Last month we looked at IT outsourcing, and suggested a few rules to help navigate what in any event are tricky waters. We emphasised getting clear about the objectives and outcomes that were desired by entering into a long term contract.
As a significant industry, IT Outsourcing has also attracted a range of ancillary service providers – consultants, advisors, lawyers, and so on, who seek to protect and support their clients, as well as earn a crust. And Differentis is one such company. We are however sceptical about much of what is written, for two reasons. Firstly many pieces of advice appear to be based on anecdotes, which whilst potentially interesting (because “real” rather than “theoretical”) may simply be misleading. Business writers often lack a healthy scepticism, and their standards of evidence fall a long way short of science, and are therefore less than fully reliable.
Secondly, the anecdotes may be misleading because they abstract from the complexity of the services and the businesses to which they are provided; so here we are not prescribing specific courses of action (buy this, make this) but suggesting ways of progressing through the process of making complex decisions.
Design a Services Architecture
You probably wouldn’t develop systems without showing how the individual pieces work together – so why would you reconfigure service elements (who does what) without aservices architecture? How will it work end to end, and how can it be managed?
Know how it will end
Define the exit strategy; very little is for ever, so work out in advance how you could disengage from your supplier(s). Some services are hard to take back without fundamentally affecting the cost structures, but could they be transferred to a third party? Many contracts provide for step-in rights, but do not extend to sub-contractors. The Satyam debacle suggests that business continuity plans need to consider fundamental failures of key suppliers. Due diligence may not be enough, (even if provided by a big league accounting firm). Plan for the worst.
Come to a shared understanding of the risk apportionment
Both parties have to understand the objectives of the deal but they also have to agree the commercial, technical and operational risks they are each signing up to. This can vary, from the provision of widgets and people at one extreme to provision of end-to-end services at the other. Risk, responsibilities, and the required skills of the retained organisation will be very different. Many suppliers who claim to deliver services thus defined do no such thing, forcing back all kinds of risk to the customer. The worst of both worlds is to pay for services but still carry the risk. (Hint: Do not believe everything you are told about “utility models”).
Understand your supplier’s cost structures
Understand how your supplier makes money and recognise you will get a bad (and worsening) service if they don’t. This doesn’t mean you can’t agree incentives to keep costs down or reduce them, but you don’t get free lunches, no matter how many benchmarking clauses you agree. You need to understand the supplier’s cost drivers as well as your own, but this means not incentivising dumb behaviour.
Understand what are you paying for
As we said last time, things will change (volume, demand, technologies). How you agree the basis of charging is an absolutely key issue if you want to incentivise the right behaviours. Paying by the widget, rather than by service, does not attract the supplier to reduce the number of widgets (e.g. servers).
Beware snake oil salesman; there are no short cuts, no standard failsafe approaches, no incredibly valuable boilerplate which will guarantee success. So when deciding what to buy and what to do yourself, be careful; think through the consequences; protect yourself.