Increase Adsense Revenue with A/B Split Testing

My Adsense earnings have been fairly flat lately, and I’ve been wondering what there is I can do to improve them. I’ve read again and again about the need to test different ad variations in order to figure out which performs the best for me. I’ve always stumbled though when it came to actually doing so as it seemed like such a large, labor intensive process.

Reasoning that someone, somewhere must have figured out an approach to solve this problem, I started hunting for a solution. I stumbled upon one on Google’s own Adsense site that is so simple that I actually laughed when I first read it. It took me all of about 30 minutes to implement it for the first time, and I can’t imagine it will take me more than 5 minutes per test going forward.

What is A/B Split Testing?

The concept, one known for years in formal market research, is known as A/B Split Testing. By testing two things, one called A and one called B, against each other, you can determine which one performs better. Simple, of course, but the trick is that for it to be reliable, you need to change only one thing at a time between A and B, and the changes need to be tested simultaneously, so that external factors like time of day or day of week don’t influence the results.

For example, you might have two ad units that are identical except that one uses a blue background and the other a white background. To use A/B Split Testing to determine which one makes more money, you would randomly alternate between the two ad variations as they are presented to your visitors. After showing the two versions many, many times, and measuring your earnings, you could feel confident about which add to keep.

Javascript to the Rescue

OK, without further ado, here’s the code:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
<script type="text/javascript">
    var random_number = Math.random();
    if (random_number < .5){
		<!--
			/* INSERT FIRST ADSENSE CODE BLOCK HERE */
			google_ad_client = "pub-XXX";
			google_ad_slot = "XXX";
			google_ad_width = XXX;
			google_ad_height = XXX;
		//-->
    } else {
		<!--
			/* INSERT SECOND ADSENSE CODE BLOCK HERE */
			google_ad_client = "pub-YYY";
			google_ad_slot = "YYY";
			google_ad_width = YYY;
			google_ad_height = YYY;
		//-->
    }
</script>
<script type="text/javascript" src="http://pagead2.googlesyndication.com/pagead/show_ads.js"></script>

Now if you have some programming language experience, this is probably speaking to you rather nicely. If not, it might take some explanation. The premise is incredibly simple. It’s just a bit of Javascript that you can use to randomly alternate between two different ads. Wherever you would normally insert a standard Adsense code block, you would use this code instead, with a few minor modifications. The only thing that has to be changed is to replace the “first” and “second” code blocks with your real Adsense code representing the two ads you want to test. Neat, huh?

Track the Results with Custom Channels

While the Javascript will rotate your ads nicely, to view the results of the test will require a few more steps:

  1. Create the ads you want to test within Adsense. I’m sure you’ve already know how to do this, but I include the step for completeness. Give them names that clearly describe their attributes and how they differ, such as TopBannerWithBorder and TopBannerWithoutBorder.
  2. Create two custom channels, one for each ad that your are testing. Again, make sure to use descriptive names so that you won’t have to try to remember what they represent. It’s not a problem to reuse the names you used for the ads. That’s what I do.
  3. Assign each ad to its own, separate channel.

Why use channels in this way? Due to a limitation of Google’s Adsense reporting, you can’t report on ads directly. You can, however, report on channels. By creating a custom channel for each ad, you can get around this limitation.

Wait a Week and Run a Custom Report

It’s important to run the test long enough to have confidence in its conclusion. How long is long enough? Well, you could delve into statistics and come up with a number of different answers, but for me I’d like to see a few thousand impressions if possible, spread out over at least a 24 hour period. I usually just let a test run for a week, and then check back to see what happened.

To actually see the results of the test, you have to create a custom report within Google’s Adsense “Advanced Reports” module. The process is simple:

  1. Choose a date range that spans the period of your test.
  2. Under the Show heading, select Channel data, and then select the two channels that contain the two ads in your test.
  3. Under the Group by heading, select Channel.
  4. Under Choose Units heading, select Combined. Doing so is necessary if you are testing standard ad units against text link units.

Click that nice big Display Report button, wait a few seconds, and view your results. As you now know for sure which ad performed better, the next step is to drop the losing ad, and create a new one to test against the winner. You can continue to test again and again until you find just the right combination to ensure that you are doing all you can to maximize your Adsense earnings.

