Today, I want to share another idea: a directory of technical help communities online. A place where a visitor can easily find links to the most active help communities of a specific topic, like HTML, math, physics, language, and so on…

But I’m not talking about a simple collection of forums, mailing lists, social networks and chat rooms. It’s more than that…

Essentially…

It’s a directory, like the 100,000 other directories on the web. But this one has a specific definition. It only lists community sites related to technical subjects.

Quality versus quantity

I realize that for each technical topic, there would be several communities scattered across the Web. The job of this directory is not to include them all, but only those most active. The level of activeness would be gauged using:

• the number of posts in forums
• number of messages per day in a mailing list
• number of users in a chat room

To reduce maintenance, these numbers should be pulled dynamically — using a Web spider of some sort…

Classification is gonna be a challenge

In projects like these, classification is always a challenge. Some help communities will overlap. For example, there could be a help community dedicated graphic design. And another one dedicated to Photoshop. But the topics discussed within each community are the same.

How are these going to be classified? This is something that should be looked into. But…

Before a criteria for classification can be determined, the raw data need to be collected. Having the complete set of data will provide better understanding of how the communities relate and how they should be organized.

Search engine optimization

A well SEO’d resource like this can bring lots of traffic. Here’s a table showing the number of monthly searches for each of the following keywords:

On-site optimization

The structure of the URLs in this directory is gonna be important. For example, if I were to create this directory as part of this website, here’s how I would structure the URLs

• http://msafi.com/online-help/php/
• http://msafi.com/online-help/photoshop/
• http://msafi.com/online-help/math/
• etc…

For each page, I would list all the relevant forums, chat rooms, and mailing lists. I would also write a little paragraph describing each one. These paragraphs would help in capturing long tail search engine keywords.

Other basic on-page SEO stuff apply here, of course.

Off-site optimization

A comprehensive, free, and helpful resource like the one described here is very easy to promote on social networks and to other webmasters. Getting link backs would require much less effort/money than getting link backs for a blatantly commercial site.

Provide supporting articles

It’s a good idea to provide supporting articles with this resource.

Research and write articles to help users ask good questions. Educate people on how to use forums. Tell them to search first if the question has been asked. In chat rooms tell them not to flood the channels with big copy/pastes. And for mailing lists…I don’t know…Tell them how to properly post a reply…

Actually, I might work on this project…

With this post I wanted to clear my head by getting the ideas out “on paper”. This is something I would like to work on; not the data gathering and classification parts (I’ll probably outsource those tasks), but the programming, development, and SEO part.

And if I do it, it’s indeed gonna be part of this website. Wanna contribute?

Related posts:

1. Create a Chat Website (Or Add Chat Rooms to Existing Website)

After reading some explanations, I still wasn’t satisfied. I couldn’t find a logical distinction between categories and tags.

So here’s what I decided to do…

Why it’s enough to just use tags and ignore categories

Categories and tags are essentially the same thing. Using both to organize posts is an overkill. Just use one — the more versatile. That would be tags. As for the categories, I’m letting all my posts fall under Uncategorized.

Tags are more concise and descriptive than categories. And after some time, the tag cloud gives very accurate visual depiction of the topics covered on a blog. The tag cloud tells what a blog is (mostly) about.

But to get the most out of tags and the tag cloud, some customization is required.

Theme customization

Themes come pre-equipped to handle categories. The theme I use on this website, showed the categories under each post. It showed the tags at the top (for SEO purposes, supposedly). I removed the categories and brought the tags to the bottom, where categories were. I renamed tags to Topics, as you can see at the end of each post.

Obviously, this required editing of the theme files.

Creating a Usable Tag Cloud Widget

The default WordPress tag cloud is ugly. That’s why very few websites use it. That’s a shame because like a said, a tag cloud can be very useful. That’s why I took some time to fix the tag cloud on this website and make a real use of it.

Here’s a sample of what I created

Custom WordPress Tag Cloud

Most popular topics appear first and with the largest font size. Then topics that are not covered as much appear later in the list and with smaller font size. Of course, this tag cloud changes as the variation of topics on the blog changes. After all, that’s the nature of a tag cloud.

One would think creating a tag cloud like that should be simple, right? Well, not if you want the topics to be comma-separated like mine are.

