whoa. GE "we're moving more than 60% of our global workload to @awscloud". that's a shitload of $$$ right there
— Mike Drop (@monkchips) October 7, 2015
GE- "like many of you we outsourced way too much". that's GE saying that… big outsourcing is in big trouble.
— Mike Drop (@monkchips) October 7, 2015
Coming away from AWS re:Invent 2015 it was pretty clear that Big Change is in the air. Rumours of the Dell EMC deal were percolating, and Amazon was touting customers going “All In” on cloud. GE is of course the very model of a leader in the corporate world, so when it sneezes others catch a cold.
Paul Downey, in his role at the UK Government Digital Service, which was set up as a pilot ship to help the departmental supertankers with digital transformations, has plenty of experience of multi-billion pound outsourcing contracts. He recently wrote about an internal client concerned about the risk of new (and more effective) ways of working.
“The problem with agile is scope creep”, he said
“We run a tight ship here. We’re on a budget, spending public money and can’t afford for things to slip. We need a detailed plan and a fixed contract to hold our suppliers’ feet to the fire when they deviate from it. We have to deliver all of the features, and on-time!”
I’m used to people challenging agile as if it’s unconventional. As if it’s a new, untried, untested thing. The present is here, it’s just not evenly distributed.
But somehow on this day I wasn’t prepared for this challenge. I was flummoxed.
Maybe he did run a tight ship and just hadn’t seen the horrors I’d seen: the service delivered feature complete even though most of the features weren’t needed, the service so complicated it was unusable, feature complete lest the supplier invoked penalty clauses. The system which cost too much to change, which needlessly instructed people to post their passports to an office, needing operational staff post them back unopened. Systems procured with a fixed 10 year contract, and which were already obsolete before they were completed.
I have seen very similar arguments from people I consider to be shysters. The idea that massive outsourcing contracts don’t suffer from scope creep and related, massive cost overruns, would be funny were it not for the fact that in the public sector at least, it’s our money, paid in taxes, being wasted. The UK government has wasted tens of billions of pounds on failed IT projects over the last 20 years or so, and one huge step forward under the last coalition government was a more sensible approach to citizen service provision.
Statements like this are frankly just as striking as the GE commitment:
“These [huge] contracts were meant to be about lowering risk. But when you package up a huge range of functions into one contract the risk becomes impossible to manage”
– Tariq Rashid, then Lead Architect, UK Home Office
Packaging of risk into supposedly AAA stuff, managed by huge suppliers, is why I often refer to Big Outsourcing contracts as Collateralised Debt Obligations. All the technical debt is wound up and disguised, and becomes impossible to unravel. Then the project is Too Big To Fail – and the enterprise customer is left on the hook, throwing more money after bad. Traditional outsourcing was designed explicitly for environments that would not change. Any changes of scope are massively expensive. It is my belief that those who accuse agile of “not working” because of “scope creep” are guilty of a bad case of projection; they certainly fail to understand how agile does a far better job of meeting changing user needs. In the deck below I talk about how structural changes in tech and now being reflected in government strategy.
Big traditional outsourcing is no longer fit for purpose. It came from the era of IT Doesn’t Matter. It doesn’t map to any of the trends currently driving us forward as an industry – Agile, Design Thinking, Digital Disruption, The Data Economy, The API economy, DevOps, Minimum Viable Product, Continuous Deployment, getting closer to the user, proliferation of infrastructure choices, falling infrastructure costs, open source, the enterprise embrace of technology built by Web companies and so on. Take CapitalOne for example – which also featured at Amazon’s re:Invent 2015 – it acquired AdaptivePath, a UX agency because it knew customer experience was becoming critical in banking as competition grows. Enterprises now routinely compete with extremely well funded startups investing heavily in software engineering talent.
Traditional outsourcing on the other hand generally pours concrete onto applications and underlying infrastructure, expecting that companies won’t need to change in the 5-10 year horizon.
I spoke with Jay Pullur from Pramati, an Indian entrepreneur, recently about the thinking behind his acquisition of WaveMaker from VMware and he said that one reason a Java/RAD platform was interesting because in his home market Indian outsourcing are only too aware they need to up their game. 80% growth rates have shrunk to around 10%. Wipro reported today Q2 dollar revenue up 2%, EBITDA margin may fall.
The Wall Street Journal has written about this trend – India’s Outsourcing Firms Change Direction as ‘Cloud’ Moves In.
“AstraZeneca PLC is sharply scaling back the business it gives to the Indian outsourcing companies that it has long relied on for tech help.”
David Smoley, AstraZeneca’s technology chief, said he expects to half the $750m the company spends annually on outsourcing over the next two years.
RedMonk has been tracking the ongoing cratering of traditional software licensing fees for some time now but it’s going to get worse for vendors as companies start to unravel their huge outsourcing contracts and bring things back inhouse. Most traditional outsourcing contracts include significant licenses and maintenance fees for traditional on prem software, which also limits operating flexibility. Then of course there is a falling off in services associated with the packaged application market – According to Forbes reporting on IBM’s financial results this week Global Business Services (GBS) “reported a 13% year-on-year decline in revenue to $4.3 billion (down 4% in constant currency), primarily due to declines in traditional packaged application implementation“. [italics mine].
Outsourcing won’t go away of course but it will need to dramatically change. Smart people say we’ve just been doing it wrong, but it is going to be a wrenching transition for an industry that looks increasingly out of step with emerging operating models. At re:Invent Accenture announced a practice taking enterprises to the AWS Cloud but I will reserve judgement before seeing some customer results. Effective services companies for cloud-enabled digital transformation are going to need a completely different set of skills – CMMI Level 5, ITIL, J2EE aren’t going to cut it. They will need to be be design led, cloud native, and help customers to rediscover competencies they have lost over the last couple of decades.
I would look to companies like FutureGov as a model for more effective public private sector partnership. Or see ThoughtWorks, a company built to help enterprises catch up with the latest in application development thinking and methodology. My friends at YLD plan to move up the stack, having begun life as a Node.js consulting shop they are now aiming at broader Digital Transformation – there is some outsourcing and offshoring in the model; Founder Nuno Job now has developers in London and Lisbon, working on customer problems. Business should be working with people like Jeff Sussna, who will help them understand that we should build systems expecting them to break. One company well set up for the change is Pivotal – beyond its Cloud Foundry product business it is one of the most successful agile software development practices out there. IBM is now looking to emulate its model,
— Millard Ellingsworth (@millard3) October 20, 2015
So cloud is hurting Big Outsourcing because it is underpinning new operating models for businesses. Cloud is of course itself a form of outsourcing, but one that allows for speed of delivery, and encourages reshoring of skills. But people and processes changes are needed to do the work. Just because customers have deployment and platform options like Azure, AWS, Google or Softlayer doesn’t mean they get are set up to get the best out of them.
update – check out Toll, cancelling a “strategic” outsourcing deal.