Ajax Libraries

In 2001, I was exploring a barren wasteland trying to figure out how to load data through http calls without reloading a web page. The pickings were slim, and about the only things you could clearly determine was that Microsoft was the only one regularly doing it in a broadly deployed app (their Web interface to Outlook), and Brent Ashley was the only one talking about it (remote scripting) in any intelligible fashion, with his iframes JSRS implementation. Titillating little gems, like the Oingo semantic web dropdown menu that changed as you typed appeared and disappeared like gossamer threads--gone before you had a chance to learn anything.

Here we are in 2006, and now everyone wants to do it. Google showed us the way with the Google Suggest tool, then Google Maps. Libraries started popping up like OpenRico (from a Puerto Rico airline of all places!), helping to bring Ajax to the snowballing Ruby on Rails. Google Map envy made the browser come into line, adding HTTP objects accessible through DHTML after the page load is complete. Now, even Safari on the Mac is supporting it, and Google Maps works their as well.

When I started MyLongTail, I had to choose a library. The implementation I chose was OpenRico, because it was the one Ruby on Rails had chosen, and I wanted to hitch my wagon to a rising star. I needed an implementation that would stay around for a long time and have broad browser support. But OpenRico co-founder Bill Scott joined Yahoo! Richard Cowin is still doing OpenRico. And OpenRico now supports Safari, but not in manipulation of HTML fragments in the DOM object, but rather in setting values within objects. It provides a much more JavaScript-dependent programming environment that I would like, but rearranging HTML fragments has broader browser dependency problems that makes it worth staying away from. For example, in FireFox, if you manipulate add new form widgets to a form, their values will not be included on the submit!

So in general, when looking for a good Ajax library, you have to expect that one of the most important things to do--altering form widgets based on actions taken in other widgets--still just isn't compatible at the browser level and there's nothing a superior Ajax library can do about it.

But that being said, my choice to go with OpenRico might not have been as smart as I originally thought. A closer inspection reveals they're checking for MSIE vs. FireFox compatibility, but are not worrying about being broadly compatible. Google just released their Ajax development platform. I'm going to re-examine this decision.

Previous Next Arrows Vs. Previous Posts for SEO

So, if previous/next arrows with keywords are so influential for SEO, why am I not using MovableType or WordPress that support them, and instead using Blogger, which doesn’t? The answer is in the “Last 10 Posts”—or in the case of my Blog, “Previous Posts” links. It’s always ONLY 10 posts. So the voting power within the page is diluted on a per-link basis. Each link you construct on a page is casting that much less of its Google-juice voting power to other pages. But because EVERY permalink page has a similar “Last 10 Posts”, and the list is slightly different on each page (by one link), it’s a zero-sum game. It all evens out, and each page is being proportionally EQUALLY buoyed by its neighboring pages as if it were the less-diluted version of prev/next arrows found in MovableType and WordPress.

SEO Secret Sauce: The Prolific Writer

Not long ago, I was a prolific writer—when I was not expected to be. Now that I HAVE to be a prolific writer, I find it more difficult. First, with Prophet 21, before it was acquired by Activant, I delved into SEO and was one of the most active posters on the then-dominant Search Engine Forums (SEF). It has since been replaced by Webmaster World for the geeked-out in-the-know SEO crowd. Everything was new then, I wasn’t in the SEO business, and I felt compelled to share—in part out of altruism, but also out of ego, and to stick it to the SEO secret-holders. Anyone holding his cards too close to his/her chest is a bullshitter, I felt. So, I said it like it was, and was one of the first people to lay out precisely why keyworded previous/next arrows were disproportionately effective in SEO—they pander to Google’s PageRank algorithm WITHIN the site. Since, and perhaps consequently, both MovableType and WordPress have instituted keyworded prev/next arrows.

Next, I decided that if I was such hot shit at SEO, I would put my money where my mouth was and go for a slice of the pie. Instead of affiliate programs, I opted for a straight cut of worldwide gross revenues of a technology company in an emerging market—digital signage. I hitched my wagon to the Scala star, and rode it for about four years. I ran into an unfortunate sysadmin during this experience, and never really received the programming support I needed to take my work to its next level. So, while I made some very nice money proving my concepts, I could never mainstream it into the next big thing as I had planned. And this was still back circa 2001, right during the Bubble Burst, and before everyone woke up to the significance of search and the power of arbitrating other people’s traffic. But I was a prolific writer at Scala, relentlessly blogging on the topic of digital signage much to the chagrin of Scala’s competitors, such as WireSpring and WebPavement. I remained prolific until August 2004, when I joined Connors and NEEDED to write.

