As startups grow and professionalise it's common for their spirit to dwindle. Without care and attention they become just another big-corp. It's painful for founders, difficult for employees and frustrating for customers. There's hope though. Some companies manage to maintain (or rekindle) their spirit after scaling to thousands of employees.
There aren't great definitions of startup spirit - though you'll know it if you feel it. Startups usually have some defining characteristics:
- easy access to people making the decisions
- employees can meaningfully affect the company trajectory
- everyone has a deep connection to the customer
- teams show initiative and self-direction
As you scale there are tactics that can act as bandaids to maintain startup spirit e.g. office hours with the founders, keep a tight leash on bureaucracy, share stories about customers and their needs, etc. Unfortunately tactics aren't useful without a strategy to guide their implementation. To put together your strategy let's talk about the research of Metcalfe and Dunbar.
Balancing Metcalfe's Law and Dunbar's Number #
Your startup is growing. As it does you'll notice systems and relationships breaking down. The answer to why is a combination of Metcalfe's law and Dunbar's Number.
What is Metcalfe's law? #
Metcalfe's law asserts that every user added to a (telecommunications) network increases the network's value exponentially. When you hear people talking about "network effects", this is what they mean.
As your startup grows, you'll see this law in action. A team of 20 people can deliver more than 20x the value of a one person startup[1]. At least that's what you'll see in the early days.
So. In theory, adding more people to a company should make everything better (higher quality decisions, faster output, etc.). Sadly experience shows us this is far from true. Giant companies are typically less nimble than their tiny-startup counterparts. Why...? Queue Dunbar's number.
What is Dunbar's number? #
In 1992 Robin Dunbar was investigating the origins of language. He discovered a biological limit to how many people any one person can maintain a relationship with [2]:
"... there is a cognitive limit to the number of individuals with whom any one person can maintain stable relationships, that this limit is a direct function of relative neocortex size, and that this in turn limits group size ..."
To put this in other terms. The ratio of a brain's neocortex to its volume predicts how many social contacts someone can monitor[3]. For an average human, that number is 148 (we'll round up to 150 because everyone else does..).
The broader story is useful - so a quick diversion.
In his 1991 paper, Dunbar showed that the more social contacts you have the more time you need to spend in "social grooming". However, to manage the relationships with 150 people we'd each need to spend about 45% of our days in "social grooming"! Obviously no species can survive this, and hunt, gather, sleep, etc. So humans invented a hack. Language. Yes - he reckons we invented language 300,000 years ago to "groom" more than one human at a time. We needed to groom more people so we could be part of bigger groups. Why? Bigger groups are:
- tougher for predators to take on
- better at out-competing other groups of humans
- essential if we choose to share resources between groups instead of competing with them
Of course the less time you spend "grooming" someone, the less close they'll be[4]. Anyone outside of your 150-strong social circle will struggle to form meaningful social bonds with you.
What's the balancing act? #
Well. As a switched on, observant startup founder you may have noticed we are surrounded by giant organisations e.g. cities, countries, and multinational companies. Those organisations are filled with people and consequently have plenty of network-value (thanks to Metcalfe's law).
But. Surely they should fall apart when they exceed Dunbar's 150 people? Well, humans are smart. We invented more hacks. We told stories[5] about our origins, our fates, our gods. We created hierarchies - where people are put into roles for social control. That way we can extend Dunbar's number. The returns diminish with scale though - hence insurrections, revolutions, etc.
All interesting. What does this mean for my startup though?
In short, plan ahead with Dunbar's number in mind. Yes your 500 person company has more network-value in the abstract. But. If each person in the network can't tease out their colleague's value, Metcalfe's law breaks down. People lose track of what their colleagues work on, and what their strengths are. That means that they can't leverage each other effectively. Worse still, the disconnection can breed discontent.
Of course companies aren't divided into chunks of 150 people. So what ends up happening is that as the company grows, people will be in a few different groups of 150. This doesn't make for good social bonds - so network quality degrades.
As the network quality degrades, so does your startup spirit. As you add more people to the mix, the zing your startup has dilutes until you're left with just another big corporation.
When you walk into a big-corp it's easy to notice when this has happened. People talk about departments in the abstract: "I'll speak to someone in IT", or "Marketing wants xyz by Friday". Sadly, in a company of 5,000 no one will know everyone's name. People will talk about "the company" instead of "we". You need another approach to make sure people are connected to the central startup spirit.
How do you keep your startup spirit as you scale? #
If things go well for your business there won't be enough people to chase all the opportunities, so you'll hire hire hire. Over time scaling problems will manifest as:
- People no longer knowing each other's names
- People talking about departments instead of peers
- People raising questions in all-hands meetings that show deviations from the cultural outline (most likely the values).
To solve the scaling problems before they appear you need a people strategy. Your people strategy must focus on bringing colleagues together regularly while reinforcing the foundations of your culture.
However, no matter how much you reinforce the corporate culture, you must accept that subcultures will form. Subcultures will be around geography, language, skillset, sports, etc. The key is to prevent subcultures from becoming silo's. To do this you need to establish, and maintain a strong cultural outline, and hire carefully.
In addition you must apply tools to the problem to reinforce the bonds between people. Tools like a regular all-hands meeting, slack, an internal blog, etc. With direct lines of communication, and a context to communicate in, it becomes easier for good ideas to surface. The tools also make it easier for guidance and messaging to permeate the company.
It goes without saying that Chinwag can help too. With Chinwag people from across the company practice working together in a fun way that produces meaningful bonds.
Footnotes
This presupposes that those people have their efforts effectively channelled and all have a strong understanding of the company's cultural outline. ↩︎
Aiello, L. C., & Dunbar, R. I. (1993). Neocortex size, group size, and the evolution of language. Current anthropology, 34(2), 184-193. ↩︎
Subsequent investigations into Christmas card lists, offices, and military organisations show that when 150 is exceeded the social group degrades. ↩︎
Dunbar himself made some educated guesses about where the inflection points exist for the relationships. These were set at 5, 15, 50 and 150 - from closest friends to acquaintances. Subsequent research has shown that these numbers and inflection points can vary depending on all sorts of circumstances. That being said - the overall argument that we have a finite capacity for human relationships holds. ↩︎
Harari, Yuval N. author. Sapiens: A Brief History of Humankind. New York :Harper, 2015. ↩︎