Tom’s Graphic Design Journal

Ok so maybe not a million dollars and once the idea gets done there will be many copy cats. So rather than go through all the hard work, I'm going to put my feelers out to see if it's something worth pursuing. Who knows, maybe someone will like my idea and go ahead and do it and then I won't have to bother coding it

As we know the biggest downside to advertising on the internet is that many of the ad programs don't give you the lovely back links that search engines thrive on. So your site doesn't necessarily raise in rank on search engines. Therefore the ad you ran might only be worthwhile for human visitors. Remember, there are two types of traffic, human and search engine.

There are text link exchanges, programs, and even just links to buy that do in fact help your site acquire back links that will help your site in search engine rankings, but they are often quite expensive.

My alternative solution is a dynamic link injection ad. Here's the catch, it can't use javascript. As we know search engines ignore that so they will miss the links. In fact, there are ads that already work this way. However, we'll be using PHP or some other language to make sure the links are followed.

How it works
1. A list of keywords/phrases are stored in a database with associated advertiser links (multiple links can be assigned to the same keyword/phrase actually and simply rotated or randomized)

2. Anywhere on the page where there's a keyword/phrase, a link gets created to the site.

3. Optionally, a javascript rollover effect can be created for further clarity...It would be nice to alert visitors that it is a paid ad they are clicking on, or let them know they are leaving the site before clicking (this is unseen by search engines but that's ok, the link is hard-coded to the page already).

That's it in theory.
In practice, a cost per click program can be used with this or many other types of payment. Monthly fees, by number of occurances, bidding systems, etc. Statistics can be recorded for tracking, etc.

Important note
This puts more relevant ads on sites as advertisers won't or can't advertise on a site that isn't related...there would be no (or not many) valuable keywords/phrases to use.

Another thought
Ads generally are shoved to the bottom of a page or the side and are quite noticably ads. People's minds automatically turn off and ignore ads. Furthermore, eye tracking studies have shown that we skip over content to the far right of our screens while reading. Not to mention ads are often invasive to visitors and disrupt the style and layout of web sites.

By including ads within the body copy of a page, we should have not only more clicks, but also more qualified clicks, due to the anchor text.

Anchor text brings me to my last point. Anchor text (the text that's highlighted on links) is important for search engines. Both parties benefit from good anchor text. This makes it possible for people to choose their anchor text which is even better.

Application
Of course this internet advertising method can only be valuable on certain sites. Blogs and news sites in particular, but any site with a lot of content. Simply because the chance for keyphrases/words to appear is higher on sites with a lot of content, but that's not to say any site can't use the ads, they just may not be able to charge as much for them.

Another idea is to make a plugin for a blog, such as WordPress, to take advantage of this advertising method.

I hope this makes sense to someone out there. I will probably code this one day, but right now I have too much on my plate. I think that this method really combines the best of two worlds. It's like a text link, but the links are not on some "links page" that has no page rank. It's also like a cost per click ad program if setup that way. Finally, it caters to both search engines and human visitors.

What's Doogle? Doogle is my custom Google search engine for graphic design and all things related. The name isn't original and the little logo isn't really either...but the point is to be to the point and easy.

I only wish putting it together was easy...at least half as easy as it seems or as the documentation tells you it is. Ok, ok, it is very easy to setup a custom Google search engine...but getting it to do what I want of course is never easy. The world just never seems to operate how I want...worse yet, the world rarely listens to me so then everyone suffers.

The one thing I dislike right now about Google CSE is the refinement system (it needs refinement). The refinement system is limiting and more specifically I want "blind" refinements... in other words, I don't want to see the stupid refinement label in order to use it. Yea, you can leave the title blank, but it still says "refine your search" and then leaves blank space. I want hidden labels.

You might think ok then the person needs to know to search "query more:hidden_refinement" YES but...what I'm then doing is making subscribed links pick up on key phrases to show urls for these blind refinements. This allows for a much smarter search that doesn't confuse the searcher with unneeded or irrelevant refinements that they didn't ask for. For example:

You search "autocomplete ajax tutorial" ... because you had the words "ajax" and "tutorial" in the query you get a box that says something like, "Search only tutorial sites." Then you click the link and it applies the refinement filter (via URL). Google actually does this with topics, go search "xbox" for example, but they didn't include this feature in their CSE. Easy. Simple. Smart.

Google's CSE has just enough pieces that the puzzle looks like it can be put together, but it's just a tiny bit short. Gives me a headache.

However, everything isn't bad. Google's custom search engines are in fact very good. The one I'm building is nice and helps me a lot. It's geared for the graphic designer and web designer/developer. I will be building a database system so people can submit their sites for indexing and specifcy "refinements" that go along with the ones I have. This will help me easily add sites (export xml to upload), alert me to dupes, and help keep Doogle very relevant. I am also figuring in a reporting system to help keep bad content out.

So go try Doogle already!

It was 70 degrees out today. 70 and it's January 6th! I've been meaning to get some more photos, but haven't had the time between all the work I'm doing. I do have some I shot a little while ago so I went through those, but only came up with two that I liked. That's just the way it goes sometimes.

It was a bit colder when I shot these, but the weather still doesn't feel like winter. Still no snow here. Maybe when I go to Stratton, VT, I can get some nice snow photos...if it's not too warm up there too!

Many people have some misconceptions about blogs and search engine optimization...and just SEO in general. I see it all the time. I can't blame anyone nor do I get annoyed because truthfully search engines do hide information from us. We can't know their algorythms, it would be cheating. We only know what we can research and that information has been fuzzy. I'll write about SEO tips later.

