Catch Me Sleeping

I logged into manage one of my Google Apps accounts today, and was greeted with a message about an upcoming upgrade. Soon, the members of my Google Apps domain will be able to access the whole realm of Google services. The included a link to learn more about this upgrade, so, naturally curious, I did.

The page mentions a service called Confucius, offering to “Ask questions and get answers from real people“. That sounds quite a bit like old Google Answers, which at this time is no longer accepting submissions.

When I search Google for Google Confucius, I don’t see any obvious link to a Google service there. In fact, the first link is to a link in the Google Directory, to a page about the Chinese philosopher Confucius.

Anyone know anything about this mystery?

Question and answer sites are as popular as ever online now. StackOverflow rules the world of programming question and answers. Yahoo Answers is still surprisingly popular, even though the average quality of questions AND answers is pretty mediocre.

Generic domain names, also known as exact match domain names, are simply domain names that describe every day ideas or things. Examples of generic domain names would be:

cars.com
fishing.com
rubberboat.com
coffeemug.com
laptopcomputer.com

Notice none of these domain names refer to a specific brand name? For instance, here are some examples of domains that are NOT generic:

dietcoke.com
fordcars.com
dellcomputers.com
shimanofishingreel.com
intel.com

These are all branded domain names, and shouldn’t be confused with generic domain names. As tempting as it might be to register these kinds of branded domain names, you shouldn’t, unless you are the actually hold the trademark for the brand name! Not only could these branded domain names be taken away from you. You could also be fined for holding these kinds of names.

Right, but generic domain names don’t have any of these scary problems. You are free to register domains like babystrollers.com, airtravel.com, and even alligators.com, if you want. Of course, it doesn’t have to be .com either. You can register generic .NET, .ORG, .INFO, .IN, etc. Along these lines, creditcards.info, fishingreel.net, and exhaustpipe.org are all good examples.

So, why are generic domain names valuable or interesting? The main reason is that there is only a single unique instance of every generic domain name out there. Once it’s registered and held in someone elses domain portfolio, it’s impossible to legally take it away from that person. Unless the domain holder lets the registration lapse, or negotiates a sale of the domain, they can hold that domain for eternity. This explains the first reason why generic domain names are valuable – because of their scarcity. Doesn’t that seem ironic, that something generic can actually be scarce?

Of course, not everything that is scarce is valuable. Any domain name is scarce, even an obviously garbage domain name, that holds no meaning. The second reason that generic domain names are valuable, are because they hold meaning. A person interested in alligators, could surely think of no better place to put his alligators website than at alligators.com. It would be easy for fellow alligator lovers to remember, and type in each time they wanted to return to research more information about alligators. The same would go for any other generic word. Where is the most logical place to expect a great website about baseball cards? Surely baseballcards.com! Looking for a new refrigerator? Why not look at refrigerators.com?

Domain names are basically addresses. Just as a street name, city, state, and zip code are enough to uniquely identify just about any populated location in the United States – a domain name is a way to uniquely identify a location on the World Wide Web. A good generic domain name would be the equivalent of an easy to remember address for your house or business. Just because it’s so easy to remember, more friends would naturally just come by to visit, or more customers would just naturally stop over to visit your store.

Now, as with any scarce, and valuable thing in the world today, generic domain names have rapidly been bought and collected. Just as you wouldn’t expect to find a giant gold nugget sitting in the middle of the street, you shouldn’t expect good generic domain names to be sitting around for you to pick up. In fact, almost any obvious generic domain name that you could think of has already been registered. It would be a shock if it wasn’t that way, just the same as finding a 100 dollar bill on the sidewalk.

In the world of generic domain names, there is a hierarchy of top level domains (TLDs). First come .COM, the come .NET, and .ORG. Then comes everything else. Since most websites use a .COM, most people naturally assume the website they are looking for will be a .COM. Going along with the addresses analogy, it’s just an easier to remember address.

Some of these domains that have been registered have actually been developed. Go checkout cars.com and perfume.com. They’ve both been developed into useful and interesting places. Many other domains are just held by domain traders and speculators. alligators.com and baseballcards.com are two examples of such domains. Their owners long ago realized the potential value of these domains, and snapped them up.

In fact, sadly, most good generic domain names are just being parked, like alligators.com and baseballcards.com. There are a few reasons for this.

The first reason, is that we’re still really in the infancy stages of the Internet era. The internet and domain registration has only be available to most people for about 15 years or so. With the amazing variety of millions of words and word combinations that could be registered as meaningful domain names, it’s no wonder that only a tiny fraction of those have been developed in any kind of meaningful way.

