The two questions most frequently asked when organisations first contemplate outsourcing are probably 1) how do I decide what functions are best outsourced, and which should be retained in-house? and 2) when should I outsource – when everything is in a mess or when everything is perfect?

What is the traditional answer to these questions?

The traditional answer to the “what” question is: “only outsource non-core processes”. The traditional answer to the “when” question is “never outsource a mess”. But can these rather cliché-ed responses be right, and do they get us anywhere useful?.

So let’s look at “what” to outsource first. How do you tell if a function is suitable for outsourcing?

Describing functions as Core and non-core really doesn’t take us very far. The first thing to understand is what you are trying to achieve from outsourcing – cost reduction, service improvement, process transformation, headcount reduction, access to resources…these are typical objectives. And then to ask for each function – can this best be achieved through outsourcing or not? This includes the critical issue – is there a market for these services, and if not are you prepared to create one (as for example BP did with Finance Outsourcing back in 1991 with Accenture and later PwC). And if the answer to this is yes then finally to overlay the question – are there any overriding considerations which would change this – loss of brand, customer relationship, or IPR are again typical issues. If the answer to the final question is no, then the function is a candidate for outsourcing.

The traditional view that organisations should only outsource non-core functions is invalid then?

Whether a function is “core” or not doesn’t come in to the analysis above, which is mainly driven by “outside-in” thinking…can someone out there carry out a given function better than you without damage to other parts of your business? It avoids the navel gazing and analysis paralysis typically triggered by trying to decide if a function is “core” or not. And the “core” argument doesn’t really reflect what we see in the market: most people would think that dealing with customers is pretty core to most businesses, but customer contact is routinely outsourced. Ditto aspects of R&D for a pharmaceutical company, or cheque clearance for a bank.

And regarding the “when” question – when is a good time to outsource in your view?

The very question pre-supposes that there is some perfect time in an organisation’s life when every process is stable and works and there are no other conflicting initiatives – such a time, of course, never arrives. And as for the traditional answer, it begs the follow-up question…if the process isn’t a mess, why are you looking at changing it? A better answer would be “Never outsource a mess without defining in the contract how and when it will be fixed”.

What is the best way to answer the timing question then?

The way to answer the timing question is to look again at what you are trying to achieve through outsourcing, and what priority that has for the business. If for example the outsourcing will deliver significant savings this year and that is what the business needs, then that would lead you to acting sooner rather than later. But if the business priority is, for example, to stabilise processes and service and that is already underway, then the outsourcing may need to wait. The key is to recognise what outsourcing will deliver at what cost in time and effort, and how that fits with the business strategy.

Are there any other factors to consider regarding timing of outsourcing?

The other key factor is to recognise that outsourcing is always a major change project, absorbing senior management and operational management time. So that time has to be available. And where processes need to be moved the main consideration is that they should be as stable as possible at the time of the change. So for example for both these reasons if a major ERP implementation process is planned, any outsourcing should probably be timed for before the roll-out or after the roll-out, but not during.

And your overall message here?

The overall message – as always don’t look at outsourcing as an esoteric magic wand. Understand why you think it is a valid tool to meet your needs, then think through the implementation thoroughly – doing that will answer both what and when.