I want to develop a masters thesis on how IBM did not become one of the top three public cloud providers like AWS, Azure, Google. Was Softlayer the wrong acquisition? Were IBM executives too focused on earnings per share at the time and not investing enough in public cloud infrastructure? Did the board of directors lack vision? Did IBM executive leadership lack vision? Was there too much pressure from major shareholders that limited investment in public cloud infrastructure? Did IBM lack the engineering talent? What are lessons learned here? Is it too late for IBM to be a top three cloud provider?
19 replies (most recent on top)
I remember a lot of material in the early 2000s around on-demand computing. Seems like the right idea but it never got executed.
I was an IT architect in Global Services (div 07, George W org) during the Sam Palmisano era, and witnessed a lot of what was going on during that time. That division (strategic outsourcing, what eventually became Kyndryl) was promoting a concept called "Utility Computing" (also known as "e-business computing") that was an early precursor to the "Cloud Computing" concept that is commonly understood today.
If you're familiar with the Gartner Magic Quadrant, organizations are rated on two things: "Completeness of Vision" and "Ability to Execute". Looking back with 20/20 hindsight, IBM rated probably 25-50% on "completeness of vision" and 25-30% on "Ability to Execute". IBM's background in selling "hardware" and "services" resulted in a "utility computing" paradigm that had customers putting in requirements that would get sent to IBM outsourcing staff for execution. Someone would ask for say "1 server, 4 CPUs, 16 GB RAM, with network connection x-x", and a set of emails would go out to Dubuque or Boulder or Bangalore or someplace like that for manual provisioning by some team of outsourcing staff members. There was no means in place at the time for automated provisioning like you have today.
That's not to say that IBM wanted automated provisioning...they did. It was one of the big reasons that IBM bought Tivoli. However, their ability to actually turn that into a functioning cloud service was questionable at best. While the chartware and customer presentations were slick (delivered by lots of highly credentialed people with loads of letters behind their names), the action behind the scenes was reminiscent of a restaurant in severe trouble. Lots of pots and pans burning on the stove (provisioning emails that never made it to their destinations, provisioning that was either mishandled or never got done), angry customers, endless troubleshooting conference calls with more highly credentialed people...you know the drill.
As the saying goes, nothing breeds success like success. The opposite also applies. Failure breeds failure. Large-scale failures make executives g-n-shy, and I think it's safe to say that the development efforts at that time really put off a lot of people within the company. New executives eventually came along, and rather than develop internally, IBM tried to play catchup with acquisitions like Softlayer and Red Hat. While those acquisitions may eventually prove valuable in other ways, they have not made IBM into the successful cloud company that Amazon, Microsoft and Google eventually became.
"I want to develop a masters thesis"
BUT
you are not doing that. You are asking for others to do YOUR thesis.
2afi Keep your eye on AWS IBM and AWS are far more entangled than you know. Google was approached, but they turned AK’s offer down, so AWS got the business. In hindsight, Google may regret their decision as IBM still has considerable influence on Enterprise customers
Yes, IBM missed a lot of boats. So did other major technology vendors like Digital Equipment Corporation (DEC) and Sun Microsystems. DEC missed the boat on the PC and Unix servers. They bet the farm on the Alpha chip. They could easily have been a leader today if not for missteps along the way by short sighted executive decisions. Cutler from DEC walked into Microsoft with a design DEC had that later turned into Azure.
brought back memories i remember around 2012??? ibm tried to build a cloud structure on Maximo, it took over 80 hours to download the software, then over 4 weeks to try and get it to work, then they did another try - different approach, again it wasn't cloud it was a hodge poge of software trying to bolt it together - again failure, I think they had 3 other attempts before they brought soft-layer, spent 10 billion on data centers - that no one used, all in all a total co-k up. they still wasn't even thinking SaaS even in 2015/16. still not sure they have it right - they appear to have capitulated and run stuff on AWS now.
The only thing they know how to do is cook the books
Look at the executive layer. They all know nothing about technology.
1, short-sighted on Cloud computing.
2, Complete transformation from Technology company to Business company.
3, Terrible acquisitions, always Blue Wash
4, Terrible management and deeeeeeeeeeeep management tiers
Discrimination! IBM layoff many talented employees that were over 50.
You don't need to write a dissertation on went wrong at IBM. It really just boils down to three things: Management Incompetence, Insatiable Greed, and Blind Arrogance.
Softlayer was not a bad company. But we did to it the same thing as we did to Lotus and Tivoli and so many others. (I _think_ the brain trust realized that the have to treat Red Hat differently than most acquisitions so it's still quite viable.)
The other names you mentioned are what are colloquially called "hyperscalers." In order to succeed, they need to be able to scale customer workloads and (therefore) their own operations in a way that was unimaginable to other pretend players in the space. I don't think that there was anyone in the Cloud organization who had a hyperscalar vision or was prepared to pitch to (really) senior management the vast investment in both hardware and software to achieve hyperscalar status. To truly do that would have been much more of a "bet the company" proposition than the sort of modest scale cloud offering that came out.
So I think that all of the points you raised are excellent. But it wasn't just the board of directors; there was no real vision of how big this could be and what it would take to get there. As a result we brought a kn--e to a g-n fight. We never realized that it was a g-n fight until it was much too late.
IBM missed so many boats you would need a couple of super computers to keep track!!!
Years ago, when IBM executives realized they were falling behind in the cloud race they reverted to false reporting (to boost their bonuses and to fool the stock holders). IMO this also contributed to the misery and should be part of your thesis.
There is currently some legal actions relating to IBM mislabeling of other revenue as cloud revenue. Not sure what is the status, perhaps others can chime in on this?
Well summarized!!
IBM lacked the vision to see the potential of the cloud. This dates back to Sam and spills over to Ginni. To have Ginni double down on Sam’s mistake is just inexcusable for a CEO. I believe the mistake was amplified due to IBM being fundamentally a HW (mainframe) company. Sam went all in on services (the next big thing) and ignored Intel due to its commodity profit margins. Ginni doubled down on that, and then gave IBM the death blow by selling off Intel hoping that Power would pick up the slack. ISV’s agreed that Power was a better alternative (performance) but could not justify its cost vs Intel, along with its unique treatment of LINUX, so Intel prevailed by being “good enough” and that’s where the ISV’s put their effort. Once you lose your ISV base due to HW costs, and you sell off the commodity base due to profit margins getting squeezed, your cloud offerings are doomed.
Same in the PC division with proprietary architecture that got destroyed by plug'n'play compatibles. Only ThinkPad survived, albeit via Lenovo. My take is that overall, IBM applies the mainframe mentality to every innovation opportunity they have. And those are running out.
not unlike the IBM bet against Ethernet and IP and in favor of 100Mb Token Ring and ATM (who even remembers Asynchronous Transfer Mode)... a trail of bad calls
When Sam Palmisano was IBM CEO he said cloud would not take off and let IBM fall way behind. Now all IBM can do is buy companies to play catch up while destroying those companies like Soft layer