18 Month Progress Report

It’s hard to believe that it’s been 18 months since I launched ESL Jobs World. My active development has had roughly two phases, one lasting about two months in March and April of 2007, and another for two months in November and December of 2007. The remaining time I’ve let the site more or less take care of itself, with periodic checks to ensure it’s still up and running successfully. I’m about to embark on a new round of development, though, so I thought it would be helpful for me to summarize the current status of the site.

Traffic Overview

My traffic has been consistently growing month after month. It has been averaging about 30,000 unique visitors per month and 100,000 page views per month. The sources of the traffic are roughly 78% from search engines, 11% from links from other web sites, and 11% from direct traffic such as bookmarks or visitors typing the address into their browser. This traffic has all been organic. While I experimented with PPC advertising for a few months in late 2007 and early 2008, I didn’t feel the ROI justified the expense, and my sense was that there was significant click fraud as well. As a result, I have not engaged in any PPC advertising in well over six months, and will likely not pursue it for the foreseeable future.

Search Engine Results Overview

I have been trying to rank well in the major search engines for a few keyword phrases, most specifically “ESL jobs” and “English teaching jobs,” as my research has shown they are the highest searched terms that most closely correspond with the niche of ESL Jobs World. My results have been quite good, especially considering the length of time my site has been live compared to its competitors. For a search using “ESL jobs,” the site is ranked #2 with Google, #6 with Yahoo, and #3 with MSN. For “English teaching jobs,” the site is ranked #2 with Google, #2 with Yahoo, and #1 with MSN. These rankings have been fairly consistent month to month, with little variation. I periodically work to establish more inbound links using those keyword phrases with directories and related sites, which is the primary means to further increase their position.

Subscriber Overview

There are two types of users that can register with ESL Jobs World, teachers and employers. At present, over 1200 teachers are subscribed to a weekly newsletter that summarizes recent job posts, and over 1400 employers have registered to post jobs. Of the employers, slightly more than 10 percent are active every week, and about 300 jobs are posted every week. Growth of both registered teachers and employers has been steady and consistent, which is a necessary condition to ensure a healthy community and the future success of the site. Every week, about 2% of the unique visitors to the site choose one of the forms of registration. Increasing this percentage is a key objective.

Income and Expense Overview

Of course, the bottom line is how much income the site is generating. At present, it earns about $400 per month exclusively from Google Adsense. Its only direct expense is $10 per month in hosting charges. I briefly experimented with Kontera and AdBrite, but did not feel either was a good fit for my particular niche as the ads they served were not well targeted. Of course the overall ROI of the site is strong, but it’s nowhere near the original goals I set when it launched. Fortunately, there a number of areas that I can pursue in the weeks and months ahead to significantly increase its monthly income.

Picking the right classified ads platform for Joomla

With the decision to go with Joomla as my CMS, my next hurdle is to figure out what additional components I’ll use with it. As I’m building a jobs site, my first priority is to evaluate and select a classified ads platform. I’m going to want to keep my costs down, so my personal selections will be somewhat biased as a result.

Deciding between home grown or off the shelf

While I can hack around a bit under the hood with HTML, I’m a little leery about trying to put together something as complex as a classified ads platform from scratch. For someone with more time or programming talent than me, this might be a solution to consider. A home grown platform could be completely tailored and customized, allowing it to stand out compared to any competitors. As the existing choices are so feature rich and easily modified, this is not a path I intend to pursue.

The solutions available for classified ads vary in price from free–always a great option–to quite expensive. As I said previously, though, as my budget is limited I’m only going to focus on the free choices that are available.

Choosing between tight and loose integration

While there are a good number of free classified ads choices available, not all of them integrate closely with Joomla. In reviewing my options, I realized that while the tightly integrated ones are easily installed and configured within an existing Joomla site, they seem to be less frequently updated and supported. In contrast, the less integrated ones seem to be far more actively developed with vibrant support communities, but can be a real chore to seamlessly interface with Joomla. It’s a difficult trade off between ease of use and a perhaps fuller features set.

I discovered a few options, such as Jobline and Marketplace, which appear to be tightly integrated with Joomla. It seems that both of these products are no longer being developed, though, and so in the end I decided that selecting one of them would be unwise. Instead, I elected to solely focus on options that were currently and actively supported.