So at Connors, I focused on delivering my secret sauce to Connors clients-only! I got a touch of the paranoia that I humorously observed in my SEO counterparts while at Prophet 21 and Scala. And I started thinking whether I could ever really write again about the secret sauce. Webmaster World was the place to be to impress peers and feed one’s ego. But building influential personal blogs such as Battellemedia and Micropersuasion was the way to build audience and attract new prospective clients. But what could I write about? This was also a problem with speaking opportunities. I was such a back-room SEO-guy that all I had to say was stuff that gave away the family jewels—tidbits like the power of prev/next arrows!

Meanwhile, things are changing at an ultra-rapid pace. Ruby on Rails comes along, and the concept of an agile framework is on peoples’ radars. There goes the special edge of my “generalized system” (where apps spring into being virtually before the product spec is written). Next, Google buys Urchin and makes Web Analytics free. There goes the possibility of selling what I built as an analytics package. The world is catching up with me on all fronts but one: how to know WHAT keywords to optimize for the biggest return. And along comes Chris Anderson with his fateful 2004 online article about the long tail distribution curve and how it applies to search.

The SEM people (those who sell clicks) rapidly jumped on the bandwagon, and said “Hey, up the keywords! It doesn’t cost you any more, and it may help”. This is a good premise, but the nuance is that it works even better when you don’t pay a cent! That’s right! Writing about an obscure but promising topic and putting the content in optimized format on your website simultaneously optimizes for ALL search engines. You don’t have the overhead of another keyword in your campaigns to manage. You don’t have to pay if a click occurs. And it’s a permanent addition to your website, working for you 24x7.

Two metaphors come into play: the snowball effect, and the iceberg principle. I will probably write about both of these separately because they are such important topics. But the snowball effect states that if you keep adding just the right content to your site at a consistent rate, your site will increase in effectiveness as it grows in size and search-appeal. It’s “a-peelin’ off a layer of someone else’s search traffic.” At any given moment, there is a fixed amount of search traffic occurring on the planet, and we’re in a battle of capturing it before the next guy does.

Secondly, the iceberg principle states that if someone searched past that magical 3-page limit and STILL found your site on a given keyword (or keyword combo), then you MIGHT just be seeing the tip of the iceberg. If you could somehow tweak your page to the first page of results for that keyword in the search engine result pages (SERPS), then you might receive tons more traffic on that term. Some determined searcher actually volunteered competitive intelligence to you through their determination and dropped it in your lap to use or not. MyLongTail is about using it! It is a powerful tool in the hands of a prolific writer.

And now that Connors Communications went public with its secret sauce for SEO, and I’m free to talk about it—watch out! The prolific writer is back. And just as I spelled out the power of prev/next arrows with keywords, I’m going to be blowing the lid off of as many closely held techniques of SEO that I can, in order to level the playing field and create some opportunity for Connors. So, what opportunity is there once all the “process” secrets of SEO are disclosed? Simple! They’re not easy to implement—especially at large companies. The old 3 c’s of the Internet used to be content, community and commerce. The new 3 c’s of SEO are consensus, clarity and confidence. In other words, large companies to stand even a minute chance in this blogging world of ours need to build consensus between the marketing and IT departments. The projects that are going to be carried out need to be spelled out in crystal clear fashion (clarity), and everyone—especially the IT department, needs to be completely confident that it can be done without breaking everything or cost sultan’s ransom to overhaul infrastructure.

And I know the prolific writer is back, because I have to discipline myself to stop there. I really could go on forever.

Long Tail Keywords

So, MyLongTail is finally undergoing a soft launch. My plan is to get a few thousand users on relatively low-volume sites. People who already have high volume sites will hardly need the extra boost that MLT provides. Its people who are just establishing themselves that will benefit most. People launching new websites will actually have the ability to watch every search hit as it occurs! In a way, it’s sad, but it is also a strategic advantage, because MLT will inform you if it needs further refinement—the long tail keywords!

One of the important principles of MyLongTail is that it does not work in a vacuum. An existing site with a little bit of content built up is a perfect starting point, because MLT will inform you of keywords that are on the edge of breaking out. But if you think you should be coming up on a diversity of terms that you don’t, you have the chicken and the egg problem, and it is time to float a few test balloons.

There is no easier publishing mechanism to float test balloons than Blogger. The beautiful part of Blogger is that it can plant your entire blog into the subdirectory of an existing site, if you have FTP access. The advantage there is that the snowball effect you are trying to achieve is on a per-domain basis, usually taking into account all three levels: www.domain.com. So, using TypePad achieves little for a corporate site, because you can’t plant it in www.domain.com/blog. You can with MovableType and WordPress, but you would need an unusual level of access to the corporate webserver. Better to FTP it into place.

