The bad

The comedian section was quite boring

The good

Opening titles were visually stunning. Mr Manhattan section was quite interesting, with nice phillip glass music

Conclusion

A bit disappointed because it had a lot more potential

2. Put some loud rock music on through your headphones. It will help with the repetitiveness of this task (there’s 2000+ sites to submit to, and most of them are inane)

3. Get your CAPTCHA patience going Hopefully you won’t be stuck in too many CAPTCHA loops

Deja.com indexed every newsgroup article ever posted, and was invaluable for finding a quick solution to a technical problem.

A third party component wouldn’t compile? Just search for the error message and class name, and in a few seconds you had a solution. Someone else was bound to have already asked your question, and got an answer.

Today, Deja still exists after it was bought by Google and was rechristened Google Groups (you can even still access it from the domain name deja.com).

But the level of posting is nowhere near what it was in it’s heyday.

Not everything was perfect with the newsgroups scenario. The main problem was the lack of control over who posts what (i.e. moderation), which led to increasingly annoying spam filling the newsgroups.

The solution that was adopted by the web in general was web forums. This largely solves the moderation issue, but creates another problem: indexing.

The wildly distributed nature of forums makes indexing harder than newsgroups.

Google web search attempts to do so, but there are several problems:

1. There is too much noise on google to actually find what you want. The spammers and scammers use every search keyword they can get to muddy the waters. Going through 50 pages of spam to find what you need is not fun. In order to find what you need, you need to actually scan through every synopsis. Some of these make you yell “Oh no! My eyes!”

2. Google is making things worse by being “loose” with it’s search phrases. If google had something like regular expressions, where you could nail down *exactly* what you’re looking for, things wouldn’t be so bad. When I was searching for “purposesoft”, I got a gazillion results for “Sony LCS-CST General Purpose Soft Carrying Case”. Why is google taking the liberty of adding a space in the middle of my search phrase? Adding quote marks or square brackets doesn’t make a difference. Even more annoying is the exclusion from the search of common words such as “to” and “a”. This makes winnowing much harder. Google engages in these keyword antics in the name of speeding up results. I often wonder if getting an instant results page full of crap is better than a slower search with higher quality results.

3. Google doesn’t index everything. No one knows the exact machinations of the indexing monster, but there may be limits on how deep it would index a site, or certain criteria for sites that are more likely to be indexed to maximum depth.

In the specific case of the Delphi developer, there are the Codegear forums (forums.codegear.com). Since I discovered these, I was happy to see that some of the old newsgroups regulars post there (TeamB).

But the Codegear forums are sort of slow and clunky. And the volume of posting does not seem on par with the old newsgroups heydays.

There are also search engines for forums such as omgili.com

Omgili seems to be a bit of hit and miss. It only seems to index certain forums, and it can be quite slow.

My ideal control would be exactly like Dreamweaver:

1. it would let you move images around

2. have templates for things like a centered layout or sidebars

3. have a html-code view and a design view (wysiwyg)

4. make it easy to embed images to send in an email

5. be a pure vcl and not an activex/ole/COM/whatever or encapsulation of MS IE/DHTMLetc

6. be able to open and render correctly existing complex html documents

At the time I narrowed the field down to

ProfDHTMLEdit by ProfGrid (ProfD)

and

HTMLEdit by PurposeSoft (HtmlE).

ProfD seemed to have the advantage:

1. it could open complex documents that HtmlE would choke on

2. it would render complex documents more accurately

3. if I recall correctly it made it easier to embed images than HtmlE did

But on the other hand, HtmlE had the advantage of being a pure VCL.

I had the sinking feeling that relying on an IE encapsulation would come back to bite me in the future, what with different versions of IE, different versions of Windows etc.

The whole thing was made moot by my procrastination binge which lasted almost 3 years.

Now that I finished my first product, I am back to the problem of choosing an HTML editor control for my second product.

I was just about to order ProfD when I came across these blog post:

http://blogs.sevensteps.com/2008/04/spending-lot-of-time-on-your-support.html

http://blogs.sevensteps.com/2008/12/when-to-take-matters-in-your-own-hands.html

In which ProfGrid’s support is described as abysmal (as well as the control being leaky).

Indeed, when I went to check out the newsgroups for myself, I found them totally broken in a sort of infinite-loop login screen (http://profgrid.newsgrouphosting.com).

Also, their “customer support” link on their web-site just takes you to Plimus’ vanilla “I didn’t receive my order” page.