Comparing Noah’s Classifieds and AdsManager

Noah’s Classifieds seems by far to be the grand daddy of classifieds ads platforms, with an immense base of installations and users. The main site doesn’t have a support forum, though, which was troubling at first until I stumbled upon the Noah’s Classifieds Unofficial Support Website which has a great forum. There is a version of Noah’s for Joomla, too, making it appear to be quite attractive. Because it was modified to work with Joomla, though, the development of the Joomla version and the normal version aren’t in synch, and I wasn’t personally comfortable with using it.

The other choice was AdsManager, which was developed specifically for Joomla. It’s website isn’t glamorous, but the software itself is impressive. It was designed from the ground up to be very closely integrated with Joomla, allowing it to share user accounts and permissions, which makes it very easy to setup and configure. It also has a very active support forum. I asked a number of questions, and got responses very quickly, enabling me to quickly move forward with the few configuration issues I experienced.

The best classified ads solution for Joomla

I felt a strong case could be made for either Noah’s Classifieds or AdsManager. In the end, I opted to go with AdsManager, though, due to its seamless integration with Joomla and the strength of its community support. Thus far I’ve been very satisfied with both its features and support, and feel it’s the right choice for the jobs site I’m building.

Getting started… again!

It’s been quite some time since I last posted an update… a far too common occurrence with blogs! I have been actively working on ESL Jobs World, though, and thought it would be helpful to post some new information to get everyone caught up with where I am.

Building Traffic

The good news is that a lot of my efforts to build traffic to the site have been successful. I’ll get into the strategies in another post, but I’m happy to report that ESL Jobs World now ranks in the top 10 on Google, Yahoo and MSN for its primary keywords. It’s home page is now PR 5 with Google and it’s secondary pages are all PR 4. It took some time for my search engine optimization efforts to pay off, but it looks like fortunately they have done so.

Along with the increase in traffic has come–as you would expect–an increase in activity on the site. While in April very few jobs were being posted, at this point in early December the site averages anywhere between 50 and 100 per day. Yes, that’s per day! This is great news, as one of the keys to the long term success of the site is to make it attractive to the teachers that will visit to find work. It goes without saying that a job site without jobs would not be terribly useful.

Building Revenue

During the first six months of the site being live, I was very sparing in trying to monetize it in any way. I wanted to just let the content speak for itself, and not give visitors any reason whatsoever to turn away. Especially in my niche of education, members can have a toxic reaction to a feeling of being sold in any way.

Recently, though, with the visibility and credibility of the site increasing, I’ve added a few Google AdSense blocks. Nothing excessive, I hope. I experimented quite a bit with their size, shape, color and placement before I got what I felt didn’t detract from either usability or the overall look and feel I was trying to present.

I’ll go into more detail in a future post, but the bottom line is that the site is now earning a small but steady amount of income. My plan is to take my earnings and put them back into the site for the foreseeable future. I’ll be looking to optimize the earning potential of the site in upcoming weeks and months, and I’ll be sure to share my efforts along the way.

Enhancing Value

I’ve always been a big believer that the only way for a site to be truly successful is for it to consistently deliver something of value to its visitors. Figuring out what is valuable within your niche, though, can often be difficult. It can be a fine line between adding every bell and whistle that pops into your head and building content or features that actually entice people to visit and return.

I feel I’ve done reasonably well in this regard by doing the obvious… asking people their opinions! In a future post I’ll get into more details, but the kernel of what I’ve done has been to try to create a natural, repeated dialogue with my visitors. I’ve then looked for common themes, and acted in areas where I felt I reasonably could.

Looking Ahead

There’s so much that I’ve done in the past few months that I’ll be making an effort to go back and review what I’ve done. I’ll try to highlight as much as possible what worked and what didn’t work, as much to remind myself as to hopefully enable others to avoid my mistakes. I hope you’ll check back again to see!

Using feedback to improve my projects

I thought I’d take a detour from discussing the development of ESL Jobs World and address a strategy that has helped me in the past on similar online projects. It’s really quite simple, but very effective when followed.

Ask for feedback, and then act on that feedback to make improvements!

A few specific tips that I have used in the past to make this strategy actionable:

Get email addresses

This is very basic, but without doing this step, you basically have no way to have a conversation with your visitors. Whether it’s via a registration process or a newsletter subscription, make sure your site has some glaringly obvious way for people to give you their email address.