The second reason that so many domains are just being parked, is that it’s quite cheap to register domain names. It costs just $10 to register a domain name for a year. In proportion to the potential value of a good name, that’s very cheap.

Finally, domainers are able to earn a good income by holding these domains. They earn a stream of income by putting ads on their parked domain pages, and another income from selling off some of the domains that they own. The income that domainers earn from these two sources is often just re-invested back into buying up more domains, further contributing to the scarcity of good generic domain names.

So, if you’re interested in starting to build your own generic domain portfolio, where do you start? Are there any good generic domain names even still available?

It is still possible to register good generic domain names. There are four techniques that are commonly used by domainers now, to continue building up their portfolio of domains.

The first, is to just hand register domains. This is the most primitive technique of all, but it still works. This is something like panning for gold in the American River in California. If you know where to look, and you stick with it, eventually you’re going to find something valuable. For hand registering domain names, it pays to stay abreast of current trends. Fifteen years ago, no one would have ever through to register a domain name gulfoilspill.com, because the event it refers to hadn’t even happened yet. By watching the news, and considering domain names related to new and interesting products and events, you might be able to pick up some valuable domain names for just the $10 cost of registering a domain.

Of course, almost no one does panning for gold anymore. To strike it rich as a gold miner, you need some more powerful tools to really search out the source of the gold. Likewise, in your search for good generic domains, you might consider using a tool help you. One such tool is Generic Domain Name Finder. Input a keyword, and it will search up to 800 similar and related keyword variations, and tell you if the .COM, .NET, and .ORG are available to be registered. A tool like this lets you be as much as 100 times more productive as a domainer.

Another way of picking up good generic domain names, is by catching domain names after the previous owner lets the registration lapse. There are some good deals to be found this way. Every day thousands of domains drop. However, it’s also quite competitive. Often, you’ll be up with many other domainers equipped with specialized tools and priority access to domain drops. But now and then, good domains can still be registered this way.

The very best generic domain names might only be available for purchase, either through an auction or through negotiation with the domain holder. Expect to pay $2,000 or more for two, three, and four word .COM combination like saltwaterfishingreel.com, and potentially several hundred thousand dollars for single word domains like alligator.com. There are several good marketplaces where you can browse available generic domains, such as domainsponsor, sedo, and domainnamesales.com

The world of generic domain names is really exciting and interesting. It really feels counter-intuitive that something generic can be valuable, but in fact, generic domain names represent some of the most valuable properties on the World Wide Web today.

Suppose you want to see which pages on your blog are the most related to each other? You could see which pages you posted into the same category, or gave the same tag too. But that’s just too old fashioned. Surely there’s a better way, right? Indeed! As a geeky programmer, you’ll most likely want to apply a statistical measure known as tf-idf to determine the most important terms in each post, and then compare the results with every other post on your blog to find any overlaps.

The td-idf weighs words by their inverse frequency. So the more common a word is (think words like A, THE, It, etc), the less weight it will have for a potential document.

Here’s how it works:
Step 1 – calculate term frequency of a word in a particular document, and divide it by the number of total words in that document. Let’s use the term forums for this example calculation.

In my previous post How to do link building, the word forums showed up 2 times, in 725 total words. The TF value = 0.002759

Step 2 – calculate the inverse document frequency (IDF). This is done by first dividing the total number of documents, by the number of documents that contain the actual keyword in question. Then taking the logarithm of the result.

For the sake of simplicity, let’s assume that none of the other 7 blog posts, at the time of posting this article, contained the word forums. The IDF value is therefore, log (7/1), or 0.8451.

Step 3 – multiply the TF by the IDF, to get the result 0.002759 * 0.8451 = 0.0023312

To put that into perspective, let’s calculate TF-IDF for the word like. Since it’s more generic, commonly used word in the English language, we expect that the term weight for like should be much lower.

I counted 4 instances of the word like in the link building blog post. Also, 4 of the 7 blog posts contains the word like. That gives us the following calculations:

4/725 = 0.005517

log (7/4) = 0.2430

0.005517 * 0.2430 = 0.001341

The result, that the weight for the term forums (0.0023312), is higher than the weight for the term like (0.001341), is what we expected, given that the word forums comes up less often in the English language, than the term like. This tells us that the word forums is more significant to this posting about link building, than the word like is – even though like was used twice as often (4 times vs 2).

I’ve seen a few interesting postings lately on other blogs, about tf-idf, which made me interested to understand it better.

