Note: Please see my follow-up post which sheds more light on what InfoWorld knew and when, including an email thread that proves the publication’s complicity in the Craig Barth ruse.
Public falls from grace. We all love to watch them unfold. Whether it’s a golfer with libido issues, or some blowhard blogger getting his comeuppance, we just can’t get enough of it. The sordid details. The backroom double-dealings. The questionable motives.
I, of course, I fall into the latter category. I am Randall C. Kennedy, former internet “shock jock” blogger for InfoWorld and current holder of the title “Most Reviled Person on the Internet, 2010 Edition.” In the past 72 hours, I’ve been humiliated, chastised and kicked to the curb by virtually every one of my contemporaries. My personal and professional credibility is shot, and my part-time career as an IT journalist is over for good. Can the urinal cake with my face on it be far behind?
Still, like every good tabloid story, the villain still wants his day in the sun - a chance to tell his side so that the record is truly complete. And while the future may see my name relegated to the role of punch line for a crude party joke, it wasn’t always this way. I once had a name I could be proud of, one that was associated with highly successful projects at some of the biggest firms in IT and finance. That it could all come crumbling down so quickly should serve as a cautionary tale for anyone in a similar position. So here, without further ado, is my story.
I’ve been a professional in the IT industry for over 25 years. I got my start in the mid 1980’s pulling wire and installing servers for a Novell Gold reseller in Southeastern Massachusetts. It was there that I cut my teeth on technologies like NetWare, LAN Manager and SCO UNIX. And after 5 years of often grueling work in and around the Bay State, I emerged with a strong appreciation for the difficulties faced by those working in the IT trenches.
My next stop was also my first real gig as an IT journalist. The year was 1993. Windows Sources magazine was about to launch as a new Ziff-Davis publication, and Editor-in-Chief Gus Venditto was looking for talent that could write authoritatively about Windows-related issues. I was brought on as a Contributing Editor – along with John C. Dvorak and others – and carved out a niche covering Windows data communications, among other topics.
IBM Comes Calling
In 1995, after two years of writing for Windows Sources, PC Computing and some extensive work at ZD Labs, I was approached by IBM about doing some consulting work for their Personal Software Products (PSP) division. Jay Sottolano was an acquaintance from the trade show circuit, and he was looking for someone to help write positioning papers and other collateral in support of their OS/2 marketing efforts. Knowing this would signal the end of my career as an IT journalist (back then, the industry frowned on such conflicts of interest – now writers just “disclose” them), I took the leap anyway, forming my first corporation – Competitive Systems Analysis, Inc. – with my new wife as my business partner.
Together, we spent the next year travelling the world on IBM’s behalf, giving stump speeches to the PSP and PSM (Personal Software Marketing) faithful and providing competitive marketing advice to the company’s Software Solutions Group (SWS). Along the way, I got the chance to brief a number of high-level IBM luminaries, including CEO Lou Gerstner and CFO Jerry York, and I was also fortunate enough to work with some exceptional executive talent, including Alan Fudge. It was a heady time for a young professional barely out of school, and I did my best to make the most of every minute.
Settling Down for a Spell
But, eventually the rigors of nonstop travel and feast/famine contract cycles – plus the arrival of my first child – prompted me to seek out a life with greater stability. So I did what most consultants do at this juncture in their lives: I bought a house in the Bay Area (Danville – finest town in the USA, IMHO) and got a real job. Specifically, I took a position as a Senior Industry Analyst with Giga Information Group (now part of Forrester).
It was circa 1997, and I spent the next year working closely with some top notch IT analysts, like Richard Fichera, as well as with more than a few egotistical blowhards (I’m talking to you, Rob Enderle). I also got to spend some quality time working with Gideon Gartner, the legendary founder of the Gartner Group. For whatever reason, Gideon took a liking to me, and I was able to learn a great deal about the inner workings of the IT research industry under his sage tutelage.