Confirm email addresses

In addition to being good form for opt-in marketing and staying on the right side of the law, you have to make sure your process to get email addresses requires confirmation. It goes without saying that a bad email address is of no use. Whatever your method, make sure visitors get an email with a confirmation link.

Ask specific questions

Asking “what should I do to make my web site better” is pretty broad, and “do you like my web site” is pretty much useless. Asking “what can I do to improve my registration process” is much more focused and as a result much more likely to get a useful response.

Ask about competitors

One of the questions I love to ask is “What’s your favorite web site in , and why is it your favorite?” It’s great to be able to find out who the competition is, and what’s helping them stand out.

Ask about best/worst

Another favorite question I ask is “what do you like best about ? What do you like least?” If you ask that question enough, certain themes should begin to emerge that can help you focus your efforts on what to improve.

Ask one question at a time

Nobody likes paging through gobs of questions. If someone feels like they’re filling out a census form they’re going to delete your request for feedback in a heartbeat. Pick one question, ask it, and leave it at that.

Don’t be a pest

Don’t spam people day after day with questions. They’ll get annoyed and will unsubscribe or unregister or worse. I typically reach out about once a month. Don’t be a pest, and people will be far more likely to respond to your questions.

Follow up for clarification

If you get a response and you’re not quite sure what they mean, say so. One of the things I like to do when I get a suggestion is to ask for examples from other web sites where it’s been done well. You’d be surprised how frequently someone will kick back a few links for you to review.

Follow up on changes

If you get feedback and make changes, let the person know and ask them for their opinion. Sometimes you’ll get a nice thumbs up, but sometimes you’ll learn that you’re not quite there yet. You need to know.

Always say thank you

You’ve just asked for help, someone has actually gone against the odds and been kind enough to respond, and then you don’t have the common courtesy to say thank you? Do you think that person is likely to help you again? Exactly. Always say thank you.

Drupal or Joomla? Picking a CMS

I’ve decided that a Content Management System (CMS) is the best foundation for my site. My next step is to choose the right one.

A great place to start is OpenSourceCMS, a site with user reviews of pretty much all the major players in the CMS space. What sets them apart is that they also provide live demos of each CMS they cover. You can actually log in to the front end or the back end of each one, reconfigure it, and make changes to your heart’s content. Every two hours they “reboot” and put everything back to a fresh install. It’s a great way to experiment without having to go through all the time and hassle of installing each system yourself.

Installing Drupal and Joomla on my host

In addition to testing each platform on OpenSourceCMS, I also wanted to install them myself to gauge how easy they would be to work with. Fortunately my hosting provider, Host Gator, uses a product called Fantastico which makes installing Drupal and Joomla as simple as a few mouse clicks. Both installed successfully with minimal effort. Purists abhor Fantastico, but for my purposes, it was a quick and easy way to get up and running quickly to be able to start kicking the tires of each product.

Installing Drupal and Joomla locally

As I’ll need a test environment before long, installing both products on my local machine is a good idea as well. Before I can do so, though, I need to install the LAMP (or WAMP) stack commonly used by Open Source software. LAMP enables my desktop to act like a web server, so that I can run everything from my local machine just as if it were running on my host.

For the curious, LAMP stands for Linux Apache MySQL PHP, and they are the four products that make up the foundation that Drupal, Joomla, and countless other products use. WAMP is essentially the same thing, but uses Windows as the operating system. Each product offers its own installer, and I got WAMP working on my local machine in no time.

The local installations of Drupal and Joomla were a bit more involved. I had to understand how to setup MySQL databases, and know the right answers to a number of questions, although the wizards that each product offered were pretty good. A complete novice would probably be overwhelmed, but I found it pretty much a snap to get both going quickly.

Picking the best CMS

At the end of the day, neither product stood out as being an obvious, dominant solution. Each had its own quirks and metaphors for organizing information and accomplishing tasks. As might be expected, there were a lot of differences between them. Like learning a language, becoming an expert in either platform would likely be a long process.