My custom made tag cloud widget

This tag cloud has the following features:

• Tags are comma separated
• Most popular tags appear first in the list

Here’s the code for the widget:

aadd_action('widgets_init', create_function('', 'register_widget("MX_Tags_Widget");'));
class MX_Tags_Widget extends WP_Widget
{
    function MX_Tags_Widget() {
        parent::WP_Widget(false, $name = 'MX Tag Cloud');
    }

    function widget($args, $instance)
    {
		extract($args);
		extract($instance);

		echo	$before_widget;

		// My widget says, "This Website is (Mostly) About".
		// You can change that to anything you want.
		echo '<h2 class="widgettitle">This Website is (Mostly) About</h2>';

		// You can change the behavior of this tag cloud widget
		// by passing different parameters
		// See a list of possible parameters at
		// http://codex.wordpress.org/Template_Tags/wp_tag_cloud
		$tags = wp_tag_cloud('echo=0&format=array&largest=1.4&smallest=1&unit=em&orderby=count&order=DESC&number=20');

		// This is what separates the tags by commas.
		// You can change the separator to any other character
		// The separater is wrapped in a designated class
		// to make CSS styling possible.
		$tags = implode('<span class="tag-separator">,</span> ', $tags);
		echo $tags;

		echo	$after_widget;
    }

    function update($new_instance, $old_instance)
    {
        return $new_instance;
    }
}

You can simply copy this code to the functions.php file in your theme, and it will create a tag cloud widget which you can add to your sidebar from WordPress’s widgets menu.

Make sure that you read the comments within the code. They provide more tips on customization.

More to come…

I think this tag cloud is already better than a simple list of categories, especially for a personal blog like this. But it could be improved even further.

For example, now users can click on a tag and see all the posts which belong to that tag. Wouldn’t be cool if they can select more than one tag and pull all posts that belong to the selected tags? That would require more customization and coding. I’ll post about it if I ever get to it

Related posts:

1. Clean Your WordPress Sidebar to Improve Navigation and SEO!
2. Fix Yet Another Related Posts Plugin (YARPP) Widget and Add it to the Sidebar
3. Give Posts Their Own Sidebar in WordPress

The domain name of DET contains the phrase “dubai employment”. I wanted to create a website that is search engine optimized for this key phrase. I learned the importance of this from The Thirty Day Challenge, which is a free course that got me to stop fumbling and truly get started in IM. They hold the 30-day challenge every year with renewed and improved material. I’m thinking of taking the new course soon.

The plan for DET was to first build a sustainable stream of traffic. I imagined that I could then throw any product in front of this stream and have it picked up. It was an accurate imagination, but it wouldn’t be that simple.

I created DET. I put some search-engine-optimized content on it. And I started getting traffic. It was time to start selling.

The First Product I Tried to Sell

What do my visitors want? Easy. To find employment in Dubai. So – obviously – my product needs to assist with that not-so-simple task. As someone who has experience in jobhunting in Dubai, I found the most valuable method to be networking with other professionals, hiring managers, and recruiters. So, I compiled a small list of contacts and wrote a guide to show how I compiled this list. I put this product up for $47. I knew nothing about sales letters back then. So, the product must’ve had a very bad sales page. I left it there for – maybe – a week. It received a bit of traffic, but I sold ZERO. I tried changing the sales page a little, but still no luck.

How can I position the product differently?

The nature of my product, which was simply a list of contacts and a guide, conveyed to the buyer that they still had to do a lot of work to get results. I needed a product that would make finding employment seem effortless. So I made a new offer:

The Second Product I Tried to Sell

Here’s what I proposed to my customers: upload your resume to my web server and I will instantly distribute it my list of “100s” of hiring managers and recruiters in Dubai.

I hired a freelance web developer to create the system to store uploaded files on my web server. And I had planned to distribute the resumes using iContact.

It has turned into a resume distribution service. The price? US$14.99! On 30th of November 2008 and only 20-minutes of this offer going live, I received the first order. And in the next two days, I received 3 more orders. Despite the low price, I was ecstatic. Aside from the one or two affiliate sales I had coincidentally made, these were my first ever sales in IM.

Time to Deliver the Product

As I said, I promised that I would distribute these resumes myself. I had a small list of personal contact information of professionals in Dubai and in the span of 2 days, I spammed them with 4 untargeted emails from a service that they didn’t subscribe to. Surprisingly, the majority of the recipients were happy! They replied thanking me and asking me questions about the service. But one of them complained – rightfully so. Here’s what he said:

Please STOP sending these – I have marked them as SPAM.
Kind Regards,

Obviously, the service couldn’t be scaled up. What if I receive three orders per day? It won’t work.

