Is moving to the cloud the new IT outsourcing? And are we making all the same mistakes?
Jim Fowler, CIO of GE Capital, said in 2014 that "We've gone too far from an outsourcing perspective." GM announced in 2012 that they would reverse their outsourcing decision and change from 90% outsourced, to 90% insourced. A new wave of insourcing was coming, we were told.
And yet today we see companies announcing a wholesale movement of IT systems into the cloud. Australian real-estate listing site Domain is doing it. The Guardian recently announced they are moving everything to AWS after failing to stand up their own private cloud solution based on OpenStack.
Are we repeating the mistakes of the outsourcing movement, by swapping in the word cloud?
Maybe. It all hinges on the decision process behind outsourcing, or moving to cloud.
At the core of a decision to outsource is an understanding that there’s little value to the organisation of performing a function itself. The classic example is of office cleaning; virtually no organisation has their own full-time office cleaning staff. Instead the work is performed by people from another firm who contract for the work. This is classic division-of-labour, comparative advantage stuff.
The decision to outsource IT follows the same logic: why do this ourselves when someone else can do the same job, only cheaper, and possibly better?
Joey D’Antoni, Principal Consultant at Denny Cherry and Associates Consulting, told me that “in my experience, the organisations that are best suited to outsourcing have rigid, defined processes. That’s very uncommon, especially in places that are trying to cut costs.”
An organisation that signs up for a five year contract based on poorly understood processes ends up discovering the limits to their understanding, and at high cost. Everything not covered by the scope of the outsourcing contract is now a change request, billed at a significantly higher rate. Purported cost savings evaporate as getting the same service as before ends up costing more than was anticipated, because the detailed understanding of what was already happening wasn’t there.
Julian Wood, a Consulting Solutions Architect, says that “people often outsource the stuff they don’t actually understand the value of. The value they then get isn’t what they wanted.”
And this is the second sad result of a poorly managed outsourcing. IT departments who aren’t good at articulating their value will sometimes hide costs elsewhere. A switch upgrade is tacked onto a project because the approvals for the switch upgrade itself were denied. Extra storage is purchased, and sits idle, because project capacity forecasting is woeful, but projects hate being delayed by provisioning times. Business as usual work is buried by fudging WBS codes, so that arbitrary utilisation rates are maintained but unfunded work still gets done. IT is their own worst enemy here by keeping the lights on, but getting none of the credit (or cash) for doing so.
Thus not only do the costs increase when the work is done by someone else, but all the hidden value suddenly evaporates, and the organisation pays more for less. A classic failed outsourcing.
But look at the hidden assumption here: that outsourcing this function doesn’t provide any competitive value to the organisation. For office cleaning, that’s probably true, but for information processing in a world we’re frequently told hinges on better decisions based on access to information?
If information is really that important to your business, why are you giving those processes to someone else? If every business is a digital business, as Accenture has said, then why stop doing something you should be grabbing on to with both hands?
Will we see a swathe of companies deciding, as GE and GM have, to bring things back in-house after years of giving their competitive advantage to someone else?