• Latent Dirichlet Allocation (LDA) and Google’s Rankings
• tf-idf + twitter = $$$

In a practical sense, calculating tf-idf is definitely a task best left to computers. For instance, if I would want to create a WordPress plugin to accurately use tf-idf values to search out related posts in my blog, it would need to recalculate the tf-idf weight for every keyword in every blog posting each time a new article was published. As useful as the end result would be, I could just imagine the hordes of tired, overworked web servers across the internet, groaning each time a spry young blogger slapped out a silly new post, and his corpus of blog posts was enlarged.

Also, as a final practical consideration, the Wikipedia page on this topic recommends normalizing the term frequency count, to prevent a bias towards longer documents, which might have terms repeated more often, just due to their length.

A lot of SEOs are content to get a top 10 ranking. They create great content, optimize their page, and then go out do link building. Over time, their page rises up. Eventually they reach the first page, and start getting traffic for their keyword. They succeeded, right? They got onto the first page, and already are getting traffic, right? In fact, that’s just like giving up right before you reached the finish line! The real rewards of your hard work will come when you reach the first position.

A few years ago, some data was leaked out of AOL about actual click distribution on their search result pages. Have a look at the data:

SERP position click through rates

Surprising, right? The first position will receive more than 3 times the number of clicks than the second position. More than 9 times the number of clicks than the 5th position, and 15 times more clicks than the 9th position.

And, in fact, I think the numbers may even be MORE skewed than these AOL numbers. That’s because of something known as Google Universal Search. Google is often inserting other kinds of contents into the search results – usually somewhere between the second and fourth positions of the search results. These are image search results, real time search results, new search results, product search results, and video search results. If you’re search result is below one of these kinds of blocks, you can assume that you’re click through rate is going to be even lower than expected.

With that said, it would be smart to look through where you rank for your keywords. With just a bit more effort, you may be able to push your lower first page rankings up to the top spot, and bring in loads more traffic.

Sad but true – Google Wave will be discontinued by the end of this year.

When it first came out, there was a media frenzy, as there tends to be when Google releases a new product. I basically ignored. Eventually, someone I collaborate with on projects sent me an invite, and told me that we would be using it. I signed up…. and basically used it reluctantly.

Now, knowing that it’s going to go away soon, I do feel some regret.

The reason that Google Wave actually ended up being useful to me, was that it was an easy way to share a thread with a selected group of people. That’s it.

All of these other features, like drag and drop, and tracking every keystroke entered, were interesting. But the main thing that became useful to me were these threads.

I really have no idea where else I can get this same threads feature. Perhaps Google will release some new tool that takes all the best of Wave, and makes it even better. Even though there are so many existing project management tools that there already are – I think there place in the market for some new innovation.

Being interested in personal productivity, as I am, I’ve come to the conclusion that multitasking at a computer really does not work. What I mean is that, when you allow yourself to constantly be distracted, constantly switch tasks, before completing any one of them. you’ll end up getting nothing done at all!

I find this is especially the case in front of a computer. In front of a computer, it’s just so easy to switch tasks. You can read an article, then send some IMs, then stream a youtube video, then send some emails, all within the spam of 5 minutes or less.

Out of pure exasperation, I decided to do a bit of research on this very topic. What I found were several articles that confirmed what I myself felt. Here are some quote from one interesting article I found:

Multitasking Decreases Productivity

Caution all you multitaskers, you may not be as productive as you think. Researchers continue to find that multitasking decreases productivity, increases stress, and may cause physical discomforts such as stomach aches or headaches. In a recent study by Eric Horvitz and the University of Illinois, a group of Microsoft workers took, on average, 15 minutes to return to serious mental tasks, like writing reports or computer code, after responding to incoming e-mail or instant messages. They often strayed off to reply to other messages or browse news, sports or entertainment web sites.

Also, NPR interviewed an academic researcher about his findings related to multitasking. You can listen to the thirty minute recording here:
NPR Story about Multitasking and Productivity

They found that multitasking was especially common in front of a computer. They also found that, the people who thought they were the best at multitasking were in fact really the worst!

They don’t really explain why people get trapped into multitasking behavior. But they do say that there might be some correlation based on personality types. Some people are exploratory, while others are exploitative. Exploratory people want to keep searching for new information. They feel some psychological reward from gaining more and more information. Exploitative people want to act on the information first, rather than continue to seek more and more information.

They also mention, one of the biggest problems of the worst multitaskers, is to filter the really important information from the information that is really less important.

