Company culture has two components. One is what you're trying to get people to do and the other is how effective you are at getting them to do it. An effective culture requires you to nail both. The two aren't necessarily correlated - you could be good at either, or both, or neither. However failure on either count makes the culture useless, or worse. Either you're very bad at pointing people in the right direction or, worse, very good at pointing them in the wrong direction.

‍

Generally people focus on the second part but in most cases I think this misses the point - the far more interesting question is exactly what kind of culture you're trying to build and why. But this doesn't seem to get as much attention.

‍

That's a shame because in the best case the first part is much more interesting. In deciding how you want a company to behave you're implicitly saying a lot about the circumstances you think the company will face and how best to adapt to them. That's far more interesting than the generic question of how to drive widespread adoption of a set of norms once you've decided what they should be.

‍

It's also the reason why the best and most interesting company cultures, far from being eye-roll inducing corporate jargon, are exquisitely precise instruction manuals for how a company needs to behave given these circumstances. The most interesting and unusual cultures imply interesting and unusual circumstances in which companies must operate.

‍

At its core a culture is a set of norms and behaviours that guide the thoughts and actions of a group of people. Companies are a particularly good example for thinking about cultures because theirs are often intelligently designed and explicitly defined. This isn't true of many other cultures. Cultures that emerge at the level of a nation state, religion or university for example, are generally more messy and less tightly defined. This makes sense as they're more likely to be a complex product of years of evolution, codified in a messy form, if at all. But a lot of what is true about company culture is true of these other cultures and vice versa, so companies are a good place to start.

‍

The most effective cultures are sets of norms that genuinely change the behaviour of people within the company. In some sense the usefulness of the culture can be boiled down to the magnitude of the difference between the way the employees of the company would have acted if the culture didn't exist and the way they act given that it does exist (assuming of course that that difference is in a positive direction).

‍

If someone was already going to act in a certain way then you don’t need to bother building a culture to encourage them to do so. But if someone acts in way X as a result of culture who would have acted in way Y had the culture not existed then the culture is impactful. Indeed the value of the culture is directly proportional to the magnitude of the difference between X and Y.

‍

This gives us my favourite single test for whether a culture is likely to be useful or not:

‍

Can you think of a reasonable situation in which this culture would be inappropriate?

‍

Or, stronger:

‍

Can you think of a situation where the opposite of this culture would be good?

‍

The power of the question is simple - the more specific the culture is, the more it guides people to act in a way that is suited to a particular set of circumstances, and therefore the less well suited it will be to other circumstances.

‍

By swinging this the other way around, if you can’t find a situation where your culture wouldn’t be a good idea you don’t have a specific set of rules for your circumstances, you’ve got a generic set of platitudes that should apply to all situations and therefore that have no particular relevance to your situation. You’re encouraging people to do what you’d already have hoped they would and have therefore have provided no specific information about how you’d like them to behave. [1]

‍

Contrast Facebook’s early culture of ‘Move fast and break things’ to the culture of one of the hospitals I trained in of ‘Deliver compassionate excellence’.

‍

Sure, it’s easy to roll for people to roll their eyes at Facebook’s culture as an example of unaccountable tech bros gone too far but it certainly passes both the tests above. In a few words you’ve got a concise and powerful set of instructions, reward criteria and defence against failure all in one go. It concisely explains to all employees that Facebook values speed above all other things, and in the inevitable tradeoffs that will occur during their work they are therefore to prioritise speed at the expense of everything else, particularly reliability. This is not an intuitive thing to do. And therefore it encourages employees to act in a way they would not have otherwise.

‍

When programmers are making tradeoff decisions about how quickly to launch new projects the culture gives them an answer. When promotions are being considered there are a ready set of guidelines by which the best performers can be identified. And when things do inevitably break those responsible are rewarded for prioritising speed rather than punished for taking risks that ended up blowing up. [2]

‍

This is an exquisitely precise reaction to the challenges facing an early stage startup. Namely that the biggest risk is moving too slowly (and therefore running out of money or being out-competed), any errors can be quickly corrected, and the upside from a few good ideas will easily outweigh the cost of failed experiments. These are a very precise set of circumstances that don’t apply to most organisations. Therefore Facebook’s culture is inappropriate in the vast majority of circumstances.

‍

In most neurosurgery for example one of the biggest risks is accidentally damaging the healthy brain, any mistakes are likely to be permanent and generally a botched operation is a much greater tragedy than a successful operation is a triumph. Move fast and break things is not a good culture for a neurosurgery department!

