Thursday, December 3, 2009

To backup, or not to backup – that’s the question

Subtitle: Backup in an Organisation – the Specialists View
What’s in a word: Backup carries positive connotations – redundance is perhaps neutral – duplication is bad. And what sort of backup am I talking about here? Not hardware, not data, not the usual meaning of the word - but rather organisational backup of human competencies in specialised functional areas. How much formal backup of that sort do an organisation need, if any?
From my experience I see three main organisational approaches, viz. -
· Backup level 0: Some specialist activities are outsourced and the corresponding competencies located outside the company. The actual contractor may, or may not, have backup. From the company’s viewpoint all kinds of backup are to be found outside the organisation, - ”in the market”.
· Backup level 2: Some specialist activities tend towards strategic significance and the corresponding competencies are systematically identified and resources duplicated within the organisation.
· Backup level 1: Some specialist activities are handled in-house without duplication of corresponding human resources and activities. There is, effectively, no formal backup. But is that synonymous with ”missing a Plan B”? Though that is sometimes the case, it is not necessarily so. – Arguably, Plan B is either to rely on finding specialist competency backup “in the market”, or to develop in-house competency ad-hoc, if needed; in both cases with a likely significant overhead cost in time and money. - The key challenge is to identify this situation in advance and to deliberately accept it where appropriate – and not appear to be overtaken by unforeseen events when they happen.
It is important to note that in the right circumstances all backup levels – 0, 2 and 1 – are valid and correct organisational choices. But they are not all equally popular and “in vogue”. Outsourcing is in vogue. Duplication is - to some extent – in vogue, where it can be justified on the basis of strategic importance and the organisation’s ability to pay. But a level 1 “no backup” setup is [too] often just considered an old-fashioned, irresponsible and undesirable result of managerial negligence and a candidate for deliberate transfer to level 0, or level 2.
I would say the relative popularities reflect middle managements point of view: Backup levels 0 and 2 represents, respectively, pre-emption and prevention as management tactics for specialist functions, whereas backup level 1 leaves managers at the mercy of their specialist subordinates, – or so they may feel.
But how does the middle manager’s view compare to an economist’s (or CxO’s) viewpoint? While it is true, according to Murphy’s Law, that all things go wrong once in a while and a few things go wrong almost all of the time, it is not true that all things go wrong almost all of the time. At the same time the “responsible” backup levels 0 and 2 both carry significant economical overhead compared to the “irresponsible” tactics of backup level 1. Consequently, there is a real economic trade-off between the chosen level of backup and the cost of operation. It is often argued that backup levels 0 and 2 make an organisation more resilient (or robust). But one could equally well argue that often a resilient (or robust) organisation should easily be able to sustain the occasional setbacks associated with level 1 backup, while most of the time being able to profit from a smaller and more effective organisation.
And how does the manager’s view compare to the specialists viewpoint? Mixed. Some specialists prefer to be external consultants to customers rather than in-house specialists to an employer. But while many specialists like to work closely together with certain colleagues, I believe few really like to work one-on-one on everything with somebody else. It is a matter of perceived short-term efficiency and of few specialist professionals being really “equal” in specific competencies. Each has his, or her, preferences, strengths and weaknesses – and often one will be racing forward and the other be struggling, relatively, to the dissatisfaction of both. And it will always be the runners-up, the backup, and not the leading and most competent of several specialists who sets the pace. This also translates to a cost to the organisation, both directly and indirectly in the form of reduced specialist employee work satisfaction.
In conclusion, where does these thoughts take us? They imply that trust should be promoted in the organisation - trust between top management and middle managers of functional specialists, and between middle managers and subordinate functional specialists – in order for backup level 1 to be a viable solution where it is economically optimal in an organisation.
Top management must trust middle management in order for the latter to take on the risk associated with backup level 1, and middle management must trust specialist subordinates in order for the latter to be allowed and able to apply their skills optimally in an environment where deliberate and informed application of backup level 1 is an acceptable option.Finally, I just like to add that very few specialists – or people in general - are truly “indispensible” in an organisation – particularly in a networked organisation. It is my experience time and again that the gap left by even extremely “important” people usually closes much sooner than anticipated. There are exceptions, of course. Curiously, they usually appear in the wake of unconspicuous people in quite subordinate positions in a heavily networked environment. Normally, persons leaving an organisation are willing, or can be persuaded, to help facilitate the ensuing transition. If that is not the case due to illness or accidents, everyone else is usually willing temporarily to make an extra effort. Only rarely are organisational fissures associated with outright animosity, since it is expensive to all parties, and even such occasions can usually be handled at a lesser expense than maintaining extensive organisational backup in the long run.