Building a car which cannot speed
By @edent
· culture leadership MSc · 6 comments · 1,000 words
As part of my MSc, I'm reading a lot of "Leadership" books. They're all pretty bad - but they have one common thesis; it is essential to improve your company culture.
I'm not sure if I agree. I feel completely divorced from most forms of company culture. I find the way that these books talk about changing people is pretty creepy and disingenuous. That's my problem, not theirs. I prefer to look at processes and systems.
Here's an imperfect analogy. What's more effective - teaching people the perils of driving too fast, or building a car which cannot break the speed limit?
We have decades of evidence that people regularly break the speed limit. We have advertising campaigns, driving lessons, tough penalties, automatic enforcement, warning signs, and dozens of interventions. And still we have huge numbers of deaths on the road where speed is a factor.
I'm sure that some people's attitudes have changed due to the above - but the culture change simply hasn't taken root in a large number of the driving community.
One of my previous cars had a nifty feature. Through a mixture of GPS and computer vision, it always knew the current speed limit of the road I was driving on. The dashboard would show me my current speed and the road's speed limit. If I exceeded the speed limit, it would start making an urgent pinging sound. And it didn't stop until I dropped to the posted limit.
So I built a system. Every time I drove into a new speed limit zone, I'd hit the "speed limiter" button. It's kinda like the opposite of the cruise control button. It prevented me breaking the speed limit.
Imagine if every car on the market had this as a mandatory feature. If it were physically impossible to speed - that would be more effective than trying to change people's attitude and culture.
A system can do what a culture cannot.
Culture Vs Systems
As Peter Drucker (never) said "Culture eats strategy for breakfast".
But I think systems create meaningful change faster and better than culture can.
In 2015, a sexual health clinic send an email to 800 people without using BCC. That means hundreds of people had their HIV status exposed without their consent.
Cultural Change might have reduced the risk of this happening. Make sure people are trained, stick up posters reminding people what to do, keep confidentiality high on the agenda. But eventually someone would have slipped up. Everything is high priority - and we can only remember so much.
But a Systems Change would have prevented this error. It could have been an email front-end that disabled the CC function, or a mail server that forced all external mail to BCC.
A culture can't stop people from making mistakes.
Systems can be overridden
The purpose of a system is that it makes it easy to do the right thing, and difficult to do the wrong thing. Sometimes, it is necessary to do the wrong thing. For example, if a ravenous tiger is chasing after you, or you urgently need to get to hospital, it may be appropriate to speed. In my car, I could hit a button to disable the limiter.
I don't mean training users to mindlessly click "OK" a couple of times to bypass a system. I mean creating a structured and auditable way to justify going outside the system.
Systems as Culture
You can argue that this is culture. A culture where people trust a system and that system is designed to support and protect them.
Sure, building a culture where people surrender autonomy to a system might be difficult. Helping people understand why a system is designed the way it is, is also important. And not everything in compatible with a rigid system.
I see the argument that culture is critical to success. But there's no substitute for creating systems which help people consistently perform well.
Get out of my way
We want to build a culture where people can "fail fast" and "move fast and break things". I agree with that sentiment. But the only way that culture works is if systems are in place to prevent people from failing in their legal duties and stops them breaking critical infrastructure.
The power to experiment and play only works if you're in a safe environment.
A safe environment cannot come about by culture alone. We need systems which let people fail without hurting others.
Thoughts from others
You can’t escape teaching people about ethics through what is essentially technology-enabled paternalism and hope they will continue to know how to make the right political decisions and won’t move to full on authoritarianism.
Things like this only work as long a they’re advisory and can still be overridden by the human and turned off completely if necessary. See the recent Boeing MCAS fiasco for a relatively extreme example of what can happen when it goes wrong.
Really interesting blog post. I'm in the camp that says "systems is culture and culture is systems". People need to be able to build, maintain and where necessary dismantle systems which make their work better quality.
Think this is a great idea! Also great for the environment, as cars use a good 10% more fuel driving at 80mph vs 70mph
Tim M says:
In education, and presumably also in work, there's a very useful distinction to be made between intrinsic and extrinsic motivation.
Intrinsic motivation is "wow, that's cool and interesting, I wanna read more about this later, I wonder what happens if we change this and try it again?"
Extrinsic motivation is "ah shit, I have to hand this in by tomorrow or I'll get a detention/the teacher will be annoyed/I'll fail the thing and have to waste time doing it again". Both carrot and stick often only bring about extrinsic motivation; "if I do this we'll, I'll get a prize/a certificate/a pat on the back from the teacher" – it's only effective for a long as there's someone handing out prizes and certificates. As soon as that stops, what's the point any more?
You seem to be advocating for the latter, which makes it hard to agree with you on this occasion.
@edent says:
I don't think that's what I'm advocating for. I'm saying that removing the motivation is better.