This doesn’t inspire much confidence in the Russian ProfGrid outfit.

I then went to checkout PurposeSoft to see what progress they’ve made on their control in the last 3 years.

Alas, the PurposeSoft web site is not working at all. It just “hangs there” when you try to load their website and then gives up with a “New Session Failed” message (and has been for the last 2 days). Judging by the google-cached version of their site it seems like they are using some kind of Cold-Fusion type of funky authoring system for the website.

Does this mean PurposeSoft is out of business?

Or are they just having web site problems?

At any rate, this doesn’t inspire much confidence either.

It seems that gods have decided to make life difficult in regards to finding a vcl for this project.

There seem to be some other options:

1. learn to code around DHTMLEdit (The Microsoft Dynamic HTML Editing Component) myself. This still leaves me with the problem of not having a pure vcl and I have no idea what complexities are involved.

2. Julian Ziersch’s WPTools claims to have HTML editing capabilites, but their demo doesn’t really demonstrate it or inspire much confidence in the product.

3. Sergey Tkachenko’s TRichView seems to have some HTML editing capabilites but I am not sure how well it would deal with complex HTML documents

I am even thinking of abandoning the cherished Delphi 7 for this project and using the dreaded Visual Studio.

Does Visual Studio have better HTML editor components available?

In short, I am stuck, which is always bad as procrastination loves feeding on scenarios such as these.

While Adwords is nice, if your product is not expensive, you don’t have much margin for the CPC.

And besides, many people don’t click on the ads and only click on the organics.

And even more, Google doesn’t show your ad all the time. It chooses to show the most profitable ads most of the time.

So, if your page doesn’t come up first or somewhere near first on the Google search results, your business is dead as a dodo.

It doesn’t matter how great your software is, or how compelling your web site is if no one knows about it, right?

Marketing is a problem for programmers.

As a programmer I just assumed that if I made a neat product, the customers would just magically appear out of thin air.

Classic mistake.

So, I need to either learn how to SEO myself or pay someone to do it for me.

A google search returns gazillions of businesses that would be happy to SEO your site for you (and also other, even fancier acronyms).

Some are expensive and some are ludicrously expensive.

Another search reveals a difference of opinions between mISVers regarding the value of such services.

Some say it worked for them (as always, one must be wary of Sock Puppet Marketeers -SPMs).

Others say it’s a waste of money, and may even get you ‘blacklisted’ by google if the SEO outfit used some black hat techniques to promote you.

The services they seem to offer are:

1. locating the right keywords for you

2. helping you make a landing page for each keyword

3. doing a link building campaign

At the moment I am totally confused about what to do. Part of me is leaning towards paying someone to do it and forget about it.

The problem is finding the right company, which would strike the right balance of being affordable and doing a great job.

On the other hand, learning how to do this and becoming good at it could be a great opportunity for me, and become a real competitive advantage.

That brings us to the second problem.

Actually understanding what SEO is.

From what I was able to glean so far, SEO is:

1. finding the right keywords people are using to locate your type of product. You can be aided in this process by looking at your competitor’s PAD file and seeing what they put in the keywords field. Let’s say you settled on the following search keywords: Bravo Sierra

2. buy a domain name that incorporates your keywords i.e. bravo-sierra.com

3. make landing pages for each keyword and incorporate the keyword into the filename i.e. bravo-sierra.com/bravo-sierra-blah-blah.html

4. use hyphens to separate keywords, not underscores

5. put your keyword in the “<title>” tags: “<title>Bravo Sierra blah blah</title>”

6. put them in the H1 tag: <H1>Bravo Sierra blah blah</title>

7. put them in paragraph below

8. do a link building campaign. This is where I get stuck. I read that you need to get links in quality sites, not rubbish directory sites.

This is where the details get sketchy on those how-to-SEO tutorials.

How do you get links on those quality sites? Send an email to the site owner? They get bombarded with requests and often ignore link requests.

How would those SEO companies do it?

Use social bookmarking such as Digg? Many of these Digg articles do get Dugg and sink to the bottom.

Issue a press release?

Create a blog and get it crawled by engines such as Technorati?

I get confused by the blog idea for this type of marketing. Say you do manage to get people reading your blog. How does that help get your product page up high on google results?

I understand it will get your blog up higher on google, but surely not your entire domain? Does google lift the rank of your entire site based on the popularity of the blog?

Or do we rely on people arriving at the blog as a landing page, and then using the navigational links on top to get to the product page?

Or do we shamelessly plug the product inside the body of the blog?