‍

If we test the stronger framing of the question (reversing it and asking if we can find anywhere where the result would be a good culture) neurosurgery clearly fits the bill. If we invert 'Move fast and break things' we get 'Take as long as you need just don’t make any mistakes' which is (within reason, for many operations) pretty much the perfect advice for a neurosurgery deparement.

‍

Notice that the same thing doesn’t apply when we test the 'Deliver compassionate excellence' culture of the hospital. This is pretty much a generic statement about how you’d like all people in all organisations to behave at all times. I certainly can’t think of many organisations where 'Be really nasty and incompetent' would be a good guiding principle. Sure there are some outlier examples where you want people to be nasty/aggressive (boxers, MMA fighters, soldiers in combat) but even in these cases you certainly want them to be excellent.

‍

Facebook’s culture, like all effective cultures, is effective saying ‘We’re in a strange situation, therefore we need to behave in counter-intuitive ways. Here’s what we’re all going to do in order to fit ourselves to this odd constellation of circumstances.’

‍

This advice is at its most powerful when it sounds odd and counter-intuitive to most people, precisely because it is a response to weird circumstances that most people and organisations don’t face.

‍

It might seem ironic to write about how 'Move fast and break things' is a good culture for Facebook soon after their infrastructure broke and caused their entire service to go down for several hours but this highlights another interesting point about culture. As the company grows the circumstances to which it is exposed change and therefore the optimal response to these, and hence the optimal culture, also change.

‍

As Facebook has transitioned from a scrappy startup moving at the speed of light, to a domainant incumbent where the downside from breaking things has grown exponentially, so has their culture shifted. Of course they realised this, and in 2014 the famous  'Move fast and break things' was quietly replaced with 'Move fast with stable infrastructure'.

‍

It’s also interesting that one of the main reasons they pressed so hard to acquire Instragram was their respect for the latter’s speed, particularly the rate at which they would deploy new code. While Facebook was still on a weekly deployment schedule Instagram were deploying in real time (reportedly running a command called ‘Yolout’ to do so). The very thing that made Instagram so dangerous to Facebook and an attractive acquisition prospect was that they were executing the early Facebook culture better than Facebook themselves.

‍

One of the most interesting things here is the striking parallel to evolution. The test of whether an adaptation/gene/mutation is good or bad is whether it better enables an organism to survive and reproduce in a particular environment. The test of whether a culture is good or bad is whether it better enables an organisation to thrive in its particular environment. By necessity adaptations and cultures that fit organisms and organisations for one set of circumstances handicap them in others. [3]

‍

But there's one massive difference between the two. While evolution is reactive company culture can be proactive. Genes are selected for by natural selection - the genes that are present in higher concentrations in a group are those which have conferred a survival/reproduction advantage over the previous generations. If the environment changes there's always a lag phase before evolution catches up and the genes that suit organisms to the new environment become common. In other words there's no way for evolution to get ahead of any environmental changes.

‍

Cultures, particularly company cultures, can be proactive and this is one of the most exciting things about them. Far from being just a dumb product of evolution, the best cultures represent a bet that founders and companies have made on how the world is going to change and therefore on what the optimal way to operate will be. That's really exciting. I think it's also one of the best examples of where vision and forecasting matter in startups. That's even more exciting.

‍

‍

‍

‍

‍

Notes

‍

1 - I think this is one of the reasons why the two parts of culture aren’t completely independent. If you force someone to sit through a corporate session on 'culture' that tells them to be kind, take responsibility and work hard you're probably more likely to be annoying and patronising than to actually generate change. If you give them a precise set of instructions and explain why they should act in a strange way you're more likely to at least get their attention and have a better chance of changing behaviour. The more your culture is specifically pointing people in the right direction the better your chance of getting widespread adoption.

‍

2 - Note again that this assumes you're nailing the first part of an effective culture: that people are actually behaving in the way you outline in the culture. One failure mode is, of course, that you fail to properly bed the culture into the organisation and in reality people who break things get fired rather than rewarded.

‍

3- There are practically infinite examples of this but one of the most striking is that the disease we call sickle cell anaemia is protective against malaria. If you’re in an environment where malaria is prevalent then the protection you get from sickle cell anaemia outweighs the harm. When removed from this environment however, this trait is on balance harmful as you get all of the haematological risks with none of the benefits of protection against malaria.

‍

Thanks to Ruairidh Forgan, Lydia Field and Jamie Strachan for comments.