Last week was the anniversary of Windows 95, and there have been few things that have impacted me as much as this OS launch. I learned a lot from the launch and post-launch support. I met Bill Gates for the first time—he nearly got me both fired and saved my job. I did a ton of media and an impressive amount of TV. I even had my own morning TV segment on the local TV news, and I did my first PC build, and this was before motherboards and cases had mutually acceptable standards. I put the motherboard on rubber bumpers because the connectors weren’t even close to matching up.
It was a fantastic, frightening time that changed my life. This week, let me reminisce about the year Windows 95 launched and how it changed my life.
The year of Windows 95
I left IBM in 1994 to join Dataquest and become one of two (joining Paul Cubbage) operating system analysts. I’d spent ten years in a variety of jobs mostly pissing people off in IBM, and once out, I was without a rudder. The operating system service was a bit of a mess with few subscribers and reports that didn’t reflect on individual brands like DOS and NetWare, but by whether the OS’s were 8, 16, or 32 bit. Even I didn’t care, and it was my job to project which platform would win.
I’d had a pretty massive research budget at IBM and had been the West Coast leader on the failed attempt to spin out the IBM software business. So, in my head, I had a ton of actual and projected operating system data, and my partner in the service had created a projection model that was pretty good at predicting the future if you put in enough accurate data. Between the two of us, we turned out the first five-year forecast of operating system sales, and I immediately made a ton of enemies.
We projected—mainly based on Microsoft’s massive demand generation budget and existing DOS/Windows penetration—that over the five years, MacOS, UNIX (Solaris, AIX, etc.), NetWare, AIX, OS/2, and of course, DOS would go into sharp decline and Windows 9x would emerge supreme. What we didn’t factor in was Microsoft’s near elimination of that marketing budget before the launch cycle was over, so our projections on Windows were overstated. At the same time, the decline of the other platforms were far more accurate.
What happened with Windows 95's post launch
There is an old joke about researchers cutting off a frog’s legs and coming to the wrong conclusion that is pertinent here. The joke goes that they start off by testing the frog’s jump—it now strikes me that this is very cruel—with four legs, and it jumps 4’. They cut off one leg, and the frog jumps 3’, then another leg, and it jumps 2’. One more leg and it jumps 1’. They cut off the final leg and conclude it made the frog deaf because it no longer tries to jump. Right data, wrong conclusion.
With the Windows 95 launch, Microsoft had created a massive amount of demand, more demand than anyone had seen for a product like this in history. There were lines around buildings to buy the OS. Still, people weren’t IT experts (heck, even IT experts want what they became). The controls over accessories, and particularly drivers, were horrible. The result was support calls that overwhelmed Microsoft’s support capabilities. Folks were kept on hold for hours. To address the long hold times, and this is where the frog story comes in, the head of the service reduced the hold queue by fiat. The result was that instead of thousands of people being put on extended hold, they got a busy signal, and Microsoft had no idea how massive the number was of people who couldn’t get help. And it was huge!
In what was an impressively short period, we went from lines around buildings to buy the thing to images of pallets of unsold Windows 95 boxes sitting prominently in the middle of big box stores.
Now, rather than taking what worked, the demand generation process, and fixing what didn’t (lack quality control over accessories and drivers and inadequately staffed support), the company seemed to conclude high demand was a bad thing. That was the last time we saw anything like the amazing launch of a tech product until Apple matured the iPod and the iPhone years later.
In a way, Apple seemed to learn more from the Windows 95 launch than Microsoft did, which I still find amazing.
My near OS/2 disaster
As I noted, I’d left IBM to join Dataquest. Before I left IBM, we’d pulled the budget from OS/2, so it was pretty straightforward. Against the massive Windows 95 budget, OS/2 was toast because IBM wasn’t going to fund a defense. But I did call my contact at IBM and warn them that I was going to release a report saying OS/2 was toast and that she should give executive management a heads up.
She responded that I was a nobody, nobody cared about Dataquest, and that my report wasn’t worth the paper that it was published on. So, not only wasn’t she not going to let her management know, she wasn’t going to worry about the outcome. My report, since it was released on top of the massive Windows 95 demand generation effort, was front-page news all over the world. I was an instant semi-celebrity, which was enhanced when the hardware side of Dataquest publicly said I was full of crap.
That extra drama made the report incredibly popular. That same IBM contact then felt the need, because her management was blindsided, to call and tell me I was done as an analyst, and I’d never work in this industry again. One of the most exciting outcomes was a call from the then CMO of Novell, who threatened to pull her funding for my service if I didn’t change the report. My boss came running into my office asking if Novell was a client. They weren’t, and the next day that CMO dropped off a check for $75K to become a client. And yet I never changed the report.
IBM later had me flown to New York and put me on the spot in front of a room full of OS/2 engineers to dress me down. What was pretty funny (though not at the time) was the engineers came to my defense and told marketing I was right and that they should leave me the hell alone. They were pissed that their budget had been pulled, and I don’t blame them.
It was a crazy year, and it often seemed like I was in a race with Dataquest’s CEO to see who would get fired first. One other near disaster was put on an international tour to talk about the report, and upon arriving in Japan, my co-workers decided to trick me into getting drunk the night before my presentation (apparently it was a Japanese tradition). Japan was a substantial OS/2 customer, and my boss told me that no matter what I did, don't say OS/2 was dead. I got up on stage so hungover that all I saw was a dull blur, and, of course, I said OS/2 was toast. He was pissed, but I didn’t care because I felt like I’d already died and was waiting for my body to get the message.
During that year, there was a severe problem that I shouldn’t have shared with Microsoft but, had I not, it would have cost that company millions. I told Bill’s folks about it, and Gates corrected the problem naming me as the source. My boss came into my office and told me I’d likely be fired by the weekend because of my mistake. Bill found out about this and wrote a commendation letter to my CEO, which saved my job.
In a way, Windows 95 changed my life. Had it not been for that product, the report wouldn’t have been controversial, I’d never have become somewhat famous, and Giga Information Group would have never recruited me from Dataquest, where I eventually became Sr. Research Fellow. I doubt I would have been able to form my firm.
In a way, I owe the life I have now to Windows 95. Happy Birthday Windows 95, and thank you for a wonderful life! And thanks again Bill Gates for saving my ass back then.