Transforming the Product

This experience has taught me something valuable. People liked believing that, to find a job, all they had to do was upload their resume and pay a small fee. I should keep this perception in whatever I’d be selling. But my model wasn’t scalable. What I needed to do now was keep the same offer, but modify the product a little bit. I ended up with:

1. A membership section where visitors can login and find contacts relevant to their industry in Dubai.
2. Resume distribution service where we send the subscriber’s resume to a list of corporate and business emails. (Note: I always gave owners of the emails the option to opt-out. More on that later)

The emails were divided by industries. They looked like info@corporate.com or a variation of that. It didn’t feel as bad spamming these emails as it did spamming the personal emails. And subscribers could still find personal contacts when they logged in to DET. The new price was $24.99.

The New Product is Kind of Scalable. But a Big Headache!

I guess now I could accept up to 3 orders a day. The list of emails we had was broken up by industry. That allowed us to send relevant resumes to relevant companies. It was valuable for job seekers, too. They saw the results and thanked us for having made their resume distribution convenient. But there was a big problem. The delivery of the product was done completely manually. We used iContact to send the emails to our list of companies. Every time we received an order we had to manually setup the email for dispatch. Once the email was dispatched, we had to manually setup the user account on our website. Then we would manually email the user a notification and a completion report.

This entire process took 25 minutes of intense focus to make sure that there were no typos in our emails and forms. We also had to come up with the cover letter that went into the email dispatch when the user didn’t specify one. We had to review the employment history of every user.

Every order delivery was a headache! After awhile, I felt that I’ve had enough. It got to a point when I’d flinch upon seeing the payment notification email from PayPal. So, in this sense, this model too wasn’t scalable.

Just when I made the final decision to pull the plug on this model, I received a complaint from a company that was on our list.

For your entertainment, here is the email exchanges we had:

Subject: Remove from your database – Dubai Employment XXXXXX

Please immediately remove any email addresses with @XXXXXXXX.com from your database. We are not interested in getting flooded with CVs and job applications.

Best regards,

John Smith

I replied

Hi,

Alternatively, you could’ve just clicked on the “remove” link at the bottom of any of the emails we send your way, which would have effectively removed you from any emailing list we have. And just FYI, all legitimate bulk email senders are required per law to provide that kind of removal link.

Best,

Mohammed al-Safi

He answered

I don’t bother about your removal link, I have never subscribed to your services.

Best regards,

John Smith

I replied

I’m just trying to tell you how a smart person that can read would go about this matter. You, however, choose to waste your time going to my website, finding my email and writing a message. With the education that I just gave you, you will now be able to just click “remove” and be done with it.

Mohammed al-Safi

You can see from my tone that I didn’t give a shit.

I thought that was the end of that. I already canceled the PayPal virtual credit card that I used for iContact. I didn’t want them to charge me since I wasn’t going to use their service anymore. But before they had a chance to find out that I canceled my credit card, they emailed me this:

Hello,

Sending for your account has been disabled due to a recent manual complaint we had received. These are weighted more heavily due to the lengths in which the contact goes to file the complaint. That being said we wanted to review your opt-in practices to make sure they are following the industries best practices. Can you please respond to the information below? Thank you for your cooperation and we look forward to your response.

* How were these addresses accumulated? What are the specific addresses of the sign-up forms?
Example: Online sign up form/paper forms.

* Where is the permission statement the subscribers agree to when signing up?
Please provide all tangible means of lead generation for us to view.

If applicable please provide the direct URL addresses or attach a copy of a paper form for us to view.

I told them

Hi there,

Sorry to have caused this to your service.

My contacts indicate their desire to receive my emails. However, they always have the option to simply opt-out of my emails per the mandatory removal link that you provide. I think the person who went through the trouble of filing a manual complaint had a personal issue with me, not because I spammed them.

At any rate, on the 23rd I stopped all my business activities that require email broadcasts and will not be using iContact again.

Thanks for the great service all along.

Best,

Automation is Obviously the Solution

Anyway, after this was over, I had to go back and think about the real problem. Orders were a headache to handle. If I could only automate the delivery of my product…

Actually, unlike how it’s presented in this post, I realized that I needed to automate the product early on.

Partnering with a Web Developer

As mentioned, to develop my website early on, I solicited the help of a freelance Web developer from Elance. He created the system that allowed users to upload their resumes to my website. In retrospect, I can see that he didn’t do much. But at that time, I was very impressed with his work. He created a WordPress Plugin…just for my site. Cool!!

Back then, I didn’t know about WordPress Plugin development, PHP, or JavaScript as much as I do now. But because I knew I needed to have access to these types of skills to achieve automation, I asked Alex, the developer, to partner up with me. He asked for 30%, I gave it to him – on the 12th of January 2009.

It was a stupid agreement because we didn’t know how much each one of us was going to contribute to this project. If I were to do that I again, I would follow this advice.

