The law of large numbers......
Roll a dice six times and the probability that it lands once on each number is extremely low (1.5% in fact). Roll a dice a million and two times, and the probability of having rolled each number a sixth of the time is nearly 100%.
Toss a coin four times, and the probability of landing two heads and two tails is 0.1. Toss a coin a million times and the probability of tossing heads half the time is nearly 1.
An outcome is more predictable when large numbers are involved - the law of large numbers dampens the random effect.
It strikes me that this mathematical law has particular relevance if you work for an established, mature organization and are looking to innovate and experiment with emerging technologies.
If you are a 'corporate innovator' - how many experiments or innovation pilots do you need to run in order to be confident of success?
Sadly ... a lot. You can apply all the normal business assessment rules to any new area - develop a business case, research the market, assess the viability and scalability of the technology, calculate the NPV and ROI. All good, sensible, important things to do.
But...... if you are experimenting with new business models and innovating with new unproven technologies ... you can develop all the business cases you like, but you're not really going to know whether its going to work until you've found out through experimentation.
I'm not saying that success is purely random - but it is uncertain. You can influence, but you can't force, customer sentiment. You don't know what alternative disruptive technologies are being hatched up in a kid's bedroom in Tallinn. You don't know what your competitors are working on - right now while you are working on your new thing. You don't know if something genuinely new and ground-breaking is going to work until you've tried it.
Which makes innovation a huge challenge for traditional corporates. Applying usual business rules is a recipe for failure. Standard approval gating processes tend to result in organizations investing in a small number of (sometimes expensive) innovations - in which the success of each matters too much. And yet the chances of failure are high.
Those leading the experiments fear the consequences of failure so hang on to a dead idea too long. The sceptics in the organization will use the failure of any experiment as a reason to ditch experimentation all together. The heavy hand of business as usual dampens the enthusiasm and limits success.
So how does the law of large numbers apply? I'd suggest three simple principles:
1. Recognize that failure is normal - you can't guarantee that every experiment is going to work. Embrace failure, reward those who were willing to take the risk and learn from it.
2. Don't bet the farm on any one thing - 'business as usual' is today's business so don't neglect it. Any new experiment might be the next big thing - but it might not. So investment should be proportionate to the risk of failure (probably higher than you think) and the size of potential opportunity (probably lower than you think).
3. Experiment A LOT - try lots of things, get to a minimum viable product or proposition as quickly as possible and don't be afraid to kill it quickly. 'Normal' business rules don't apply and volume matters....
I'd be interested to hear examples which prove (or disprove) my theory.... any black swans out there?
Transformational CRO | Driving Revenue Growth for SaaS/B2B Startups | Expert in Go-To- Market Strategies
1 年Mark, thanks for sharing!
Senior Consulting Partner
8 年I'm not sure organisations are set up for this experimentation or have the skills to test really drive a culture of experimentation. Maybe like all major studies a good pilot area is worth thinking about as a starting point ?
Digital Artist & Creator
8 年If you haven’t already seen it worth looking at astro tellers talk on ‘the unexpected benefit of celebrating failure’ https://www.youtube.com/watch?v=2t13Rq4oc7A it is about approaching each experiment or idea with a view to trying to kill it quickly, and realising if you do, you have minimised the investment and released resources to move on. There is also a good book on how cheap experiments are better than good ideas, good ideas are often pitched by people that are confident in their success and not always able to let them go if they start to go awry, often leading to ‘well it seemed like a good idea at the time!’. Low cost, volume (innovation or workplace) experiments play to the numbers game, there will inevitably be failures, but if you minimise the investment and maximise the volume – the benefits of the successes easily out way the cost of the failures. Both examples support your theory – you have to play the numbers, high volume low cost and failure is expected and not negative – it allows for quick release and progression on to the experiment . . .
Chief Revenue Officer at MeatBorsa
8 年I reckon I've got a few in my back pocket for you, Mark ;-)