursory look, and consider buying a few trusted links by doing things like joining trade organizations.
  • Create linkbait and get it exposure on the popular social news websites. Also consider promoting your best content via Review Me buys on popular channels. (disclaimer: I have an equity stake in Review Me)
  • If you have enough time to manage it (or know someone else who can) consider adding a forum or other user generated content sections to your site. Active forums can have hundreds of thousands or millions of page views each day.
  • Have your pages refresh once every few minutes to rotate through advertisements. I believe MarketingVox does this.
  • Make sure your site design is clean, you publish quality original content, and your about us page looks trustworthy. These will make it easier for people to buy ads directly from you, and will make it easier for the media to use you as a source.

If you have a high value market position in a large competitive sector and are not making much from AdSense make your site brand ad friendly by pumping up your circulation numbers. Many of those ad buys are cheap on a CPM basis and will lead to additional recurring traffic and links. In many high value markets as your get more exposure your ad revenues increase logarithmically because you get more traffic and people are willing to pay a higher CPM rate.

Writing for Multiple Audiences

Many sites tend to steer clear of controversy, but want the traffic controversy can bring. What are the solutions?

  • Accidentally put things in / leave things out of your story that make it easy to take it the wrong way. Leave an easy angle that you know people will take so that when they do you can crush them. Depending on your strategy, it may even help to leave comments off on some of these types of posts.

  • Write your own controversial comment as though it was from someone else and then let people debunk that person.
  • When you syndicate your story to a social site that is politically biased, place blame for whatever is wrong on someone. People will vote on hating or liking that person without even clicking through to the story to see if it relates to the headline.

Even the Search Engines are Bad at Optimization

Yahoo! allows paid inclusion members to buy quick links in the search results and indexes sites more fully if they use Yahoo! for their site search. While neither of these may seem like a big deal they both are. Allowing businesses the chance to buy a greater share of the search results or giving them greater indexing priority for them using your search service are both things that would make search purists cringe.

While Google is busy turning the text link into a performance based ad network Yahoo! sells their search credibility and marketshare too cheaply. Buying Del.icio.us and MyBlogLog are not going to make up for the undermining effect on credibility paid inclusion offers.

Pages