The advice says, to determine equity or share in a business between you and a partner, talk about it in this way:

Today, right now, your contribution is worth 5% of the company and my creation of the company is worth 5%. The other 90% is based on what each of us does over the next 18 months. Here’s a list of what has to get done, and what we agree it’s worth…

Brilliant.

My job was to handle everything there is to handle. Alex’s job was to automate order delivery. His job required good knowledge of how WordPress works, but, like many other developers, Alex was more comfortable working from scratch. I guess he didn’t like having to cater to WordPress. Three months passed and we weren’t even 30% done with automation.

Alex has a day job and a family to support. I’m completely devoted to DET. I’m married, too, but I don’t support no body.

The point is, our partnership didn’t work. Against Alex’s preference, we had to dissolve our engagement on the 23rd of March 2009.

Breaking the Shackles

I’ve always felt that I was unfairly held back by inability to materialize the concepts and ideas that I had.

Ah, that’s a very broad statement. Of course, our imagination can fly while our tangible abilities are restrained by the limits of realities.

But when it came to Web development, I felt I could break some of my restraints. Due to my education and past experience with computers and programming, I was comfortable with programming concepts. All I needed was to apply those concepts and theories to my real-world problems…

So, I dug up my old programming toolbox from the attic of my brain. I pirated books on PHP, JavaScript, CSS, and Photoshop. And I started making Web.

Learning this stuff took me about 3-months. During this period, I shut-down the paid services on DET.

I’m broke and I can’t find a partner to compliment my skills. The only solution is to make my own Web. Besides, it’s something I’ve always wished I could do. So, for the first time in a long time, I was determined. I put everything on hold and focused on this one thing.

On the 15th of June 2009, the paid services of DET went back live.

Complete Automation

Now my system is completely automated. Users upload their information, their accounts are automatically created, they receive automated emails with their username and password, and they get exclusive access to the premium section. All we had to do to complete the fulfillment was approve the uploaded information. Once approve is clicked, the system distributed the resumes automatically. The order fulfillment process went from 25-grueling-minutes to easy, less than 5-minute review and approval process.

During this phase, DET was making $700+/month. My target income was and still is $3,000/month.

Better Conversion Vs. More Traffic

Now, I have a much more scalable model. I can market the product and not worry about being flooded with orders. I had a decision to make: should I work on turning more visitors into customers or should I work on bringing in more visitors. I decided that I will work on the former – increasing conversion rate.

Improving conversion rate is something we can do without having to talk to anybody. Getting more traffic, on the other hand, required that we went out and talked to Website owners or we had to spend more money.

Redesigning DET to Increase Conversion and Profits

During the months I was working on DET, I was frequently being told that I should redesign the website – make it look more professional. I agreed with those comments. During that time, DET used a WordPress theme by Brian Gardner. The theme was Revolution I. I thought that if I redesigned the Website, I could increase conversion rate. So, I embarked on a new project. I called it DET 2.0. It was a complete overhaul of not only the design, but also the content and presentation of every page on DET.

From July 4, 2009 to August 7, 2009, I was intensively working on DET 2.0. The new DET used an excellent WordPress theme called Education. According to some people who were shown the new DET prior to its launch: it looks “credible.”

Great! That’s just what I want.

On the 8th of August 2009, a different DET was fully deployed. It definitely looked more professional, but it was more salesy and included lots of corporate speak. The old DET looked like a personal blog.

I thought the new professional design – and the other improvements we made – justified a higher price for the product. The price was bumped up to $67. Less than 12-hours later, the first new sale was made.

There’s a peculiar pattern in the way sales occurred on DET. Whenever I launched new sales copy or made a major modification to the product, I would make several sales in a very short period. Then, sales would dwindle…or worse yet…almost completely vanish. And that’s just what happened with DET 2.0.

I worked my head off for more than one month just so I could kill the income that DET steadily provided.

I dropped the price to $47 to see if that would make a difference. It did help, but still…the conversion rate has actually worsened. It was ironic. Just as I was expecting sales to increase, they shrunk severely.

At the time of this writing, I’m getting an average of one sale every 4 days. A dismal situation.

Should I now go get more traffic?

Obviously not. The website has become a complete failure. I should fix this mess that I worked hard to create. I should now really focus on fixing the conversion rate. And that’s what I’m currently doing. I’m working on new sales copy and making some modifications to the website. I will write about that later. Stay tuned.

Related posts:

1. A Website Struggle in Internet Marketing – Part 2
2. A customer asked me for a refund, I had to tell him to…
3. Best Way to Organize and Cloak Affiliate Links in WordPress