M Moser’s Steve Gale reasons that we don’t have to be bad with numbers to get our sums all wrong.
Organisations grow, change shape and sometimes shrink – buildings don’t. So workplace designers always have to fit a living business into a very static and inert building. Freed of this conundrum the design business would be a breeze.
A simple view says that as a business grows, the building should grow with it, which is why we need lycra buildings. A more nuanced view of accommodation encourages us to use flexible working practices so a change in headcount does not put unbearable pressure on the required floor area. But the starting line for calculating an organisation’s space demand is invariably headcount, so how reliable is this number which drives the spreadsheets?
You can usually get a headcount number as soon as you ask for it, but all I know about this number is that it is wrong. If this sounds a bit harsh, it is completely normal for a headcount to be wrong, because there are too many weak points in its construction. It is dodgy on day one, and gets even less reliable as it is projected into the future.
But we do need a best guess of how many people need a roof over their head, and when we get it, we can be sceptical, and tweak it some more, because experience shows it can be wildly inaccurate. There’s no need for an accurate figure, because you can see that would be impossible, we just need a working approximation and a way to design around it. The fuzziness in a headcount estimate seems to come from three places.
First, the prediction itself, no matter how carefully calculated, will change before the finished project is delivered, as expectations and requirements evolve in the months it takes to deliver a new workplace. You could call it scope creep.
Second, workplace occupation is a long game, and we need to design for a whole period of tenure, not just the beginning, and this is often too long for an intelligent guess. The need to look five or more years into the future is a big ask, and yields only approximate data. The rough range of possible headcounts needs to be acknowledged.
Finally, the third area of inaccuracy comes from the often forgotten ingredients in a headcount estimate, but which make a big difference. The number of people that might actually need accommodation is not the same as the number of people on the payroll. We must winkle out the real users.
What about non-permanent employees like temporary staff, part-timers, long term absentees (maternity, sick leave, secondments), agency or contract workers, job-shares or multiple roles? Are there interns, visitors, partners, customers or suppliers needing space? The gap between payroll and user numbers can easily be as high as 100% – and it’s best to avoid errors that big.
So organisations never accurately know how many people will be on their premises, and over the period of tenure this will vary a lot anyway. They know roughly, and that’s all – so even the current headcount is very approximate, and the future one is, let’s say, challenging.
How do you deal with uncertainty? You fudge it. You admit you know a bit, but not everything, and offer a number with an error bar that reflects your ignorance. The target changes from a point to something like a zone, and becomes more hittable.
Once a fuzzy estimate is out there, we can discuss behaviours, workstyles and management practice to allow the business to adapt within its home as it inevitably changes. It would be a lot easier if buildings were as elastic as the headcount number, or made of lycra.
Steve Gale is Head of Business Intelligence at M Moser Associates. SteveG@mmoser.com