The job that a contract has to do ought to be simple, even when the suppliers have to manage lots of complexity to deliver the services.
A contract will say what responsibilities the supplier takes on, and what they will be paid for doing so. Of course there are caveats and exclusions, and assumptions, and customer provided elements, and all the usual paraphernalia of a contract, but ultimately it comes down to the question: What is one party obliged to do for the other?
The problem is writing down exactly what you want and exactly what is supposed to happen when things go wrong. We’ve talked before about contracting with suppliers to sign up to outcomes, not inputs ( here ), and last time whilst addressing target setting and performance management we said that there was a real danger that the targets used to drive behaviour could backfire, sometimes spectacularly. People chase targets, in particular if they are likely to suffer financially from not hitting them.
This is made even more complicated in the case of an outsourcing arrangement, as not only are the supplier”s targets enshrined in the SLAs but they are likely to be subject to service credits, thus entrenching input–related measures with formalized financially driven targets – the worst possible combination.
So what can you do to increase the chances of getting what you thought you were buying..
There is a dividing line (which may be hard to draw in practice) between the services that companies want to perform themselves for a range of reasons (security, IP, excellence, focus, and so on) and those where there is a fungible deep market which can be exploited pretty much at will (lawyers and office cleaners for example).
In the middle – indeed straddling this divide – are the large complex outsourcing arrangements, where a whole range of services are provided and architected by the supplier. Early customers were forced to give away a range of control in order to get the economies of scale they craved; this in turn has led to a desire to buy services (and not widgets), but reality still limps behind the wish. Buying services is less problematic if they can be easily defined, especially if switching costs are low. Remember, the whole service credit and SLA regime was driven in part by a recognition that switching costs would be very high, and so customers needed some leverage over their suppliers short of termination. But complex services are just that: complex, and therefore hard to define. And the more complex they are the more they are likely to fall foul of the whole measurement regimes we talked about last time.
As anyone who has ever negotiated a complex contract knows, the devil is not simply in the detail but in the unforeseen detail. And both parties may understand perfectly what each are trying to get at without ever finding the means of writing it down. You might think that the one thing a lawyer should be able to do is close down those loopholes, and they often can. But they aren’t omniscient, and the client is often working to a budget, and the day has only so many hours in it to come up with recondite and relatively unlikely scenarios and then put language in place to deal with them.
This appears to be an impasse – it seems practically impossible to determine how unforeseen events will be managed within a contract, yet adding more and more specifications of what has to be delivered (in the SLAs) can lead to the wrong behaviours. We sometimes hear the much abused word ”partnership”, but usually only when one party wants something from the other; otherwise there is little need to mention it. Of course in any relationship there is give and take, but what can be given and what taken is still subject to a calculation; and beyond certain limits one party will walk.
So what can be done?
Firstly, both parties have to put in place some robust governance to address any fundamental issues that they have, and the formal mechanism for this must be addressed in the contract. But how it is actually lived cannot itself be the subject of a contract (and so it can’t be left up to the lawyers).
Secondly, both parties should agree, at a senior level, to describe what success looks like from both perspectives, and to lay out a set of principles they will adopt in managing the contract. These will not just describe the services (which will be detailed in the contract) but the expected and desired outcomes, as well as those outcomes neither party wants. This of course isn’t a contract, but can be useful to pull out at times of stress. Forcing key players to articulate and discuss what they were looking for from the deal, makes it harder for either parties to subsequently complain that they didn’t know what they were getting into.
Thirdly, the current or proposed SLA regime should be reviewed to understand whether it drives the right behaviour or not. Is it too complicated? Will people just try to hit the numbers and forget about everything else? Does the supplier ever suggest that a particular SLA – or perhaps the service credits that go with it – doesn’t actually act in either party’s interest? Or would this be seen as special pleading by a customer that just wants the supplier to deliver the damn services? Everyone knows that the point of the credits is not to recompense the customer (and we can forget covering consequential loss), but do they help the right people do the right things, or are they just built into the P&L of the account, and therefore fail to influence what happens on the factory floor?
Fourthly, consider what services can be (or soon will be capable of being) delivered in the Cloud, because most of these genuinely do look like services, and switching costs are relatively low. They are less likely to require complex contractual wrangling, lawyers or consultants. Fundamentally Cloud computing is a commercial proposition, even if it requires technology of a particular kind to enable it. The service levels are simple, and concentrate on availability, rather than on any abstruse system of measurement. The economies of scale are open to view, and the switching costs are much lower than trying to get out of an outsourcing deal with a full service provider.
So, a contract will only take you so far, and the fact you can’t predict everything is something you will have to live with.
Contracts do influence behaviour, but the better ones include operational guidelines and examples of how they are supposed to work to make them less contentious to apply. Continuity of management on both sides helps, but is hard to assure and impossible to enforce. There is little substitute to agreeing (and documenting) at the highest level in both organisations the aims of the contract.