I spent a considerable amount of time in each platform creating content, changing around templates, activating modules, and doing my best to get a reasonable feel for what it would be like to work in each environment. I downloaded a number of add ons that were commonly available to assess how easy it would be add to their core functionality. I evaluated the search engine friendliness of the URL’s they generated. Above all, I tried to get a sense of how comfortable I felt in each product, understanding that I would likely be spending a lot of time with my final choice.

It seemed that Drupal had some incredible abilities to define different types of content. I could create a “job” object and define what attributes (title, country, description, etc.) it should contain. It was pretty powerful. I liked how everything was available from one page, without loading and reloading all the time. Adding and formatting content was easy. There were a large number of included modules that could be enabled, from forums to blogs to comments, so that I could extend its functionality quite easily. There’s a lot to like about Drupal.

Joomla draws a hard line between “front end” (what a visitor sees when visiting the site) and “back end” (what an administrator sees to control the site), which is more consistent with other applications I’ve used in the past. The types of content are essentially fixed, although there is a considerable variety in how you can display them. Joomla also comes with fewer bundled features than Drupal, which can be a good or a bad thing depending on your perspective. The number of unique add-ons, though was impressive. In fact, Joomla has a whole section of their site dedicated to Joomla extensions. It took some time to figure out the difference between a component, a module, and a mambot, but once that was clear, I was able to extend Joomla quite easily.

To make my decision, I finally decided to look at the community that surrounded and supported each product. As a test, I made a general inquiry into the support message boards of each product. My Joomla post received a number of helpful responses in a relatively short time frame. My Drupal post languished for days, and even after asking a second time for feedback, was still unanswered. As I’m going to be learning each system, and relying on others to help me, this experience ultimately swung the pendulum to Joomla.

With this key decision out of the way, my next task will be to choose the right mix of extensions to add to my base Joomla installation.

Deciding between a forum, a blog or a CMS

My competitive analysis of the major and minor players in my niche has given me a much clearer picture of the web site I plan to create, and a clear set of requirements has emerged. A fundamental question, however, remains unanswered. Just exactly how am I going to organize and present all of this information in a manner that is both scalable and appealing?

Evaluating different ways to organize and present information

In the land of the Internet, it’s rare to come up with a really unique problem so a little further competitor comparison can help me understand how other job sites have solved this problem in the past. Going back to the large and small competitors from my initial research will help my narrow down my choices quickly. As expected, I find that there are many different ways that I could enable schools to post advertisements for ESL jobs, and teachers to find and respond to them.

Considering a forum

I could use a forum as the foundation, and set up different categories for the different regions of the world. Employers could then register and post freely, and teachers would have the ability to browse and search through the listings. Controlling the quality, though, would probably be an issue, as by definition a forum post is completely free form with little in the way of consistent look and feel. It seems that although a few competitors are using this approach, it’s not terribly popular. I’m also not very pleased with the overall look and feel of a forum approach, which at least to me doesn’t come off as feeling professional. For my own purposes, this is not an avenue I intend to pursue.

Considering a blog

A blog is another possible approach. They’re very easy to set up, and I could enable tags to classify the different types and regions of jobs to help teachers find them more easily. Creating an open format for schools to post, though, would likely be a nightmare. A blog would probably be acceptable for a single school that wanted to post a running list of current jobs, but I could quickly see such a method becoming very unworkable as the site scales. Surprisingly, a variation on this theme is used by one of the biggest and oldest job sites in my niche. Why is it so successful? My assumption is that when that particular site was created the blog format was one of the few solutions available that could be implemented easily. The approach feels considerably dated at present, though, and I think it would be hard to attract new visitors using it.

Considering a CMS

Another general method is to use a content management system, or CMS. These are a bit more complex than forums and blogs to get set up, but the trade off is better control over the organization and structure of the content. They’re great when the need is for multiple people that need to post information, yet with a consistent look and feel throughout. No competitor is using a CMS in isolation as a solution, but a number seem to have used a one as the core to which greater functionality was added. This possibility in particular is attractive to me, as most CMS platforms have a lot of third party developers contributing added features. It’s likely that someone has created code that works with a CMS that makes posting jobs easy.

Considering a niche solution

The final choice is to look for something tailor made to the exact type of information I want to organize and present. As expected, there are systems available that solely focus on classified advertisements. At least one of the competitors in my niche is using a very popular one called Noah’s Classifieds. I’m tempted to choose this path, as it would appear to offer the shortest path to get up and running quickly. I’m concerned, though, that while it obviously excels at ads, there are other areas where it may be lacking. While a niche solution would get me up and running in this one area quickly, I’m also afraid it could limit my future options. I sense it would be a real chore to bolt on a forum or a blog without going through a lot of custom effort to integrate them.