The other concept I’ve seen hyped up is ‘fresh content’: i.e. “google loves fresh content”.

“Fresh content” means writing one article every day and slowly building up google’s affection.

My question is this, how do you come up with enough Bravo Sierra to fill up an entire article every day that would somehow relate to your product and attract Diggers or other readers?

The other advice is “just build great articles and people will come” and “write about what people want to hear and they will come”.

According to Google Trends, people want to hear about (i.e. “Volcanic” keywords):

1. ickett (whatever that is)

2. orson charles

3. shock sites

4. benjamin bronfman

5. danica mckellar

6. vanessa rousso

Now, I have to say that I am quite proud that I don’t know who any of those people mentioned above are.

It means I’ve done a good job keeping the mental junk food out of my brain.

So, how do I write about what people want to hear about if I shun the tabloids?

Tongue in cheek aside, I understand that what that advice means is: write articles that are specific to the software you are trying to sell or within that field of interest, and include the keywords in the articles.

But how do you come up with that content and keep your integrity?

How would these SEO companies mentioned above create these articles? Or would they just tell me to do it?

I have done many of the SEO things mentioned above, but haven’t seen much results so far.

Meanwhile, my competitor comes up as the first result for my keyword.

Does this mean I have to wait several months until Google digests my efforts and things get reflected in the search results?

I feel that answers are forthcoming.

Setup seemed to be going great, and the general feeling was ‘my ad will be online within a few minutes’.

After everything was set up, I sat back waiting for the ads to come up.

After waiting a few hours, I went back into Adwords to investigate what was going wrong.

Buried somewhere inside the Campaign management functions is a magnifying glass that shows you what is going wrong with your ad.

The magnifying glass was telling me that “You have reached and exceeded your budget”.

I was thinking, “Google must have shown a bunch of ads so quickly that I didn’t get a chance to see any. A lot of people must be doing searches on my keywords. But, why is the impressions counter on zero?”

A gnawing suspicion was telling me that whole “exceeded your budget” thing was a funky way of telling me that my account was under review and waiting to be approved.

Then why not tell me that in the error message, Google?

A search of the support groups discovered that this has been an ongoing problem for several years!

To think that a company the size of Google couldn’t get around to fixing a simple error message for this long, and was just now getting around to addressing this in their beta version was incredible, to say the least.

The poor people who regularly post on those forums have been acting as auto-responders to the daily question of “why has my budget exceeded”.

I’m seeing a pattern here of large companies ignoring recurring issues like this for no apparent reasons (see in earlier posts regarding Hostgator and Plimus).

It took about 10 days until my ads started showing, and so far I made one sale as a result of an ad.

I’m not sure if the numbers are going to work for as I’m paying quite a bit per click and my product doesn’t have much margin for this kind of advertising.

I may have to put the price up for my next product to leave a healthy margin for Adwords.

I’m also troubled by the thought that many people ignore the ads and only click on the organic results. I know I’m not paying for impressions, but still, I’m missing out on many referrals.

I luckily found a post mentioning the ‘-free’ and ‘-freeware’ negative keywords. These tell Adwords to filter out people looking for free products.

Google should be showing a message advising people to put those two negative keywords in, if it wants to live up to it’s image of an honest corporate citizen (or at least add them automatically to the keyword list).

Yes, that would cost them some revenues, but leaving things the way they are is a little bit scam-y.

To my delight, I made the first sale a day after releasing the product (I believe it was a referral from Tucows).

At this point I wasn’t checking the logs to see where the customers were coming from and haven’t set up Google Analytics yet.

The second sale came from an Adwords referral a few days later. The Pay per Click commission ate up a big chunk of the sale amount, but it still felt good to make another sale.

Later, I would regret choosing Plimus and being so hasty in choosing a processor.

Everything seemed to be working fine, the setup was straightforward and I chose the option of uploading a list of license keys to be dispensed by Plimus (less chances of something going wrong).

Then I made my first order, as well as letting my friends and family know they can put some orders in if they want to.

To my horror, all the orders we made were put into Plimus’ Under-Review status. Which means every buyer has to receive a robot Callback for verification.

In all the years I’ve been buying stuff online I never had to go through this process as a buyer.

The worst part of it is that the robot does not call immediately. It may call several hours later.

Which means you have to sit by the phone and wait for the call.

Because, if you miss the call, it may not call again for several days (if it’s the weekend, especially).

A few of the family members that ordered the product got a Order Rejected message because the robot was not able to get hold of them (3-4 day after making the initial order).

