Get data, get loads of it, heap it up, as much as we can and when we think we’ve got enough we’ll go and get more of it because we’ve convinced ourselves that the more data we have the more chance we have of being right and the more our backsides are covered if we get it wrong. It’s a ‘win-win’ in old money. More is better.
Workplace is disappearing under just such a burial mound, its own Sutton Hoo.
Like the 9th Century monk Nennius who said he had ‘simply made a heap of all I have found’.
Then you analyse it, cut and dice it and turn it into a report and some charts and you can quote from it knowingly and build some confidence from the investment. All good so far.
Yet for those who watch A Question of Sport – what happened next?
There are a couple of fundamental and related problems with Big Data.
First, the human being has to decide what to do. Monster digital agglomeration and segmentation is followed by good old-fashioned linear thinking with all the vagaries and bias that entails.
Your utilisation data comes with a host of ‘you might do this or that’ (and often ‘we can help you’). Your colleague survey data ‘points to’ a few responses you could make, if you want to but it’s your call and of course your responsibility. Those who have sanctioned the not-inconsiderable investment in collecting the data want to know what you’re going to do with it all, given you made a passionate case for obtaining it. Doing nothing is invariably not on the table because it couldn’t be countenanced that having spent the wedge you’re doing just fine. In the world of data analytics, no-one’s doing just fine, are they?
Just because we have collected and analysed wheelbarrow-loads data, it doesn’t mean we know what to do with it. We need solutions. They’re difficult. Our linear-programmed minds find it just as taxing to work out what to do with all this analysed data as they did before it arrived. And by solutions it doesn’t mean simply digitising the thinking we undertake to make it look prettier. It’s like the Ant Hill Mob – they had a big car, but to make it move they stuck their feet through the floor and ran with it. Creating applications that can arrive at solutions is where it gets problematic, time consuming and expensive.
Secondly, at the point at which the brakes squeal and the rubber burns we also realise, somewhat uncomfortably, that more data isn’t automatically an advantage. Data has to have a purpose, has to be useful. Otherwise even though its good and clean, and perhaps even chin-strokingly interesting, it’s useless.
Those workplace utilisation studies are a classic example. Workplaces are ridiculously predictable places even in the wacky world of agile, with consistent occupation patterns based around a working day that has been the same in essence for decades. The soft camel curve of occupancy peaking around 11.00-11.30am and 3.30-4.00pm renders the rest of the data irrelevant, as we need to plan for workplace capacity at the peak. Not the average, or what happens at 9.05am, but the maximum space needed by each team at any time. Just that. The overwhelming proportion of those bytes the under-desk tachos are collecting is useless too. Their little red eyes are staring in disbelief.
Most organisations still determine they need their own data to be sure, which ends up looking remarkably like everyone else’s data. No-one is that different. We also used to collect ‘signs of life’ (SoL) – desks with stuff on, and no warm body. Yet the only thing that results from having SoL data is a policy or a protocol telling people to clear up if they leave it there for two hours or its bin-bagged, which just annoys everyone. It’s rarely an adult conversation, despite your intent when you set out. Finally, data is often collected on what people are doing at their worksetting, in the belief that it can help determine what settings light be required in future. Which means using a laptop, using a laptop and dock, or talking to someone or on a phone, or talking to a real person. What else do people do in a workplace?
In your colleague survey 59% of people aren’t sure about the colour of the carpet in reception and 47% of people would like some more plants. Do you change the carpet and but not buy plants? You’re just over the average for being able to adjust the temperature. Is that a good thing – based on what you expected to see or would tolerate as acceptable?
We are still terrified of confidently opting for less data when there is an opportunity to have more. Simply amassing more won’t save us. We need data, but the right data. Less but better. And it needs to have a purpose, a practical use – and be available to applications that can generate solutions from it.
Or we’ll be singing our overdubbed-to-oblivion ballad into the sunset, wondering why what was promised never arrived. And nothing will have changed.