Settling on the right foundation

In the end, I decided to use a CMS as the core of my web site. Doing so offers me lots of future flexibility to easily offer some of the additional features I’ve determined will be necessary to differentiate myself, and keeps the door open for even more features that I might not have thought of originally.

My next step will be to review a few of the major CMS platforms available, and see which one offers the best off the shelf solution for my own needs.

Analyzing both major and niche competitors

I now have an actual live domain name attached to my hosting account. My next step is to get a little more specific about the type of content that should appear on my web site, as well as its general look and feel. My answers to these questions will help me determine what tools I’ll need to use to build the site, and also help me flesh out my ideas about my next steps for moving forward.

Reviewing the big name competition to determine best practices

To start, I like to look at some of the successful big name companies in my general category. For online job listings, Monster.com and Career Builder are clearly on top. My thinking is that these folks have probably invested a huge amount of time and money tinkering with different formats to figure out the right way to connect employers and job seekers. By observing their overall approach, I can benefit from their investment and shorten my own learning curve.

For example, a common theme is a fairly clean home page with a really big search box front and center. Both have a white background and use fairly subdued color palettes, although Monster is a little flashier. The focus of the home page is also pretty much exclusively geared at job seekers–both sites have a different “front door” for employers.

Another key distinction is that they also both offer more than just job listings. There are articles geared towards helping people figure out the best strategy to find a job, and even how to choose one career over another. Both have a salary calculator tool. And they each have a method for visitors to request a periodic email when new jobs that meet their criteria have been posted.

Assessing the web sites already in my niche

After getting a good sense of what the large players are doing, I also want to look at the smaller sites operating successfully in my direct area of ESL Jobs. Identifying common themes across this group will help to ensure I meet the minimum bar of what the current market expects. I can also use my visits to see if I can find any opportunities where I can offer something to differentiate my own site. In particular, I will look to see if there is anything that the big companies are doing that the niche companies are not.

To find these top competitors, I simply head over to Google and type in my keywords. The top 10 web sites returned are the ones I will survey. A quick pass shows how incredibly different they are from one other. Judging by some of the dates on their job posts, they are also experiencing very different levels of popularity and success.

Looking deeper, the majority of their home pages seem to be incredibly long, requiring scrolling down repeatedly to view all the information. For a first time visitor, I think this might be a bit overwhelming. In contrast, the big name companies have most of their home page on a single screen, a feature I plan to emulate. The colors and fonts of the niche players are all over the place as well, with only a few having what I would consider to be a clean design and the others being incredibly busy. Most have a fairly dated appearance compared to the current web sites of the big companies.

Somewhat surprisingly, only a single site had a large search box at the top of their home page, and only two had a clearly visible way to receive new job postings by email. It was interesting to note that quite a few provided the opportunity for a teacher to post a resume that could be viewed by interested employers, a feature I had not even considered. In an odd twist, one of the top 10 sites didn’t even appear to be functioning, as a lot of its links generated script errors.

As far as providing material above and beyond just job postings, quite a few seemed to be part of much larger sites with content that ranged all over the map. A few had articles on teaching conditions in various countries. A few had tips for finding a job. None had a salary calculator.

Learning lessons from the big and the small

My time spent evaluating both the large and niche players helped me narrow down the features and appearance of the web site I intend to build. It’s interesting to note that while much of what I assumed going in turned out to be correct, a good bit was a complete surprise to me. It underscores the benefit of taking the time to really understand the market before jumping into it.

In summary, I determined that my minimum bar for success is a clean, compact, professional look with a big search module at the top of the home page. To really stand out, ESL Jobs World will need to present more than just job listings, but also tools to help teachers find their next job. I need to separate the teacher part of the site from the employer part of the site. I need to have an easy to find email newsletter to notify teachers of new jobs. To really cover all angles, I also need to enable teachers to post their resumes so that employers can review them. Finally, providing supplemental content not found elsewhere like a salary calculator would really help to differentiate me.

All in all, a tall order, but at least now I have a target in my sites.