But the truth is, I was bored at Giga. The type of content we were asked to produce – dry, color-free analysis of IT trending minutia – was taxing to produce. I missed the chance to get my hands “dirty” working with technology directly, especially Windows NT, which was my first true love. And it was during this time that the pseudonym, “Craig Barth,” emerged for the first time: As a pen name I used while moonlighting for Windows NT Magazine as their News/Analysis Editor.
Note: Contrary to popular opinion, I am – and always have been – a huge Windows NT fan. I was one of the first journalists to jump on the Windows NT bandwagon, even going so far as writing a book promoting IT’s “Migrating to Windows NT” (that’s the title – from Brady Books - look it up) in 1993. In fact, by the time I left my first gig at Windows Sources, I was writing the “Windows NT” column for them on a monthly basis. And, of course, the potential synergy with Windows NT Magazine was a no-brainer.
Intel Comes Knocking
So when Intel Corporation came knocking in mid-1998 with an offer to work with them as a performance engineering consultant to their Business Desktop Marketing (BDM) group, I once again took a leap of faith. I resigned from Giga and resurrected CSA (which had lay dormant during this time), then spent the next two-and–a-half years designing and testing new benchmarking scripts to push Intel’s high-end hardware through its paces.
During this time I worked on numerous projects involving chip launches (the Pentium III, Mobile Pentium II/III, the Pentium 4), networking gear (desktop GbE), and multiprocessor systems (in conjunction with Dell Computer, another client of mine). An overarching theme throughout this engagement was the concept of “Constant Computing” – i.e. the idea that PCs are never really idle, especially under more complex OS like Windows NT/2000 – and I produced over a dozen white papers for Intel cataloging my findings and conclusions.
Note: My primary contact throughout this time period was Tom Harper, a maverick technical marketing manager who reported directly to Pat Gelsinger. Pat was also aware of my work and even used one of the test cases I developed as a Constant Computing demo piece for his keynote speech at the 2000 ISMC (International Sales and Marketing Conference). They later licensed the test case code from me for a tidy sum.
Eventually, my services were farmed out to Intel’s Desktop Architecture Labs (DAL), where I continued to refine my methodologies and also began formulating the idea for my first stand-alone test tool: Benchmark Studio.
The Wall Street Connection
Note: The following section has been heavily modified at the request of our client. It is a violation of our contract to identify them publicly, and we are hereby honoring that request.
They say that all good things come to an end, and when the dot com crash hit silicon valley I found myself out of my primary consulting gig and looking for a new challenge. Working for Intel had left me flush with cash (all told, they poured nearly three quarters of a million dollars into my small, two-person consultancy), so I had time on my hands to work on my test tool ideas. Eventually, I released the first version of Benchmark Studio as a low-cost, commercial test suite. And, lucky for me, it caught the eye of the lead tech in the PC Engineering group at a large financial services firm in NY.
That foot in the door turned into the biggest success story of my career. But back then, circa 2001, I was just happy to have them as a customer, period. And eventually, after cultivating a strong support presence and generally proving myself as a reliable technical resource, my contact recommended me as the best person to help one of the company’s largest divisions develop a new performance monitoring framework for their high-end Windows-based trading workstations.
Needless to say, I jumped at the chance, and over the next three years I developed and refined what ultimately became a commercial performance monitoring product known as Clarity Suite. And as we moved from pilot project, to limited production deployment and finally a business unit-wide site license in 2006, I was rewarded with a steady stream of consulting contracts culminating in the aforementioned licensing deal.
And while I’m not a liberty to discuss the value of these business transactions, suffice to say that they far exceeded my total compensation from Intel. Add to this a smaller scale deployment at CSFB (Credit Suisse First Boston, which was what they were still called in in 2001) and a pioneering study of workstation scalability conducted at Kent State University (under the direction of Hewlett Packard and Intel Corporation – the white paper is still available), and I was quite busy during the first half of the last decade. Again, heady times for a now older and more seasoned IT veteran.
