Imperfect data – perfectly useful

Kai Hopkins

Kai Hopkins

Having seen a recent and rather controversial weight-loss advert here in the UK, I decided, along with the majority of others I suspect, that I was not ‘beach body ready’. I didn’t really need the advert to draw the conclusion, but none the less I decided I needed to do something about it. More specifically, I decided to do more exercise and to find out once and for all if I too had a rippling six-pack buried underneath my ample midriff.

As with so many things in life, I felt it was a basic lack of information that was to blame – sure, I feel walking to the pub is a really long way, but is it? Am I really in the gym as much as it feels like I am?

So I downloaded a free app called Pact. The premise is simple; I make a promise to exercise a certain number of days every week, and I have to pay $5 for every day I miss. All the money paid by the lazy layabouts is then split among the self-righteous who do achieve their targets. On the surface, it relies on an inherent incentive we all respond to – money. I get about $1 for every week I complete, and risk losing $5 for every day I miss. But more importantly, it allows me to track, in real time, how much I am actually exercising.

The advert certainly provoked a reaction…

The advert certainly provoked a reaction…

It can count my steps up to the daily target of 10,000, or can use my GPS position to know when I am in the gym, or connect to a whole range of other monitoring systems to tell me what I have never properly known – how much I am actually doing?

Whenever I tell anyone about this app, they have two observations.

First they ask: ‘can you cheat?’ And the answer is yes, very easily – I can spend an hour in the all-you-can-eat buffet directly under my gym and it will count as a workout (By the way, who puts a gym above an all-you-can-eat buffet?!) or I can shake my phone gently while sitting on the couch and it will think I am power talking through the park. All will allow me to hit my targets, but at the end of the day, I am only cheating myself. My aim is to increase how much exercise I do, not make money (which at $1 a week might take some time anyway).

Second they ask: ‘how accurate is it?’ And here the answer is mixed. While walking in a straight line it counts steps well, however it sometimes also counts me bending down or reaching for another bar of chocolate. Also, while it might count my time in the gym, it doesn’t measure the intensity of my workout.

So if this data is as imperfect as it appears to be, how can it be of any use? And the answer is that even imperfect data can be useful. It makes me consciously think about my daily routine, where as before I didn’t; it makes me actually plan to go to the gym rather than just end up there as if by chance; it makes me go for an evening stroll if I have not yet hit my 10,000 target – it has simply and cheaply elevated my exercise routine in my conscience and in real time prompts me to do more. If I had enough time and money, I could measure my exact heart rate, calories burned and inches moved but time are not always in abundance. It helps me manage something I can readily influence in real time and while it may not be the most perfect data, it is still very useful and can be used to modify behaviour for the better.

Now re-read that last sentence…

It helps me manage something I can readily influence in real time and while it may not be the most perfect data, it is still very useful and can be used to modify behaviour for the better.

Now, instead of picturing me panting and puffing in the gym, think about your own lives and the work you do. Think about your own goals, and whether or not you have the time and money for perfect data. If not, perhaps take a look at the imperfect – it might tell you more than you expect.

That is what our experience at Keystone has taught us – that in the absence of inexhaustible time and money, imperfect data can still be actionable and, indeed valuable. And the more we collect it and use it, the more robust it becomes.

Our Constituent Voice™ methodology is designed to get affordable quality real-time perceptual feedback and convert that into data about an organisations’ development performance. This not only enables them to improve how they work, but also improves the relationships they hold with those around them. Moreover, it has built-in mechanisms to minimise cheating and improve accuracy. Public discussion and sense-making of the data and triangulation with other sources of evidence enhances validity and helps generate creative ways forward through dialogue with constituents.

If Constituent Voice is the social change equivalent of Pact, then watch out for a leaner, fitter development sector.

Leave a Reply