Enterprises and Vendors can build more effective Outsourcing Solutions, based on evidence showing that dividing programmes of work by packages is more effective than dividing it by lifecycle stages and technologies.
Typically, Enterprises whose business is not technology (i.e. the technology is there to support/operate the business aims, rather than being those aims) have an ongoing stream of new development in various sizes, plus maintenance of what’s already there.
When these Enterprises seek to outsource, they manage their Risk by outsourcing parts of the software lifecycle. So typically, one vendor might be given Dev, another Test, a third Support, with Analysis and some level of Design often being kept in-house. Vendors may or may not be allowed to perform overall system integration and project management.
You might also split by technology, so each vendor will be given what they’re strongest in (or at least, perceived as being strongest), and there may also be distributed inhouse teams.
So the result is that you’ll have something like an onsite design team, a SAP dev team in Bangalore, a Chordiant dev team in Kolkata, a Web dev team in Vilnius, a mobile centre of excellence in Glasgow, a test capapility split between onsite and Mumbai and so on.
But this runs contrary to the evidence: where you have co-located teams you get between 3-1 and 10-1 productivity increase.
There are two major factors in this:
The obvious one is communication. Close teams simply communicate better and work more collaboratively. My experience resonates with this; even non-co-located teams within the same company show better communication and effectiveness than transational hand-offs between organisations as they can rely on lots of shared formal standards and informal norms that makes communication much more effective.
The less obvious one is about limiting WiP. Pulling a team of 30 across 10-20 projects, and trying to maximise productivity on a simple utilisation and capitalisation of cost basis is not a basis for any kind of stable environment that leads to learning, specialisation and effectiveness. It leads to all the multitasking problems we’re all familiar with.
This would also allow vendors to build centres of competence in business capabilities, rather than simple technology ones. This would give the people working for those vendors an increase in their Sense of Purpose, as well as adding a second dimension to Mastery, with a result in increasing levels of committment and investment in customer success, and Customer First is always a useful method in increasing sales. This is borne out by what I’ve heard customer after customer plead for: “Understand our business better” although rarely accompanied by the partnering behaviours that would make it likely.
This isn’t to argue for any form of Agile-As-Usually-Practised, by the way. This improvement works works for Waterfall teams just as much as for Scrum, or Kanban, or XP, or any other methodology.
So why do we do this? Why don’t we see more circumstances where the work is divided by project and not by type of work? Even given the duplication between vendors, even a 3-1 productivity increase is a clear net benefit.
While there’s a degree of “The Customer Is Always Right” — at some point, you have to bow to the wishes of the people currently with the money if you desire to make it your money in the future — I don’t believe we as vendors challenge this enough. I like to think my employer is generally better at challenging and putting over a Point of View when our beliefs and experience are in conflict with the customer’s, but in this case, I don’t think we have that conflicting belief at all. We genuinely think — contrary to the evidence — technology and lifecycle splits are non-problematic.
The clear risk is that we end up back in transactional procurement, competitively bidding for individual pieces of work without the partnership relationship required to build the centres of business functionality competence needed to make this approach work. So we need to keep the partnership conversation going and strengthening, seeking to be both the customer’s trusted advisor in in how to deliver best, and their trusted agent in successfully doing so.
All solutioning in Enterprise Outsourcing is based on a dialogue on best meeting the customer’s needs. As an industry, Enterprise Outsourcers need to start paying attention to the evidence and acting accordingly both internally and in dialogue with customers.