Devil Mountain Software Emerges
It was during this timeframe that I decided a new corporate presence was required to help differentiate my consulting past as Competitive Systems Analysis, Inc., from my long term goal of productizing Clarity Suite and bringing it to market. So I once again collaborated with my wife and long time business partner, and together we created Devil Mountain Software, Inc. – with me as the public face of the company and her as the silent partner working behind the scenes to manage the business.
Eventually, we brought in other partners to invest in the venture, but we kept the management team limited to just ourselves. And when a site license came through from one of our biggest clients (essentially guaranteeing we wouldn’t have to work again for the rest of our lives), I started thinking about alternative ways to leverage what was now DMS Clarity Suite – options and scenarios that existed outside of the traditional commercial resale channel.
One idea that I had always wanted to explore was taking DMS Clarity Suite online – essentially providing the same kinds of monitoring and analysis functionality we were delivering to our commercial clients on their in house servers, but in a more limited, less feature-complete format. The goal would be to create a community of users around a set of free tools and services, and then to mine the data they uploaded in order to gain insight into trends and developments affecting the broader Windows community.
Thus, the exo.performance.network was born. But not before I made a fateful detour back into the world of IT journalism – a wrong turn I would eventually regret in ways I could never have imagined.
Early InfoWorld Involvement
Throughout the early part of the recent decade, I kept a toe in the water of my old haunt, IT journalism. It started quite small. From time to time I would collaborate with contacts at the InfoWorld Test Center – then still a real, physical lab space in silicon valley. I got to know some of the lead contributors, like PJ Connolly, quite well, and we’d get together at the lab sometimes to run benchmark tests using the aforementioned tools I developed for Intel, etc.
Eventually, the lab was shuttered, and InfoWorld started drastically downsizing its operation. It was during this time that I struck up a collaborative relationship with Doug Dineley, who to this day remains a class act and the one person I had the most respect for at that publication. But back then, it was all about product reviews and testing. Doug would present me with a list of possible story angles and I would pick and choose based on what struck my fancy at any particular time. Eventually, I became one of his more regular contributors, and he remained a good friend and close confident right up to the bitter end of my involvement with the publication.
But even when I started splitting time between my day job supporting my commercial clients’ deployments (now pretty much just a software maintenance role), my hobby building xpnet.com and these occasional freelance reviewer gigs, the relationship with InfoWorld remained casual. It wasn’t until 2007 that things got serious. And that year, more than any other, will go down as one of the worst I can remember.
Wooed by the Dark Side
Late 2007 is a time period pivotal to this story because it signaled a series of beginnings. It was when I first started thinking about blogging for InfoWorld. And it was also when I first approached the publication about partnering with DMS on the promotion of an online service, one built around the still evolving precursor to what would ultimately become the exo.performance.network.
And at first, neither venture went very well. Newly promoted Editor in Chief Eric Knorr, who I had never met and had barely heard of prior to his ascension, was resistant to the idea. He didn’t think it would fit with their still undefined editorial focus (InfoWorld had only recently decided to drop print and go online only). Meanwhile, the blog became tedious to maintain, especially since I wasn’t being paid for the work.
But eventually, things changed. Eric settled in as Editor in Chief, and a new Executive Editor, Galen Gruman, emerged to forever change my life. For starters, Galen took a liking to the xpnet.com idea. He began championing the idea internally, working with me to refine the messaging and coordinate with the various sales and marketing groups to achieve buy-in. At the same time, Galen took it upon himself to become the primary editor of my now paid blogging gig. He helped me to identify which topic areas were having the most impact – and thus started me on my descent into internet “Shock Jock”hell.
You see, what Galen and I discovered was that the topics that were most effective in drawing readers were also those that skirted the edges of both legitimacy and taste. For example, if I wrote an entry detailing some deeply held belief about a particular IT vendor or technology, nobody paid any attention. However, if I simply vented about something that was bugging me – a mysterious crash in Vista or some piece of VDI “marchitecture” coming out of VMware – the attention level shot through the roof.
