What Does It Mean ‘To Experiment’?

This work is part of a series of pieces around ‘The Experimental Organisation’, which together explore the practicalities of how an Organisation can be curious, and unlock learning at scale. Or indeed, to see if such a thing is even possible: when culture tends towards safety, and ‘experimentation’ veers towards the unknown.

An ‘Experimental’ Organisation is one with a deep seated capability to learn, through a variety of approaches, from the formal, visible, and highly structured, through to the informal, hidden, and loose mechanisms of individual activity and collaboration within Communities.

An Experimental Organisation is one that knows what it means ‘to experiment’, and is clear when, how, and why, it does so.

Whilst many Organisations talk about the need to experiment, they are not always clear which of two primary definitions they are referring to.

An Experiment may be a structured approach to exploration by which we gain insight: the formulation of a hypothesis or position, typically based upon research into what came before, and the subsequent testing of that hypothesis, or by trying a novel procedure under controlled conditions.

Or we may talk about an Experiment when what we mean is to try something and see what happens as a result. A process of wading in and seeing how it goes. When i ‘experiment’ with making a sourdough loaf, i’m largely doing the latter. Although if i took my first failure and carefully tested each element in turn (ratio of ingredients, temperature of the oven etc) i am veering towards the former.

Both approaches are valid, at least in an Organisational sense: if we ‘teach’ people something, we may wish for them to ‘wade in’ and try it out (although some may call this ‘prototyping’, or ‘rehearsing’), OR we may seek a more structured approach to gather data, gain insight and generate understanding of a broader problem.

It’s generally easier, in an Organisation, to just wade in: partly because that is an inherently easier thing to do, and partly because Organisations are complex environments, within which it is remarkably difficult to learn anything in anything like a structured way.

In the language of science, Organisations are messy, and it’s very very hard to control that mess: and experiments rely on us ‘controlling’ certain conditions.

If i am trying to learn to juggle, i can just ‘wade in’ and try it. Or i could watch a YouTube video. Or you could teach me (assuming you can juggle…). If i wade in, and succeed, we can say with confidence that my ‘wading in’ led to my success. If i watch the video and then manage to learn, we can say that the video ‘taught’ me.

Or can we?

We have already ascertained that i could learn by ‘just trying’, so simply getting the same outcome by also watching the video is not enough. But say i learnt in half the time, that would be significant. We can half learning time by using videos!

Or can we?

The problem there is that now i can juggle: i taught myself by just wading in. So i cannot ‘relearn’ it with a video. You need someone else to have a go. But someone else is not me: perhaps they were great at cricket as a child (i was not…), so they have some legacy of hand/eye coordination. They are not me. So we have the first factor that we cannot control: people are different.

Alternatively, imagine you try teaching me, and i have a go myself, and i watch the videos. Maybe i read a book about juggling too. And i learn in half the time of a control group who just ‘have a go’. Something went right yes?

Well, yes, it did. But what?

In complex environments, with many changing factors, it’s hard to know ‘what’ went right. You cannot control the conditions. And if you cannot control the conditions, you cannot really learn ‘what’ went right, or ‘what’ went wrong, to replicate it in any meaningful way.

It’s not to say that you cannot drive an effect: you may well get an effect (i can juggle), but not know ‘how’ you got it, which tends to lead us into the worlds of magic and superstition (i scratched my left ear before my last job interview and hence if i scratch my left ear again, i’ll get that next job…).

But there is more to the challenge than simply understanding what an ‘Experiment’ is (and choosing which approach you wish to take): there is a second layer of complexity in that Experiments (of any flavour) take place within culture, and culture is typically excellent at doing what it knows how to do, and sometimes terrible at learning how to do new things.

To give you a sense of this: over the last two years i have worked on programmes in over a dozen Organisations that have involved participants running experiments. In around half of those organisations, they have been utterly unable to do so. Not because they were not bright enough to scope and run them, and not because the Organisations themselves did not wish them to learn, but rather because facets of culture bled away the momentum, caused a thousand cuts, or simply delayed, discussed, and conferred their way to stasis.

Which makes sense: Organisations are not built to change. They are built precisely not to. They have a specific competence to continue to do what they do now, HOW they do it now. So ‘change’ is bad enough, but the type of unstructured, exploratory, untidy behaviours of experimentation are particularly toxic.

Even if we are able to address the cultural challenge, there are further layers of complexity to unravel: maybe i CAN run an experiment, and maybe i learn from it. But what do i do then? Do i just go away and ‘perform’ by myself. Or do i teach my friends. Or do i somehow loop this local, tacit, tribal knowledge back around into the Organisational realm itself? Is there a mechanism by which my knowledge can help the Organisation to learn?

Well: this reflects back to the core challenge: why would i share, and does anyone want to listen? A scientific experiment is only useful it it is peer reviewed and published. If Newton had discovered gravity, but only used that knowledge to ensure he did not fall asleep under apple trees, we would not have been much better off in a collective sense.

So beyond the rhetoric, what would an Experimental Organisation look like?

• It would understand what is meant by ‘An Experiment’, as both a structured, or exploratory, event.
• It would hold the specific capability to support those who wished to learn, by either mechanism.
• It would have a culture within which we could find the space to do so.
• It would have the space, support, and recognition, to share the learning, and recognise those who learn.It would have the ability to value ALL efforts, not just those that show a prize at the end: so failure, and success, would be held equally IF that effort took place within an agreed structure. In that sense, they do not avoid risk, but they remove the reputation based risk of learning and failing.