Test balloons usually entail taking a writing recommendation under the Suggestions tab of MyLongTail, and only investing a half-hour to an hour of writing time on that topic. One of the dirty secrets of SEO is that your main objective is achieved merely by using blogging software and getting your title tag correct. You can almost leave the page blank and get SOME results. But it’s better to put out a quality page with quality content. The reason it’s a test balloon is because it will result in MORE suggestions being issued in MyLongTail. There will be word combinations you never anticipated, with a word from the top of your page combining with a word on the bottom of your page, and some very determined searcher going many pages into the engine and finding you. This registers with MyLongTail, and begins to solve the chicken and the egg problem.

For those people who have been lucky enough to discover the soft launch of MyLongTail, feel free to get started. Put the MLT code on every page of your site through the template system. With Blogger, it is a breeze. Just copy and paste it as-is into your template somewhere between the body tags—but not between special blogger tags. Seeing the results should be almost instantaneous. Take the first reasonable suggestion, and make a new blog post base on it.

Writing for SEO

Have you ever gotten into the mood where you are just a writing machine? If so, you are one of the most valuable people in marketing today. Just about anyone can plagiarize off the web, and change a few words here or there. And low-priced outsourced employees can put target keywords together in new sentences and combinations for AdWord campaigns. But the plagiarized pages will eventually be filtered as spam, and the AdWord campaigns will eventually stop running. Only the original content, containing valuable and preferably timeless information will continue.

Such the pages that achieve the largest double-whammy objective: acquiring new links. New inbound links that occur spontaneously without link-begging or link-exchange are like gold. They play to the Google BackRub formula, but they also leave a trail of references that actual Web Travelers can follow back to your site; search engines aside. As long as they are not reciprocal, and you’re receiving links from an average distribution of people around the world, your site will be achieving this critical objective in the most desirable organic fashion.

And the whole process begins with writing—and writing at a decent rate. Why? Because search engines are biased to consider newly discovered content and float it temporarily as a test balloon. That means new content always receives a temporary boost, and explains why new site launches/re-launches are so often accompanied by unsustainable natural search gains. The effect wears away in a few months. Google’s patent application from March of 2005 showed us that Google actually considers the “delta” information between one full site crawl and the next.

New writing goes into this set of fresh content. And if you’re publishing with blogging software, you have the additional boost of the new post ping system and the whole additional set of crawlers and news feeders that it brings. This is part (but not all) of the reason blogging software brings an unfair SEO advantage over crusty old content management systems that never had search friendliness as part of their criteria. And a company who hires a professional “writing cannon” as member of their marketing team will be very well served. Their work, instead of just transient press releases, will become a permanent part of the tapestry that is the Web, and their work will be put to work 24x7 generating new prospects.

Google Trends

Until today, Alexa has been the place to go to get a semi-authoritative and free gauge on traffic. But with the release of Google Trends, there is another must-see site. It is along the lines of BlogPulse, but much more authoritative, because it is tied into the most used search box in the world. It is however interesting to note that it has a similar problem as Alexa, in that all data is relative (not absolute). In other words, they’re not giving actual traffic numbers. They’re just comparing relative search volumes across times and between words. This is so similar to Alexa’s relative rankings, and I cannot help but feel is a general strategy that they use to keep themselves from being called to task on their numbers.

Whenever you’re looking for hard and fast numbers on Web traffic, based either on keywords or on site visitation, you have to turn to such companies as comScore and Hitwise, who each have dubious methods. comScore essentially runs software on enough individual’s PCs to get an average cross-sampling. Alexa too runs special software in the form of their toolbar. comScore has the superior method of cutting deals with ISPs to get a direct and pure sampling, but potentially skewed by region due to ISP service areas.

Alexa is the only one so far that has turned around and put its data out for popular general use. But their numbers can be challenged if put in absolute terms. The solution is to make it relative, and I am never done with explaining their Traffic Rank, Reach and Page Views. Of these numbers, Reach is the funniest. They call it reach “per million viewers”. They are essentially saying “parts per million”, measuring traffic that visits a site as one would measure ammonia levels in a fish tank. The trick is, you can’t tell how much ammonia is in the fish tank without being given the total volume of water. So, you can’t know how much traffic is reaching a site in absolute terms unless you’re given the total number of Internet users.

Similarly, while the Google Trends tool is amazingly useful in comparing keyword traffic on a word today versus last week, or between keywords A and B, it’s not telling you traffic on a keyword in absolute terms. This is still a process of piecing it together from the clues.