My Tech

…the truth is however that I have some very cool toys…

Ok finally we are going to talk about my innovations. You have had to endure me constantly referencing “my tech”,”my tech”,”my tech” and it sounds so pretentious that it makes me cringe every time I say it.

FACT:

Before I learned to code (Python), I used to use spreadsheets with formulas that were so long and conviluted that it was pretty funny. Gotta start somewhere.

An Edge

Obsession, and a complete inability to gracefully accept any type of defeat, was combined with some very lucky breaks at the beginning of my R&D efforts.

This opened a Pandora’s box full of opportunity and cost (For example my findings have been used to develop very sophisticated predictive models of various search engine mechanics. On the other hand I’ve received over 20 cease and desist orders for publicly sharing some of those findings.).

An element of paranoia can accompany the discovery of trade secrets and I have had some close calls with breaches and encountered some very motivated people hell-bent on getting at my stuff.
You learn. You improve. You innovate. You buy a gun.

So I keep my tech close to my chest.

I get down on a homebrew, multi-teraflop cluster that can operate within a totally closed network using my private index of the web as part of my training data.

My R&D efforts have produced some interesting utilities especially in the web marketing space. I’m not going to talk about everything, but here are some highlights of the technologies that I have developed.

This is my search engine modeling platform. It allows me to model an infinite number of scenarios and predict campaign performance with scary accuracy.

This system has its own internal link graph with trillions of links, a snapshot of the web with hundreds of billions of documents, a team of always-on web crawlers constantly indexing the web and a retrieval engine based on the technologies used by the primary search engines.

Extensive Text Analysis systems, text processing and interpretation; yielding highly valuable, actionable information.

My indices have all kinds of additional data as well. For example:

Local and community Visual Link Graphing that allows me to create data rich detailed visual maps of websites and analyze them based on the visual geometries that represent their architecture. This concept is not new. My system however allows you to mine non-salient information about the website. Presenting interesting opportunities for predictive measures of a website’s/communities life-cycle.

Some Details
Developed in many diverse languages. My preferred language is Mathematica but Python, PHP, C++ and others are represented in my work.

These tools and techniques have been developed and refined throughout my career.

All my innovations were born out of a need to have better data and more cutting-edge methods than my competitors. Creating novel techniques for reverse engineering search engine tech and so much more.

Mining and utilizing the very best Information Retrieval science and developing innovative approaches for applying that knowledge to the web as it pertains to my clients needs and my interests.

So much blood goes into my R&D. The sweat and tears could fill an Olympic swimming pool. When I look back at the orchestration of diverse resources. Crazy compartmentalization sequences. Forcing myself to learn these sciences from scratch (as well as the dependency sciences), their languages, the culture of their academia and then beating down their doors until I finally recruited help from these exceptionally talented men and women of a world normally closed to outsiders. It’s been a journey and I am very proud of what we have built.

•Social Media Object Projection System (currently in production): A system for predictive analysis of social media activity.

•Word of Mouth Marketing Automation Software: Designed for use in client physical location/office where customer interactions occur. It allows customers to do both automated WOMM and online review as part of their normal business transaction process.

•Access Probability (AXP): a novel KPI for SEO campaigns.

•Perception Index: A reputation management KPI that allows perception and opinion scores to be generated and evaluated for a given entity, brand, business, etc.

•Perception Engineering Platform (formula to production use): (Linguistic) Content Optimization System for maximizing impact, improving branding and adjusting public perception. Uses sophisticated predictive sentiment models, unique language databases, NLP, current events, colloquial data and proprietary algorithms to maximize the impact of branding messages, improve public perception and develop customer evangelism.

•Actionable Analytics Systems: By using advanced KPI’s and data analysis techniques to measure campaign performance against campaign goals; actionable intelligence can be distilled. Techniques include: proprietary social measurement like Buzz Index, regression models (linear, logistic, etc.), discrete choice models (visitor motivation, conversion motivations), identification of previously untapped responsive visitor segments by analyzing historical data using machine learning systems, visitor relationship management (usability, best practices, etc.).

And of course novel techniques for keyword, competitor, architectural research, etc.

That’s as far as I want to get into this subject for now. It is my hope that I have offered a little insight into my web marketing technologies. If you have any specific questions please contact me.

Why all the R&D?

To answer that we need to have a little context. You can explore this below.

Some SEO History and My Humble Beginnings

I have watched the evolution of the SEO industry since its inception in the late 90’s. By 2004 the problems were apparent. SEO’s were seen as snake-oil salesman by much of the business world. There were widespread irresponsible practices within the industry at that time and it gave us all a black eye.

In those early days SEO had a rough go. There was allot of mistrust. Results could be very hit and miss and the there was the ever-present threat of a major algorithm update which could annihilate your business overnight (anyone else remember the morning of the “Florida Update”?).

So I focused on finding a better way. I focused on the weaknesses inherent in the SEO world at that time and worked very hard to resolve them. The main weakness at this point was the volatility inherent in offering services controlled by search engines (which were largely a mystery in the SEO world).

My initial focus was understanding search tech from the underlying sciences (Information Retrieval) and then the software engineer’s perspective.

I developed relationships with IR pioneers in both the academic and private sectors. Slowly formulating techniques for discovering non-salient search engine features. The path I took focused on understanding search engines in an unprecedented way within the search marketing industry.

Today there are some great SEO’s out there. They have survived and thrived in this industry by sticking to the basic foundational search optimization concepts and evolving them to focus on higher quality and greater utility content. In the past creating keyword optimized content used to mean plugging your targeted keyword into every nook and cranny of your content.

Now optimized content is created by making the best, most useful example of content on a given keyword topic.

The Now

Fortunately SEO has evolved a great deal and there are lots of SEO’s doing great work. Some of them dive into search engine technology and have made some very cool finds and even cooler tools that have become industry-recognized essentials in the SEO’s toolbox. Other SEO’s have specialized in other areas that are not technical and they are also doing great work. It’s clear that there are different approches to effective SEO.

Over the years as search engines got better; it became less important for SEO’s to have a good understanding of the complexities and nuances of search tech. Developing outstanding, relevant, well cited content takes you far in the search results and elsewhere.

The difference of my SEO systems is centered around the extent, scope and direction that my R&D in this area has taken and the resulting advantages. It is to my knowledge simply unparalelled. The initial core technology that I produced to power SEO efforts took nearly 10 years to develop and is still constantly evolving. I spent that first decade of research with a very tight focus on good data, sound R&D, cutting edge IR, a couple of early lucky-breaks when I was first striking out on my own and very high standards of quality.

The result has been something unique and powerful.

It turns out that there is tremendous value in understanding those details. My collective web marketing technologies have some unique features and bragging rights that no one else can claim.

Bona Fides

de Apibus Ulva…