IBM spins out services - The next revolution after Gerstner’s?

I usually leave IBM earnings presentations to another kind of analyst – I’m just interested in whether its technology stories make sense in a technological sense rather than in how much money it is making from them. But this one has big implications. IBM is now planning to spin out managed infrastructure services into a new public company. But, as my old co-editor and fellow Bloor analyst Martin Banks would put it “what this means is” is of the essence.

For anyone following the Bloor Hybrid Cloud story, IBM is a serious player – all Cloud players support Hybrid Cloud these days, but IBM is one of a select few that actually seems to believe in it deeply. In contrast to IBM’s multi-cloud “integrating and connecting” approach, one gets the impression that AWS and Azure etc. would really be happiest if customers would just stick on their cloud infrastructures, importing stuff from outside onto their platform when necessary. Well, that is what I think. Fellow Bloor analyst Fran Howarth has reminded me that not everyone thinks this way – here, for example – but these people who really dislike IBM simply aren’t seeing the IBM I see (they’re probably more positive about , say, Azure and Salesforce than I am too). I do think the issue here is  “whether IBM’s “cloud” is actually cloud in the sense of what the rest of the industry calls cloud” – op. cit. – I think that it probably isn’t quite, and that that might be no bad thing. There are probably workloads that would be very difficult to run purely in Public Cloud; and lock-in to a specific public cloud (by a combination of services; complexity; and even licencing of associated on-premises tools – last time I looked SQL Server licensing was much easier if one chose Azure cloud) is entirely possible.

I have recently become convinced that IBM’s purchase of Red Hat and adoption of “OpenShift Everywhere” by default (obviously, it still supports other customer choices, but OpenShift is preferred) marks a fundamental revolution in IBM, commensurate with the Gerstner “IBM is a Services Company” revolution starting around 1993 (see a short account of this in the Harvard Business review. It is almost, it seems to me, a cultural reverse-takeover of IBM by Red Hat and the Open Source movement. One thing however is similar to that last century revolution: back then “change came to IBM in large part due to the pride and energy of the employees themselves” (op. cit.) and Gerstner’s role was as a catalyst – “People at IBM were very smart. I didn’t have to [look] outside; I had to find the people already there” who were ready to turn things around”, Gerstner said. People at IBM are still very smart and they are now being “fertilised” by equally smart people from Red Hat – such as Jim Whitehurst, author of “The Open Organization: Igniting Passion and Performance”, who might be catalysing IBM’s latest revolution.

In passing, I might just point out that IBM is one of the few large companies to have survived near-death experiences and to have reinvented itself not just once but several times. It is, in many ways, the epitome of a Mutable Business™ in practice. I put this down, in part, to its investment in fundamental research – it has added to the fundamental knowledge base in computer science in ways that other big IT companies mostly haven’t. Fundamental blue-sky research isn’t everything, of course, but I think it helps – if there was no IBM many genuine innovations (from scanning electron microscopes to modern relational databases) would have been slower coming.

So, what is this new IBM view of Private Cloud? To quote Arvind Krishna, IBM’s CEO, “When I say hybrid cloud, I am talking about an interoperable IT environment across onpremise, private, and publicly operated cloud environments, in most cases from multiple vendors. Why do we believe this is the future, rather than a pitstop along the way? Clients find that choosing a hybrid cloud approach is two-and-a-half times more valuable than relying on public cloud alone. This considers factors such as innovation, development time and portfolio”. What is worth emphasising here is that hybrid cloud is the destination, not a stage towards achieving public cloud. Some workloads may never move into public cloud, because regulations force them to stay on premises; or because they are so large and profitable that moving them simply isn’t worth it (in regulated financial services, the cost of regression testing to prove that workload behaviours haven’t changed may be prohibitive); or because they need cutting-edge technology that simply isn’t available in the Public Cloud yet.

On the whole, I share fellow Bloor analyst David Terrar’s view that this “is a great move to separate out the GTS services business and turn IBM back to more of a software, hardware and cloud based company.  Haven’t figured it all out yet, but a great thing”. I too think this is a huge move for IBM, and a good one, but that the full implications will only become fully apparent over time.

And so, back to the IBM earnings presentation that started this. There is a good, and reasonably positive, financially-oriented discussion in the Financial Times – it reports that “Big Blue’s share price jumped 8 per cent on the news, as investors welcomed the spin-out of a business that has become a drag on IBM’s growth”. Nevertheless, it also reports Arvind Krishna as saying that “different people inside its customers made buying decisions about infrastructure services, and that [the new company] would continue to have a close partnership with the rest of IBM after the separation”.