Finding a web hosting provider

Selecting and registering a domain name is a great first start, but without a place for it to live, it’s basically worthless. I need a hosting provider, a company that will provide a server where my web site lives, including its hardware, software and connectivity to the Internet. There are quite literally thousands and thousands of companies that provide this type of service, but how to choose? Fortunately there are a few ways to narrow down the choices.

Free hosting or paid hosting?

Choosing whether to pay for hosting is my first decision. Although it may seem counterintuitive, there are actually a number of free hosts. Many of them specialize in either web sites or blogs. They are limited in the type of software I can use to run my site. If I just need a few static web pages, or an out of the box blog, though, they can be great choice.

Of course as in real life, there is rarely a free lunch and free hosting is not without its caveats. As a rule, free hosting sites like GeoCities won’t let me use my own domain name, and as mentioned earlier, a big piece of branding is my domain name. They also place ads within my content, and receive the revenue from anyone that clicks on them. They put strict limits on the amount of visitors my site can receive within a fixed period of time, and will effectively turn off my site if I exceed them.

For my purposes, a free host isn’t an option. I want to be able to fully control the look and feel of my site, ensure its always available despite its popularity, and make sure the only ads that visitors click are ones that provide revenue to me.

Shared hosting or dedicated hosting?

My next decision is to decide whether or not I want to share a host with other people running different kinds of online businesses. As might be expected, sharing a host is much, much cheaper than the alternatives. With a good hosting company, a shared host can be a completely acceptable solution. Some companies put way too many users on their shared systems, though, causing performance to suffer for everyone.

On the other end of the spectrum is choosing a dedicated host. Buyer beware, this can be very expensive! Of course with high cost comes high performance, and you really do control the system. As long as you aren’t breaking the law, you can pretty much do whatever you want with one. Need to run some scripts that put a heavy load on the system, or send out lots and lots of emails to list of your subscribers? With a dedicated host, how you use it is up to you.

A middle variation that some companies offer is semi-dedicated hosting, where the total number of shared users is limited but you still don’t have exclusive use of the server. This can be a good middle ground for people whose web-based businesses are beginning to take off, and have outgrown their initial shared hosting plan.

For my purposes, a shared host is where I’ll start out, although I’ll make sure to go with a company that can upgrade me to a semi-dedicated or dedicated host should the need arise.

Finding the right web hosting provider

With a seemingly limitless number of choices, how do I decide upon a hosting provider? As with everyday life, a strong referral from a trusted source is a great method. Barring that, there are some other alternatives to narrow down the field.

A number of discussion boards or forums on the Internet focus on web hosting. My personal favorite is Web Hosting Talk. I don’t need to be a member to read the posts, and I can go straight to the forum discussing the exact type of hosting I’m interested in. After perusing a few pages of posts, a few names started popping out as worthy of further consideration.

Finally, the Better Business Bureau is as good online as offline. I went to their web site, searched for a few of the promising hosting providers, and looked for complaints. Any business of any size is bound to have a few dissatisfied customers, but a large number of complaints should be a red flag.

Pulling the trigger and purchasing a hosting account

In the end, after reading a number of reviews and checking up on several companies, I settled upon a shared plan with Host Gator. The purchase process was very straightforward, and took all of 5 minutes from start to finish. I had now completed two big steps: purchasing a domain name, and purchasing a hosting account.

The last step was to link my domain name to my hosting account. I went to GoDaddy, the company where I purchased my domain name, and filled out a short form changing my name servers to the ones assigned to my hosting account. Name servers control how a web browser knows how to find your web site. Once the name server changes were complete, my hosting account and my domain name were linked, and typing in www.esljobsworld.com landed on a generic Host Gator welcome page for new accounts. Mission accomplished!

Next, I’ll need to do a little competitive research to help determine the content and appearance of my new site.

Selecting and registering a domain name

With the topic of my first web site decided, I now need to figure out what to call it. Like any business, picking a name is a critical decision. Unless you have a huge marketing budget to give meaning to a somewhat nonsense word like Dell or Amazon, my thoughts are that your name needs to tell people exactly what your site is about. For example, if presented with a name like Festigle or Discount Pet Food, both in the same line of business, which one is easier to figure out? Exactly.

