On this, the third and final day of my work at IBM Interconnect, I have enjoyed several intelligent discussions and sessions that have my grey cells firing. This morning, I attended a session conducted by Steven Dickens of IBM (video interview with Dickens to follow shortly) and Guy Shone of Explain the Market (a UK technology economist) that threatened to explain how a CFO might decide to authorize the migration of core applications to the cloud.
I spent about an hour before the session getting to know Shone, who strikes me as a bright and hype-free analyst. He writes a column in a London paper and is a frequent commentator on the British airwaves regarding the business value of technology — one of my favorite subjects.
What he said got me thinking, which is what IBM says we all should do. There are signs all over the IBM Interconnect venue advising us to THINK, and others to EAT, ASK, SLAY (???), etc. For a moment, I thought they were doing an homage to John Carpenter’s classic film, They Live, and I kept expecting to see signs like OBEY, or MARRY & REPRODUCE.
But I digress.
Shone spent a lot of time noting that a lot of change is going on, perhaps too much to neatly define, categorize or model in order to predict outcomes. CFOs tend to be conservative and slow to jump on new tech, preferring to play it safe and take advantage of tech that demonstrates mainstream adoption and generally accepted value. Except, in these days of disruptive micro businesses and unbridled mobile commerce, waiting is not always the smart move. There may be some first mover advantage. A conundrum.
He was critical, as am I, of what substitutes these days for sanguine (economic) analysis. There is a major (and deserved) trust gap among CFOs in reports from industry analysts (who too often sell their opinions to vendor clients), to stilted social media endorsements (how many likes a product or service receives can be distorted deliberately by vendors), and even in use cases (which rarely provide replicable outcomes). What’s the poor CFO to do?
The amount of information to digest to get at a kernel of truth in analysis is huge and growing. Signal to noise is getting pretty awful. Vendors toss around statistics that were created under contract to other vendors which they privately disparage as so much BS. It’s hard work to be a CFO who is constantly under pressure to let loose the hounds of technology but always cautious not to let the dog bite the hand that feeds it.
So that’s one thing that has me thinking. Is there a model with dependable and replicable outcomes that a CFO can leverage to guide decisions like moving core apps to the cloud? That question was never answered in the session, by the way. At best, Steven Dickens offered to make consultations available to CFOs who were contemplating a move to IBM Clouds.
As I listened to Dickens extol the virtues of mainframe cloud technology, I was hearing something a bit more subtle. Last year and the year before, Interconnect presenters showed “hybrid cloud” slides that depicted traditional organizations running most core apps (legacy apps, systems of record, etc.) on premise. To the extent that public cloud services were leveraged at all, it was for ancillary uses — like storage of archival data or for DR as a Service, that sort of thing. IBM was fond of saying that larger firms were not ready to move wholesale to the cloud. Hybrid was a way for them to dip their toe in the water before diving in.
Not now, apparently. This year, IBM seems to have redefined hybrid cloud to mean the placement of some core apps in the public cloud (an IBM one) and some on premise. They have spun the discussions of the sessions at Interconnect to include lots of references to “cloud native” apps and why companies should be shifting their development efforts toward this model.
Yet, there is no model to encourage CFOs to embrace such an endeavor. There are, theoretically, some short-term gains. If getting skilled workers to come to your firm wherever you happen to have put down stakes is a hassle, that may go away by going to the cloud, where communities of developers and operators live. So there’s that OPEX value, as there is with any outsourcing arrangement.
And of course, if you no longer want to pay the price for equipment upgrades, that will go away with the cloud…or with any outsourcing arrangement…too. So there is some CAPEX value there.
But these are older use cases and not always guaranteed to offset the costs of cloudification, including denial of service attacks, outages, security breaches, etc. They certainly do not speak to the granular issues of the overwhelming diversity of cloud technology initiatives that force you to hitch your wagon to an industrial cloud service provider’s star. Is joining an IBM Cloud the equivalent of a lock-in? Do we need to be concerned about that or is it the price for doing things in a way that delivers predictable outcomes? That has been a question on my mind, honestly, since the Open Technology Summit that I attended at the start of this show.
Still Dickens has a nice Midlands voice — a true cloud whisperer. And while he might not have a rigorous economic model for cloud adoption yet, he might just sweet talk some CFOs to increase the 17% of revenues that IBM currently derives from its cloud services.
BTW, I am attending IBM Interconnect as a guest of IBM, who has paid for event registration, transportation and lodging. My words here, however, are my own.