Personally, I find that it’s really important to be disciplined when I sit down at my laptop to do some work. For one, I really need to have a plan of the specific things that I really want to accomplish. Without a plan, it’s easy to drift from one thing to another. Then, at the end of the day, there’s no real progress on the projects that really matter!

If you search online, or read forums, you’ll see lots of techniques for how to do link building. Everyone likes to talk about exciting things like link baiting, social media, article directories, and even plain buying links. Those techniques all can work. But it surprises me that so many people don’t do the most basic link building technique. I see it discussed so infrequently, that I wonder if it’s supposed to be a hushed secret that no one talks about.

Actually, the reason that this technique is not talked about so often, is that it doesn’t have a simple and catchy name. It’s a just a systematic process, which needs to be followed.

Step 1. Do the search that you want to rank your target site for.
Step 2. For each result in the SERP, run the google link: operator for that domain.
Step 3. Look at every one of the results for the link: operator search. Ask yourself, can I get a link on that page?

There are a few reasons why this is such a valuable process. For one, it really opens your eyes to all the different places on the web where you can get links from. When you run the link: operator search, you’ll see parts of the web that you didn’t even know existed! You’ll come across interesting old forums, directories, news sites, organization sites, etc.

Another reason this is valuable, is that you’ll get a feel for which links Google feels are more valuable. It’s common knowledge that Google withholds results from the link: operator search results. A website might have several thousand backlinks, but Google might only show a hand full of those when you run the link: operator search. Google values links differently, but I think it’s a safe bet that the links that do show up with the link: operator search are counted.

So, to let this make more sense, here’s a concrete example, following the 3 steps outlined above:
1. search google for “online merchant account”
2. the first result today is MerchantExpress.com – so, we’ll then run this search in google –> link:merchantexpress.com
3. as of today, Google shows –> About 186 results (0.27 seconds). We’ll go through all of them, and evaluate if we can get a link from the same page.

Again, sometimes you’ll be surprised by the links you’ll find. You definitely will see some free for all links pages. In the online services industry, you’ll often find footer links. For instance, a design company might put a ‘designed by XYZ Studio’ in the footer.

You won’t be able to get a link on every single page. Maybe from doing a search, you might only uncover one or two good links. But, that’s just part of link building. It’s a slow process, and you have to take a long term view about things.

In general, you should avoid building links too fast under any circumstances. It could show that your site was being aggressively SEO’d. If your site was banned or blacklisted from the index, all of your hard work would go to waste.

Following a system like this, you should be safe building 2 to 3 links per week to a site. Over a few months, that should be enough to help you to rank for some less competitive search phrases. Over the course of a year, that would be enough to help you to rank for some moderately competitive search phrases.

If your site is new, don’t expect to rank for very competitive search phrases for the first year or two. Again, if you try to push things too aggressively, you might just get your site banned from the index.

Also, here are a few additional tips to find more links for your site, using the same exact technique.
1. Go deep. Obviously, you want to check the links for the top 10 domains in a SERP. You could go as deep as 20 or 30 domains, and still find some pretty decent links that way.
2. Go broad. Do other searches related to your main phrase. In addition to ‘online merchant account’, you could phrases like: online merchant account bad credit, online merchant account services, online payment gateway, online shopping cart. For many searches, Google will show you a list of related searches at the bottom of the page.

I’ve noticed something about my work lately. Sometimes I feel it’s just so hard to keep focused on one thing at a time. For instance, I’m chatting with some person about a project. While I’m chatting with that person, I feel compulsion to surf some websites that I might have been looking at. I might also read an article, while carrying on this conversation with one or two other people.

In actuality, I should keep my focus on the chat conversation that I’m having with this person. But there are a few problems here, that make it easy for my attention to be divided, and divided again.

For one, it’s easy for any person to interrupt you when you’re online. There are definitely times when I want, or even expect to be interrupted. If there’s something important happening, which requires my quick feedback, I would be mad at a person if they were too timid to interrupt me.

On the other hand, being online all day, and being a person with many online contacts – I find my day is constantly getting interrupted. It’s hard to really sink into my projects, and really get into the flow of working. If I don’t get into the flow of working, my projects will move forward very slowly, and I end up getting frustrated at the quality of the results!

I think I’m not the first person to ever notice this. I saw an article online today called “The Acceleration of Addictiveness”. The author talks about how more and more things are designed to be addictive. One of the most addictive things is the Internet!

