Isn't this a well-known part of a company's lifecycle. You stop innovating and try and live off of your previous reputation. Yes, there is inertia and sticky customers but ultimately, people will eventually ask why they pay a premium for something that they can get equally well elsewhere. Usually, by that time, you cannot recover your position.
At least Microsoft learned this before it was too late and went hard into cloud, Linux and open-source.
yea when you've built a reputation to a certain point there is less motivation to innovate. any mistake you do have less consequences as well, so if you fail your clients will just shrug and go "well it's IBM.... if they can't do it then nobody can"
I would guess that one reason so many legacy big tech companies failed to compete against AWS is that they had a critical mass of sellers who were not equipped to sell cloud. Combine that with existing relationships with CIO/IT folks who were incentivized to maintain the status quo makes it very hard to actually do things differently. I believe the key thing AWS changed was that the sellers were not gatekeepers to purchasing. AWS sellers had to figure out how to add value beyond just doing demos and negotiating pricing and contracts. I doubt IBM was willing to change it's sales process or who it targeted for building relationships within the customers organizations.
> Over and over again during the last decade, IBM engineers were asked to build special one-off projects for key clients at the expense of their road maps for building the types of cross-customer cloud services offered by the major clouds. Top executives at some of the largest companies in the country — the biggest banks, airlines and insurance companies — knew they could call IBM management and get what they wanted because the company was so eager to retain their business, the sources said.
This! We literally build & prioritize feature work based on one-off customer requests...
I recall back in the early days, working with IBM professional services on a large contract Lockheed was doing with DOW Chem, but hosted in IBMs datacenter space at a very large DC in South San Jose...
It was a FN nightmare to get anything done - it took a month just to get static IPs allotted to the project, way too many meetings, too many PMs on the IBM side and the pricing was absolutely ridiculous. IIRC we were charging DOW ~$40,000 a month to host two Racks in the DC when prices for bandwidth were ~$900/MB at the time...
I then worked with IBM and Intel when Intel was deploying a an OpenStack setup for RedHat. Same level of BS to get simple things accomplished. It was one of those cringeworthy projects where you had PMs on the team who didnt know jack about anything we were doing and were super slow to get anything done. You just cringed when on a call and people throughout the project were so disjointed and disconnected that you felt embarrassed for them on the calls...
I can relate to this. I did some contracting work for a large bank (let's call them client #1) and had to deal with IBM mainframes connecting their OSA-adapters to an IP-network. These adapters talk OSPF to the network and we needed some changes. No one from IBM allocated to my customer then (this is essential..) had any clue. I had to arrange via another colleaugue of mine working for another customer (client #2) who also dealt with IBM that an IBM engineer for his client (#2) talked to the IBM engineer for my client (#1) to have this change done. IBM seems to be silo'd as hell and no one from their team (PM's whatever) seem to be in any rush to contact within IBM to help solve an issue.
At the time (2016-2018) their IBM Bluemix cloud offering was a joke. Custom made and in no way a match for AWS/Azure. Luckily I persuaded my client to refrain from using Bluemix and go with Azure/AWS instead.
16 comments
[ 14.5 ms ] story [ 773 ms ] threadAt least Microsoft learned this before it was too late and went hard into cloud, Linux and open-source.
This! We literally build & prioritize feature work based on one-off customer requests...
It was a FN nightmare to get anything done - it took a month just to get static IPs allotted to the project, way too many meetings, too many PMs on the IBM side and the pricing was absolutely ridiculous. IIRC we were charging DOW ~$40,000 a month to host two Racks in the DC when prices for bandwidth were ~$900/MB at the time...
I then worked with IBM and Intel when Intel was deploying a an OpenStack setup for RedHat. Same level of BS to get simple things accomplished. It was one of those cringeworthy projects where you had PMs on the team who didnt know jack about anything we were doing and were super slow to get anything done. You just cringed when on a call and people throughout the project were so disjointed and disconnected that you felt embarrassed for them on the calls...
At the time (2016-2018) their IBM Bluemix cloud offering was a joke. Custom made and in no way a match for AWS/Azure. Luckily I persuaded my client to refrain from using Bluemix and go with Azure/AWS instead.
Suppose I'm sent back to undergrad, and I want to host a side project I'm working on. What's the path of least resistance?
Is IBM anywhere in this conversation?