In the Internet world, of course, your business name is also your domain name. And just as in the real world, your business name needs to be unique in your market. As the Internet is global, that means your market is global, which makes it much more difficult to find a unique domain name that accurately describes your niche that hasn’t already been taken.

Choosing the right keywords

Fortunately there are a lot of methods that can help you select a domain name. The first area that I like to focus on is keywords. Keywords are the basic building blocks that define the topic of your web site. For me, the keyword that describes my general topic is jobs. There are a lot of keywords that describe my niche, including ESL, EFL, TESL, TEFL, and TESOL. A combination of these keywords would make for an ideal domain name.

Of course, not all keywords are equal. Some will be more popular that others, meaning they are more commonly used when people use search engines like Google or Yahoo. Using a more popular keyword will be more helpful in attracting visitors to my web site, as its more likely it will be the keyword someone uses when they search.

There are a number of tools that will show you the popularity of keywords. One, The Keyword Suggestion Tool, will show you the rank of a series of keywords across the major search engines, in addition to a lot of other information. Entering in ESL jobs, EFL jobs, TESL jobs, TEFL jobs, and TESOL jobs, I discover that ESL jobs is far more popular than the rest. As a result, I’ll try to create a domain name that includes those two keywords.

Considering top level domains, hyphens, numbers and length

One factor in choosing a domain is selecting its top level domain, or TLD. There are a number of TLD’s, including .com, .net., org, and .info. I’m a huge fan of the .com TLD for a commercial site, although the others are valid for certain niches. On balance, though, I’d always go with a .com. The .com TLD most strongly resonates with people, and demonstrates–at least in my opinion–that your web site is legit. If my ideal domain name was taken in .com and available as .net or .org, I’d forego the ideal name and look for another that was available as a .com.

Another consideration is the use of numbers or hyphens. For examples, it would be possible to register www.esl-jobs.com or www.esljobs1.com. Again, I’m a fan of keeping it simple, and unless there’s a compelling reason to use hyphens or numbers, I steer clear of them. I’ll keep looking for a unique domain name that doesn’t use either of these elements. There’s nothing wrong with using them, but from a style perspective, they don’t resonate with me.

Finally, the length of the domain name is a factor. The last thing someone wants to do is type in a horrendously long domain name. Granted, most traffic coming to a web site is via links, and so domain name length isn’t a deal breaker, but as a general rule, shorter is better.

Evaluating possible domain names

With keywords in hand and some guiding principles, it’s time to discover what domain names are available. In a perfect world, I would just try different www.keywordkeyword.com combinations in my web browser and a perfect domain name would be available. No such luck for me, as both www.esljobs.com and www.jobsesl.com are taken. That means I’m going to need to get a little creative.

I’m going to need to add a word or two to ESL and jobs in order to find a domain name that hasn’t been registered yet. One resource I like to use is The Most Common Words in English. My reasoning is that adding a common word will help ensure my domain name is easier to remember and spell correctly. It also feels better than just pulling out words at random. Of course the word needs to relate to my general topic, or the combined keywords are going to be confusing. Spinning through the list, I settle on word #195: world.

Going back to my browser, I’m pleased to find that www.esljobsworld.com is available! My search for a domain name is complete, and now its time for the next step, registration.

Registering the domain name

This is the first step where it’s going to cost a bit of money. Fortunately, domain name registration is not that expensive, and there are ways to make it even less so. There’s a huge amount of competition in the market, and that means there are always deals around.

I feel it’s important to stick with a name brand registrar like Network Solutions or GoDaddy, two companies with solid reputations that have been around for ages. I personally prefer GoDaddy, as they’re a bit less expensive, and are very active in distributing promo codes that you can use to reduce the price of many of their services. A helpful list of promo codes is available in a discussion thread in the forum at DigitalPoint. Jump to the end of the thread and work your way backwards; you’ll find a number of promo codes that you can use.

The only thing left to do is actually register the domain name, and fortunately that step is by far the easiest. From the GoDaddy home page, just type in the domain name in the form at the top, and follow the steps. Be careful, though, as they will relentlessly try to sell you additional services. Just decline them as you go along, and in less than a minute or two, you’ll complete the process.

So after a lot of research and trial and error, I’ve succeeded in registering my domain name. My next step will be to fined someplace for it to live, a hosting provider!