You can sit in front of the computer, and surf website after website, and actually ‘feel’ like you’re working. But, you’re not. You’re only receiving input. Your brain is not being forced to process the information, and produce any creative output as a result of it.

In some regards, endless surfing of websites online is similar to watching TV. It’s an activity where the person receives information, instead of produces information.

In my case, realizing this was the first step. I’ve since tried to create more rules and discipline for myself. I try to catch myself if I find myself straying too far away from my daily goals.

One of the work habits that’s been the most useful for me so far is the Pomodoro Technique. The idea is to set a timer for 25 minutes. During that time, you keep a laser focus on your goal. No matter what instant message pops up, no matter what interesting article is waiting to be read, you don’t take your focus away from your goal. If ever you find that you did stray from your goal, you have to stop your timer, and reset it back to 25 minutes again. You don’t give yourself credit for any work at all.

It’s a strict framework, but I find it is a good way to keep my mind on track with all of the things I need to manage every day. Without a way to keep your mind disciplined, it’s just too easy to get distracted when sitting in front of a computer connected to the Internet!

I recently registered a domain on the topic “coverdell education savings account”. From my keyword research I found it to be interesting because it had a high cost per click on Adwords. Also, it seemed to have less information published about it on the internet. At the time of publishing this posting, a search for this phrase in google, without quotes, finds 53,200 pages. This number is on the lower end of what I usually come across. With all of this information, I assumed it would be a phrase I would be able to rank for quickly.

After a few months of work, I’m not ranked anywhere in the top 700 results. I thought this was surprising, so I decided to take a closer look at the SERPs, and understand who the competition were.

Perhaps, unsurprisingly, the competition here is indeed stiff. The top result, IRS.gov, followed by Wikipedia.org. Given that both sites had very relevant pages about the topic, and are exceptionally strong domains, I don’t think, in reality, I have any chance of bumping them out of the top 10.

Two other domains here were also very strong – statefarm.com and finaid.org. Finaid.org was registered in 1994, and shows up as a PR8.

The weakest domain in the SERPs here is a site called Good Financial Cents.com. It was only registered in 2008, however, it seems to have picked up a relatively large number of links from other blogs in that time period. It’s strength appears to be primarily these other links from other financial type blogs.

This SERP turned out to be relatively surprising. It could be because this is a government sponsored program, the SERPs are filled with government agencies, and major companies that are partners in this program. This isn’t a strategically important keyword for me, so I think I won’t pursue aggressively.

Lately, I’ve been getting interested in how to calculate keyword difficulty for SEO. More or less, I want to have a reasonable and quick way of determining how much effort would be required to make some keyword rank in the top 10 for Google. Why is that important you ask?

A few reasons. One, is to estimate the opportunity. There might be some economic opportunity for ranking for a certain keyword. Perhaps, ranking for that keyword would bring my business a steady trickle of new leads. Maybe one or two of those visitors would become a new customer every year. With that hope, if I could move my website to rank in the top 10, I might be willing to invest a few hours a week for the next year to get that ranking. On the other hand, if it would take 3 years, to have any reasonable hope of ranking my website for that keyword, I might as well forget it, and focus on something that would have an impact sooner. 3 years later, the market might have changed so significantly, that there might not be any value in ranking for that keyword anymore.

With that in mind, I’ve started looking analytically at various search engine result pages (SERPs), to see what I could learn. One keyword I tested was “credit cards”. I feel it’s a good example of a highly competitive keyword. It would not be easy for any site to crack into the top 10 of this SERP.

In the top 10, all 4 major credit card systems are represented (Visa, Mastercard, Discover, and American Express). In addition, there are two banks credit card offers pages represented (Citi and CapitalOne). The other 4 are content or comparison websites. The top ranked site is an exact match domain – creditcards.com. The other 3 content sites all have the word “credit” as part of their domain.

Of the content sites, 3 are pretty major players. One is owned by Bankrate, which is a major provider of mortgage and financial information online. Credit.com and Creditcards.com both have professional management staff, and seem to be pretty active in providing up to date information about the credit card industry. The only weak result in the top 10 here is credit-land.com. From studying the links, there are obviously a number of purchased links, such as purchased blog posts, and just plain text link ads. However, it is a 10 year old domain (first registered in October 2000). Google does give some weight to older domains.

So, my impression here, is that it would be virtually impossible for a newly registered domain to ever rank for this site. The weakest site here, credit-land.com, is a 10 year old domain, and has purchased a number of links. They most likely will fall out of the rankings once they were discovered. However, the other 9 seem pretty solid, and very relevant to the search.