A few years ago, I was part of a collaboration doing research on communities of soil microbes. In the research, I was taking sequence data from soil samples, building phylogenetic trees, and trying to find interesting hints of the universe within the exotic soils from mountain tops, glacial fields, and Antarctic deserts. I remember one day meeting with one of my committee members to try and learn some things about an interesting phylogeny I was looking at. I remember asking, So, what do Verrucomicrobia normally do in the soil? My phylogeny had a huge branch sticking out the side of it, full of this microbe I’d never heard of before. He answered, It’s hard to tell, they are diverse and we haven’t had much luck growing them in the lab to study. To which I exclaimed, But the soil is all Verrucomicrobia!
About a year ago, I saw a paper published about an almost accidental discovery (some of the best are) that the entire setup process before growing cultures in the lab created an environment in the petri dishes that inhibited the growth of some microbes – Verrucomicrobia among them. Growing cultures in the lab petri dish is considered a necessary step to describing microbial species and studying their role in the environment. So when I saw the paper I chuckled thinking of all the grad students that would be racing to better understand all those Verrucs for the coming year.
The processes inside startups can be a lot like those lab petri dishes – sometimes you set the perfect conditions and grow the exact thing you wanted. While other times it takes a lot more perseverance, experimentation, or even a mistake to make it work the way you want. Until you have success in your petri dish, many aspects of the startup environment can remain mysterious and unexplained. And just like science, the times you get it right (and wrong) are worth sharing.
The startup experiment
The experiments you run in a startup petri dish are diverse and often have pretty big impacts. You find yourself running experiments to answer complex questions like, What type of measurable characteristics separate normal job candidates from future leaders? Or, How should teams be managed when they are small versus when they are large? And huge questions like, What is our market? How do we remain innovative? How do we attract the best hires?
In some of the more interesting experiments, you realize that the startup environment isn’t static. Your startup grows and changes the environment around it. When this happens, you need to get good at scrapping bad ideas, recognizing ideas that work, and changing your experiment regularly. Recognizing inflection points becomes a core competency.
One set of environmental conditions within the startup that fascinates me deeply, are the ones the control how communication happens. Now, I don’t mean blogging and Tweeting, nor do I necessarily mean Slack and Gmail, although I do mean them all a little. I mean the type of communication that enables idea creation to become idea execution efficiently. To understand how it changed over time, it helps to start at the beginning.
Communication in the early startup environment
Early stage startups aren’t very much like full-fledged companies. In fact, they are so different that you really have to experience it to understand what makes them work. In our company, we had a level of informality and fluidity that you can’t find in many corporate adventures. Some days you’re an engineer, some days you’re salesman. Together, we were just a collaboration of smart and hard working individuals wanting to get something done and realizing that it could be done bigger, faster, and prettier if we all worked together.
The informality of these early days also created an informal flow of information. You didn’t need meetings to learn from your team. Another team member’s ideas seemed to enter your brain through a process osmosis. Any person on the team could have an idea in the morning and by lunchtime everyone else in the company would be familiar enough with the idea to understand how they could contribute or provide feedback.
That informality effects the way that good and bad ideas are evaluated. In those early days, if you thought an idea was good enough, you just put your head down and worked hard until you could say, “Look, I can prove how good this is!”.
This process of company ideation is exciting but probably pretty dangerous. It is only possible with the right environment. We fortunately had a team of smart individuals that cared for each other and respected one-another’s ideas enough to build space and support through those times of ideation. I’m not sure what happens when startups don’t have the right environment at this stage, but my guess is that to succeed, they work a lot harder and move a bit slower early on.
Communicating as a company
As a company grows the edges on its internal social network necessarily grow exponentially. There is an important inflection point hidden in that growth and related to the idea flow within the company. Ideas will begin to move a little bit slower or worse, get lost completely. Osmosis no longer works.
This is a risky moment. The obvious risk is that some good ideas may become lost in your organization before they reach the right individuals that can execute on them. The less obvious but potentially more dangerous risk is that some members of your team may miss information and not be able to contribute effectively. We definitely experienced this frustrating moment in our growth.
Side note: how singular organisms come together to form collective individuals with shared motivations and livelihood has been a fascination of mine since reading Leo Buss’s work many years ago. I borrowed the images in this blog post from those essays.
My feeling is that a lot of people probably get to this stage and throw Slack (or another communication platform) at the problem. While I think Slack can be great, I don’t think Slack alone solves the problem of moving ideas efficiently at this stage in a startup.
Team members need to know who to talk to about each area of the company. Who’s working on what and what’s their role. Team members need to know where we are all going, what is our short and long term goals, and what are the key milestones to know we are getting there. And finally, team members need to know and celebrate when we reach those milestones.
With the right new environment in place, team members are more capable of self-evaluating ideas and then pushing their best ideas directly to the other team-members that can help make them reality. Knowing full well that if they succeed they will all celebrate. At this point, Slack can then be very helpful to reduce the distance across the startup petri dish and to allow ideas to move quickly.
We are still working at building that environment. Processes such as company wide meetings, team level virtual stand-ups, and various other tactics are all being tested. I’ve started a couple of mini-experiments myself that I’ll write about here in the future.
Come this far? Congratulations, you have something like a full-fledged startup in your petri dish. So what’s next?
Building an idea machine
For a while I thought that the transition from osmosis style ideation into a new environment that facilitated better parallel ideation was the final big move within an organization. Then I dug deeper.
What you find when you dig deeper is that there is something more profound out there related to startups and the flow of ideas. I’ve read a lot lately on the concepts, many of which are caught up in Holacracy and new (non)management practices but not inherently dependent on them. To understand what it is, we probably need to go back to the beginning.
In those early days of a start-up, you are a cult of individuals. To prove an idea was good, you just did it. That cult has long lasting effects, both good and bad. The good is that many of those individuals are probably some of the most talented people you will ever collaborate with. The bad, is that people not from the cult of individuals will find it harder to build ideas into company strategies.
The mathematics of an organization (a startup in our case) hint to me that you will reach a point where the entire organization will always be smarter than any single individual… every time. What this means in practice is complex, but it means reducing the dominance of the cult of personalities.
I’m not sure I know how to setup the right petri dish for such an experiment, but I’m watching the startup world closely to see who does.
If you liked this post help me out by sharing it on Twitter
Thanks to Aaron Steele for reading my drafts.