Eventually, I found myself enjoying the buzz that my “angry missives” would generate. Little did I realize how quickly such a model could deteriorate or how much it could damage me, personally, once it fell apart.
A Slippery Slope
As the missives kept coming, and the traffic numbers kept climbing, Galen and I – along with Eric Knorr – worked to evolve the persona of “Randall C. Kennedy.” I was now to be the lightning rod of the publication, the guy who puts the most provocative spin possible on every story with the intention of aggravating as many zealots as possible. The net result was gobs of page views – I was the single biggest draw, site wide, for all of 2009 – and also a great deal of scorn from my contemporaries.
Ironically, It was the growing disapproval of my peers in the industry that first gave me pause. I realized that I was now regularly espousing opinions and viewpoints that had almost nothing to do with what I truly believed. Rather, they were simply extensions of the RCK persona. I became the “Microsoft basher” when, at heart, I held the company in the highest regard. I became the “Vista basher” and the “Windows 7 basher” when, in truth, I used both every day and found them to be excellent products (yes, even Vista). The whole persona had taken on a life its own, and I was terrified that it would ultimately spin out of control.
Which of course it did, in the most spectacular fashion, and just in time to nearly destroy the one project that I truly cared about and believed in: The exo.performance.network.
In fact, Windows Sentinel (the co-branding nomenclature that Galen came up with for our collaboration) would never have happened if my colleague Mr. Gruman hadn’t pushed it through the various layers of IDG bureaucracy. The man was a bulldog, and working together we managed to launch Windows Sentinel in April of 2008 to little fanfare and even a few snickers.
Note: It’s important that the public understand the nature of the contractual relationship between DMS and IDG. The arrangement was strictly one of cross-promotion – DMS would provide the service and InfoWorld would promote to its readers. And while both parties would share in the registration data and collected metrics, at no time did any money change hands.
This was strictly a marriage of convenience, and the only side to ever see even a dime of revenue (from advertisements and sponsorships associated with the registration pages and related collateral) was IDG. So my detractors can put away their evil conspiracy theories of greed and avarice – they simply do not apply here. I gave everything to make Windows Sentinel a reality, and got virtually nothing in return.
The Return of Craig Barth
But back to the story. From the beginning, Sentinel had a credibility problem. Though it was being promoted as an independent service and research entity, it still had my name attached to it. In fact, InfoWorld made a point of identifying the solution as the product of a collaboration between the publication and its Contributing Editor, Randall C. Kennedy, the founder of Devil Mountain Software, Inc.
They even plastered as much across the registration page. It doesn’t take a genius to tell you that having the industry’s most notorious internet “shock jock” as your only front man was not a formula for success. So I took drastic action. I created a fictitious spokesperson by resurrecting my pen name from days gone, Craig Barth, and assigning him the title of Chief Technical Officer for Devil Mountain Software, Inc.
It all started fairly innocently. I would receive an email inquiry from some media person asking about a piece of research I had published through my official exo.blog, and I would reply – not as Randall C. Kennedy, the “shock jock” that nobody took seriously anymore – but as Craig Barth, the ever helpful and deeply knowledgeable CTO of DMS.
Over time, these interactions became more frequent, and I began to enjoy my newfound anonymity. No longer fearful that my hard research would be rejected out of hand, I became bolder, even going so far as utilizing my alter ego when fielding phone calls from the likes of Gregg Keizer and others.
And all the while I wondered to myself why nobody was making the connection? How could a legitimate service that was so publicly launched by one of the most reviled personas in the IT media sustain such a ruse? Couldn’t they see the absurdity of it? Devil Mountain Software, Inc., was the company that Randall C. Kennedy formed. There was no attempt to hide this fact.
Of course, someone did see through it all. And it was with fear and trepidation that I fielded a phone call from my close colleague and co-architect, Galen Gruman. He had seen one of Gregg Keizer’s earliest mentions of me – a report on on benchmark results that showed Vista SP1 failing to provide a promised performance boost – and he wanted to know who the hell this Craig Barth guy was.