None of the people who ordered received their key sooner than 2 days after ordering.

I then did a search online on ‘Plimus Phone Verification’ and found people complaining about this in 2005! 4 years later, and Plimus still hasn’t done anything about this.

This is a classic case of a company not listening to it’s users needs and generally not caring at all.

I then agonized about this for several days. On the one hand, I was told this was only done for the first orders and then verification would be turned off  (except for suspicious cases, which seems to include anyone not in the US).

So part of me wanted to wait and see if things would improve (especially since I had already done the work setting things up with Plimus and didn’t want to go through the whole process again with another service).

On the other hand, I had no idea how long this ‘probationary’ period would last. And I had gnawing suspicions about the quality and the general philosophy behind a company that would let things go on the way they are since 2005.

I just couldn’t decide. I was stuck.

Finally, I made up my mind to switch to another service.

I can’t stress enough how liberating that decision was. Just making a decision, any decision, was better than procrastinating and agonizing about it.

I switched to Fast Spring (my other option was 2checkout.com), and it was quite easy to switch.

Luckily Fast Spring supported the same method of uploading a list of keys, otherwise I would have had to make changes in my program.

Fast Spring also quickly set up an order page that conformed to the layout of my website withing a couple of days which was a bonus.

They don’t seem to have Plimus’ infamous chargeback fee as well, which is a plus.

A minor complaint about Fast Spring is the ‘upload license keys’ function which seems quite slow and only able to handle about 5000 keys at a time.

And I wish that they supported showing the license key on the ‘order completed’ web page, instead of only supporting it via email.

The email-only situation, leaves the user hanging there at the end of the transaction, waiting for the email (although Fast Spring seems to be very quick at sending that email – within a minute or so).

This could be a problem with email spam filters.

All in all, I would recommend Fast Spring at this point in time.

I’m using CS4, and I find the filters and effects really enjoyable to use (especially the glow and shadow effects).

Dreamweaver does a good job of synchronising everything to the host server. Although, I did notice that some of the site templates have disappeared since CS3 or CS2.

I’m using Hostgator as a host (Baby account), which I found quite affordable and quick to setup.

I began submitting to sites using the good old manual approach (later I will find out about the wonders of PAD files). While submitting to SoftPedia, I found out they require a secondary download link.

I scratched my head a bit and then decided to make the secondary link an FTP link, as the CPanel in hostgator had an Anonymous FTP option.

I enabled Anonymous FTP, but when testing it, no files would show in the public FTP directory.

After wasting a lot of time trying to get this to work, I called Hostgator and found out I have to upgrade to Business account to get anonymous FTP.

I asked why wasn’t I shown a message in the CPanel about this feature not working in a Baby account (there was no error message when I enabled Anon FTP). I was given some excuse as an answer.

But in general, I am happy with the Hostgator service.

I made the classic mistake of only putting images on my front page (getting carried away with Photoshop and embedding the text in the images), thinking the ALT tags would be enough for the search engines.

To my chagrin the search engines ignored the front page as a result.

I then had to add some text to the front page to get indexed.

Now, coming up anywhere near the first result page in google for my chosen keywords is another challenge altogether. More about SEO in another post.

PAD files are a real time saver for submitting to many download sites. After doing a few sites manually, I wished I found out about PAD earlier.

I also submitted to the ASP database (I wonder how many download sites crawl this database).

I quit my job as a Delphi developer in 2006 intending to make it big in the mISV world.

Since then, almost 3 years were lost due to procrastination and World of Warcraft. After building up a hefty debt, I managed to pull myself together.

These two factors helped me take control:

1. Steven Pressfield’s Book: The War of Art – highly recommended for people in the predicament I was in

2. The pressure of the building debt and the prospect of the money drying up (coupled with inability to get more loans)

3. I decided to scale down my super-duper first product, and release something more modest as an incremental first step kind of approach to it

I have now finished the first product and am getting ready to launch it.

Some background about the project: It is 2009 and I am still using Delphi 7. There just doesn’t seem to be much impetus to upgrade from the cherished Delphi 7 which has proven so reliable on so many projects.

I am using the excellent Developer Express controls, which have saved the day more than once in the past. Can’t say enough good things about them.

For an installer, I am using the fine Inno Setup (kudos to Jordan Russell). Don’t get me started about the horror stories I endured using the awful InstallShield in the past.

That’s about it for this first post. I hope this blog is of some use to someone walking a similar path.