In the Summer of 1993, I taught myself HTML by printing out source code from the browser, and also printing out the actual web page. I’d compare the two closely enough that I wrote my own 8-page html guide that year in middle school. I must have been pretty proud of it because I still remember showing it to my aunt…who lived 3,000+ miles away. Apparently it, too, made the trip.
After a fantastic nine months of college, I leased a 14,000 square foot warehouse in Austin, Texas and threw raves and DJed progressive house from 2000-2001. I then headed out to Raleigh, North Carolina, got my start in SEO, and met some amazing folks like Andy Beal, Jenny Halasz, Garrett French, and many others who you likely know, or know of their work.
In 2001, I was the Senior Director of SEO at KeywordRanking/WebSourced, quite possibly the largest SEO agency to date, peaking at over 1,400 SEO clients. I was responsible for designing all of our SEO packages and keeping the 175+ person company up-to-date on SEO, making sure our clients were succeeding, etc.
In 2005, Andy Beal asked me to join him at Fortune Interactive. That was a very easy decision. As Vice President of Operations, within nine months of starting up, we had profits of 25% on a cash basis and 23 employees, more than half of whom reported directly to me.
After designing and executing multinational campaigns for Motorola, national campaigns for Lowe’s, and many, many others, I set out in 2006 to begin consulting and doing other things single guys in their mid-20s do. And I guess some things a few years later that maybe most guys don’t do, too, like…
After traveling the country with a friend for the last chunk of 2007, in a van we bought off of eBay (I sure miss that rickety thing sometimes), I began writing some code in July of 2008. I was in Boston for a month; the start of another trip around the country. By the end of that year, that (horrible, horrible) code became the first version of ontolo.
In the end, I simply wanted technology to automate all of the parts of the process folks said you couldn’t automate. Luckily, I’ve been stubborn enough to not give up on it, but have also been naive enough about the complexities of high-performance technology to not be so intimidated by it that I don’t try.
In the fall of 2014, I set out to learn everything I could about how to make my original vision of ontolo become real. I’d never been a very competent programmer (even simple concepts like object-oriented programming confused me and locked up my sense of logic), was generally incompetent at systems administration, and knew next to nothing about networking.
But that combination of stubborn and naive…sometimes that one works out. Not always. But sometimes.
Suddenly, programming began making sense. After diving into Perl for two months and putting together a prototype that was 10 times faster than anything I’d ever built before, I was ready.
Except there was another little thing that kept gnawing at me… I knew it could be faster and more efficient, even though I didn’t quite know how to do it. Or even if I could figure it out. I just knew it wasn’t as fast as it could be. And I absolutely hated knowing that.
I’d resisted going into learning what I needed to learn, over and over, week after week. Every friend who’s a professional programmer (except one) advised me strongly against getting into the the stuff I needed to learn in order to build ontolo.
But I couldn’t help myself. My obsession with speed and efficiency got the best of me. It consumed my mind and it wasn’t letting go.
So. Nine months later. And here we are.
ontolo downloads data faster than any other similar marketing service. And ontolo extracts hundreds of times more data from web pages, hundreds of times faster, than any other publicly-available technology.
(If you’re curious about the nature of this sort of problem – “article extraction” -, you can see here where moz developers managed to separate the main article content and comments from web pages at a rate of 70 web pages per second. ontolo separates out several-hundred sections of web pages, not just the main article and comment block. At the same time, ontolo parses individual elements like links, URLs, text, deeper text analysis, form data, script data, etc, and does so at a rate of over 4,000 web pages per second on a single core in a 5-year-old Intel CPU [Xeon 5650].)
So, after a number of years down a few rabbit holes and detours, ontolo is now where I’d always imagined it could be. In short: you now have the ability to analyze millions of web pages in just a few minutes (ontolo averages four minutes per million URLs downloaded, analyzed, and indexed), and and have stored in a way you can search instantly on the data to find very specific marketing opportunities.
The fundamental problem of market research has always been getting targeted data. But you need a lot of it. And the problem with a lot of data is the time it takes to go through it all. With ontolo, that huge time cost is now virtually eliminated. Research that used to take days and weeks can now be done, quite literally, in minutes.
And now that ontolo is here, a new vision is forming about what’s next to bring significant efficiencies to your marketing process…
I hope you enjoy using it as much as I enjoy building it.
– Ben Wills, November 2015