After all, if anyone knew me well, it was Galen. He was there when I first pitched the idea for Windows Sentinel. He was there as I wrestled with how to separate the hard data from the “shock” persona. And since he was intimately familiar with DMS and its management team of 1+, it was only a matter of time before he made that call.
To be fair, Galen was not pleased when I confessed my actions. He felt that I was pushing the ethical boundaries by misleading the public in this way. However, for whatever reason – personal loyalty, a desire to maintain the status quo with the “shock jock” persona – he agreed to keep it to himself.
Frankly, I’d figured he’d expose me on the spot. But instead, he turned a blind eye, even as I referenced my own research in my blog –data that had been promoted by Craig Barth and consumed by countless other media outlets ignorant of the ruse at play. Somehow, Galen managed to hold his tongue for over a year, even though secretly he must have wondered when it would all come crashing down.
Note: While I can’t say unequivocally when InfoWorld Editor in Chief Eric Knorr was made wise to the ruse, I’m pretty sure it was well before the whole mess spilled over. Eric was as intimately familiar with the nature of DMS as Galen, and for him to pretend to have been oblivious to the situation – when the persona of Craig Barth from DMS had been plastered all over the Internet for a year or more – makes such a claim hard to swallow. And despite Galen’s apparent loyalty to me, I still can’t see him keeping this from his immediate superior, if for no other reason than he would need a way to cover his own ass during the inevitable implosion.
Crash and Burn
And implode it did. After publishing a particularly alarming set of findings – which I still stand behind while continuing to evaluate new data – the internet became engulfed in controversy. As the furor grew, and as more and more media outlets questioned just who this Craig Barth fellow really was and what made DMS tick, the house of cards came crumbling down. The persona of Craig Barth was exposed as one Randall C. Kennedy, and the entire web of half-truths and misdirection was exposed as the ruse that it was.
Frankly, I was relieved it was over. Balancing the two worlds had become almost impossible, and I longed to escape from the “shock jock” persona that had been created for me so I could once again embrace my core beliefs. But what surprised me was the level of anger expressed towards me for what I saw as nothing more than a very poorly executed attempt to escape from the proverbial rock and hard place. Simply put, the level of vitriol expressed felt way out of proportion, and the claims of “egregious ethics violations” and “insufferable breach of trust” were simply over the top.
After all, it’s not as if I had trafficked in nuclear secrets or or stolen someone’s credit card information. I merely tried to shield what was important to me from the fallout of the world that had been created for me. And in the end, I failed miserably. It was a dumb move, born of frustration at feeling painted into a corner of my own making. I should have just walked away earlier – it’s just a blog in the end – but I lingered too long on the edge of the razor, and eventually it cut the heart out of everything I had tried to accomplish.
Please note that I’m not looking for sympathy or even understanding. My goal here is simply to clear the air – to tell my side of the story and to hopefully clarify both my professional background and the nature of the very legitimate products and services I’ve developed.
At the end of the day, this whole affair is just a blip in the timeline of a a career that spans two decades, one which saw me working with a bunch of amazing people at a some of the most revered companies in the world. I’m proud of my many accomplishments, and I’m happy that I can finally close this chapter of my life.
It’s been one hell of a ride…
So, what next? For starters, neither the exo.performance.network or Devil Mountain Software, Inc., are going anywhere anytime soon. I will continue to develop and expand what has become my true labor of love, but now with a renewed commitment to the integrity and authoritativeness of the data that makes the service so special.
What I will not be doing is venturing back into the field of IT journalism. Not because I couldn’t do so if I chose to – you’d be surprised at how many emails I’ve received offering to host my “shock jock” persona on a different site (some people will stoop to anything for a few page views) – but because I never want to compromise my integrity that way again.
At the end of the day, I really am Randall C. Kennedy – a passionate fan of all things Microsoft Windows-related. Thank you for taking the time to hear me out.
Get more charts like these for free at www.xpnet.com.