A bird does not sing because it has an answer.  It sings because it has a song.

16 October 2012

15 years ago today

1000 ISPs Live With x2 Technology -- October 16, 1997.

In honor of the 16th anniversary of the launch of x2 today, I have gone back and set up links on my x2 story posts so that you can start at the beginning and click seamlessly through to the end.

(Actually, it was because one of my former coworkers asked for an easier way to read the story, but hey -- any excuse will do.)

I find it fascinating that I can still locate 1996 and 1997 press releases about this stuff.  There's still a website called 56K.com , and although it's probably not as useful (or as widely read) as it was 15 years ago, it's a fun little trip down memory lane if you're into that sort of thing.

After today, though, I think it's time to move on to more current affairs.  Time to choose another tale to spin.  I'll give it some thought, and I will take requests.  For today, though, happy x2 anniversary, especially to all my old Robobuddies.  Download a toast to the great time we had.  It's on me.  :-)

08 October 2012

Postscript to the story of x2

What was the point of all that, anyway?  As of 2010, according to the FCC, 6% of internet users connected via modem.  Although weirdly enough, Wikipedia reports that AOL added 200,000 dialup customers in 2011 (presumably because it's cheaper than broadband for those hit hard by the recession).    The slowest broadband connection you can get (ISDN) is double the speed of a V.92 modem, and the fastest is measured in hundreds of megabits. So I certainly didn't write down this story to brag about the technology.

Years later, 3Com spun off what was left of U.S. Robotics, and the "new" company is basically a marketing shell for foreign-built commodity hardware.  (It was a big point of pride at USR that we built every one of those modems right here in the good ol' U.S. of A. -- Skokie, Illinois, in fact.)  So it's not about the happy ending, either.

I wrote the story down for three main reasons.  Number one was catharsis.  It was a high time of my life, working in a dynamic environment with supremely cool people, and I miss that, and them, every day.  Writing it down let me relive some of those memories and perhaps preserve them for the future (see also "fandom").  I hope the people who were there with me have followed along and enjoyed a little look back at our history.

Number two was teaching.  We did some things really well, and I hope I touched on those in an instructive way.  Don't look at the details, look at the concepts and methods.  That entrepreneurial spirit, strong sense of teamwork, respect for customers, and drive to solve problems would serve any organization very well, no matter what project it was undertaking.

Number three was as a cautionary tale.  3Com and U.S. Robotics were technologically very compatible.  The merger made all kinds of sense if you just looked at the numbers.  But their cultures were fundamentally at odds with each other, and as soon as either company started telling the other one how to do its work, it was doomed to failure.  Remember that U.S. Robotics acquired Palm Computing in 1995, and that worked out pretty well for both organizations, primarily because USR left Palm alone to do what it did best.  When 3Com started meddling in Palm's business, its founders left and created a new company.  Do you see a pattern here?

Please don't make the mistake of thinking I intend to vilify 3Com.  3Com happens to be the perpetrator of this particular evil, but that doesn't mean 3Com is evil.  More importantly, it doesn't mean 3Com is unique.  Failure to understand the role of culture in a company's success lies at the core of many business failures and almost every failed merger.  If the merging companies don't share similar cultures, or at least cultures that can find common ground and adapt into a single community, the company will fail.  3Com, a company founded by the co-inventer of Ethernet and once possessing enough cash and status to buy naming rights to a major league baseball park, is now completely gone, having been absorbed into Hewlett-Packard in 2010.  U.S. Robotics, once an employer of over 7000 people, now employs about 125.  And Palm bounced around over the years, ultimately being spun into a wholly-owned subsidiary of HP called Gram.  While technological advancements and a few strategic errors are certainly at play in the ultimate demise (or serious marginalization) of each company, the massive failure to get the culture right was a significant factor; probably more significant than any executive at any of the companies ever really realized.

So let this be a lesson to would-be masterminds of M&A -- ignore culture at your peril.  It's not just about what people wear to work.  It's about every choice they make, every word that comes out of their mouths, and ultimately every action they take to ensure the success or failure of an organization.

03 October 2012

3Com: Crossing The Bridge

We had barely recovered from our six-month sentence in NDA hell over x2 when we once again found ourselves under NDA.  This time the story was that U.S. Robotics would be merging with 3Com.

I seem to recall that we learned this in January or so after the x2 announcement, and the merger was consummated in June (so saith Wikipedia -- I had remembered April but had to look it up to be certain of the year).

So underneath the excitement of the x2 rollout lurked the uncertainty of the merger.  There were some layoffs in the last quarter of 1996 and early 1997, and the mood was a bit schizophrenic.  That only got worse after the merger news surfaced, but the x2 momentum carried the team, and the team spirit, through to the 1-year anniversary and the 1000 ISP party.

Oh, yes -- we had a party.  Once the independent research confirmed that we did indeed have 1000 ISPs live, we threw a massive tent party on the lawn of the UState building and invited everyone.  We printed up t-shirts and tapped more than a few kegs.  There was music and food and a special Connect NOW message that I got to deliver on stage.  It was probably the last time USR employees got to party like USR before the stodginess of 3Com started drawing the life out of it.

I believe it was December of 1997 when the holiday talent show featured the Product Management Singers -- Burk, Rick, Barb, Don and I -- performing my lyrical masterpiece, "Santa Clara's Comin' to Town." ("You better watch out, you better not doze; you may find your desk in Rolling Meadows -- Santa Clara's comin' to town....")  We were still having fun while we were still in the heritage buildings.

We moved into that massive eyesore of a building at 3800 Golf Road in June of 1998 (I had to look that up too -- thank you Chicago Trib online archive).  It was right around then that the best and brightest talent of USR started heavily bleeding out.  A few of our really good engineers were already gone by then, along with a few treasured folks from other departments, but the worst of the talent hemorrhage started about mid-1998.

Something you need to understand about the "merger" of USR and 3Com is that it wasn't a merger.  And that, in a nutshell, is why it was such a miserable failure.  The USR folks were told it was a merger.  The 3Com folks were told it was an acquisition.  And this fundamental failure of communication drove everything that happened from the moment the deal closed.

3Com didn't understand how we worked at USR, and didn't particularly care.  The new building came with 6-foot high cubicles that isolated us from each other but didn't block noise (the worst of cube life and the worst of office life all blended together).  We were told not to put anything on top of our wall cabinets or outside our cube walls.  The place was like a rat maze; it could take a considerable while to find anyone in the identical endless aisles.  (On a business trip that summer, I learned that Santa Clara HQ was even worse.)  It was relentlessly beige and enormous and lifeless -- except when we got our summer thunderstorms and discovered that those gigantic walls of windows leaked like sieves.  Then it was exciting!

Because Santa Clara insisted on having its say in everything we did, our cycle times and customer responsiveness slowed down to 3Com's pace.  We lost our best engineers, our best sales people, and our best product managers, and eventually we lost our edge in everything we did.  When the V.90 56K standard was adopted in spring of 1998, the last nail was driven in the commoditization coffin of modems, and by then, we'd lost the people and the entrepreneurial spirit that would have generated the compelling follow-on product.

We tried our hand at videoconferencing systems, conference phones, ISDN, cable, and DSL, and did OK in those markets, but never anywhere close to what we had achieved in analog modems.

It was an amazing run, but it was over.  I left 3Com in December of 1998.

Next up: a postscript