So what am I talking about? Blogs and SEO. Everyone thinks that a blog will help their site like some magic pill. True enough a blog can help since it increases the content on your site/domain. "Content is king" remember that. However, it depends on the type of content you write and how you set your blog up.

1. Links from other blogs to your site
It won't help your site's rank if you have a link in someone's comments. There's a "nofollow" on links there. Also, if the blog doesn't have search engine friendly urls, then even a link inside an article may not help too much. The actual article page needs to have a page rank assigned to it and if there's & % = etc. all over the url, then search engines have a hard time getting to it sometimes. Look at my urls, they spell everything out and use backslashes. aka "mod_rewrite" ... you want that.

2. "Throw-away" blogs linking to your site
A "throw-away" blog is when you make a free blog (WordPress, Blogger, etc.) and link to your site. This will not help you unless it's active because it won't have a "page rank" assigned to it. Stop wasting these people's server space.

3. Tagging sites
Tagging on Technorati, etc. only helps you bring people and not search engines (as much). What? You must think I'm out of my mind. Ok, try this...go to Technorati and go under a category tag...let's say "ajax" that's super popular these days. Being on the first page of results for ajax gives you a link from a page with a rank of 3. That does help. The subsequent pages all have 0 page rank. No search engine love there. HOWEVER, it does help bring you actual people to your site and have other blogs link to yours. So yes use tagging systems, but just understand how they are going to help you. In other words, you need content people want to link for them to help you.

4. RSS
Oh, boy...this is an under rated technology. I need to get more active with it, but syndicating content helps...but only if it gets syndicated. Just having RSS turned on isn't going to help your site magically. It needs to be used. What can you do? Well, you can hope people use it, but there are also sites out there that allow you to submit your feeds. These can in turn give you some back links that may get picked up by search engines.

Also, don't forget the blog design. If it's designed poorly for people or search engines, it can hurt you. While my current blog theme may not be XHTML valid, the design can be spidered and actually does have optimizations made to it. It's also very easy to read and I feel attractive. So I'm willing to trade up XHTML validation.

Yes blogs can help your site...and they can also be a waste of your time. You merely need to know what does and what doesn't help.

While I'm at it here, I figured I might as well share my two cents about dynamic text replacement. Or in other words, how I get all the pretty titles to show up on my blog automatically without creating images.

This is not a tutorial on "exactly" how. You can read some of the following sites and figure out how quite easily, the people who created these powerful scripts also include instructions. However, I wanted to review some of these better known ways as well as their pros and cons.

There are two popular ways dynamic text replacement is done. One is by using server-side software to create images from font files (there's several ways to do that) and the other is through Flash. All of these methods do require javascript though which may or may not be an issue for some viewers. Fortunately there will always be a default display so it degrades well and doesn't ruin things. This also makes all of these methods search engine friendly.

sIFR is the way through Flash. It is quite nice and gives you the most control I think. It will allow you to fine tune and adjust spacing between words and padding/margins all around which comes in handy...but not letters, there is no kerning with sIFR becuase dynamic flash text did not support it (sIFR version 3 apparently has it, but it's beta). However, you can control leading.

Another method is by using a javascript and PHP script by Stewart Ulm, an article can be found on A List Apart about his method. Another similar method would be by using Python. You can find out more about Peter Bengtsson's approach here. Yet another variation is a script done completely in PHP and CSS that is based off Stweart Ulm's script, found at joaomak.net. There is no kerning support (currently) for these methods either...yet

sIFR
Pros: Flexibility, anti-aliasing control, leading control, margin/padding control. Really nice looking because it uses Flash's engine for graphics (can technially use gradients or something for color).
Cons: Requires Flash, takes a bit of fiddling especially with odd fonts and text wrapping, larger base file size (but reuses swf file so in the long run it can be smaller), links can't use right click...to open in a new window, etc.

Stewart Ulm's Script
Pros: Fast (small file sizes..less than 5kb for a title you see here), easy to configure, doesn't require Flash to be installed.
Cons: Not easy (apparently it's somewhat possible) to kern/letterspace.

Peter Bengtsson's Script
The same pros and cons of Stewart Ulm's script...Just uses Python instead which may or may not be a con.

joaomak's Script
Pros: Doesn't require Flash or Javascript, valid XHTML and CSS, text/letter wrap (after page has been loaded...so if someone resizes their browser window), small file size (since there's no javascript and PHP is mostly server side, I'd imagine the only thing you're loading is very tiny images for each word...meaning this should be fastest loading script).
Cons: You can't specify class or id attributes on your text that you're replacing. That one hurts a bit.

One thing I didn't touch on (but hinted at) was transparencies. The sIFR technically could allow for it, so you could put your text overtop of an image background as you can with Flash movies...however, not all browsers support it. There is no (to my knowledge) support for the transparent backgrounds using the other scripts. I'm unsure if you can generate png images with the scripts...if you could, it would then be a matter of using the CSS filter property to get it working for IE6...but you then run down that road which has complications when you bring links into play.

sIFR is probably the most robust and best looking with the most options. I may in fact use it for my blog over Stewart's script. I'm not using Peter's because I'm not familiar with Python. Again, I personally did not use sIFR because of load times and because it doesn't allow for people to right click on links to open in a new browser window (and yes I have seen one attempt to fix this, but alas, it does not work on Firefox2).

I don't believe in any one "right way" for anything on the web. Any web designer/developer should know that. Even with "standards" there are just too many possibilities out there. The only two things that matter is what you can do...and what you can't do. Ultimately, it's your call...but don't go crazy with it because it will slow down load times if you were to replace ALL of your text on the page.