Silos and Civilizations
Problems in Dreamland zeroes in on the operational reasons once promising startups melt down - and how to avoid it. This is perfect for founders, leadership teams, or anyone in a highly cross functional IC role (Chiefs of Staff, product managers, biz ops, etc).
The first meltdown trigger: silos and civilizations.
You planted the seeds of your startup's demise when you made your first N+1 specialist hire. By which I mean your second sales person, or second engineer, or second customer success manager. This feels counter intuitive, because building an [X] team is exciting! It's a sign of progress, of things working, of gaining traction - especially for a founder who is doing all the things as a generalist up until this point.
But alas, we humans are frail, and we need to belong. We need a group and an identity. We need an Us (and where this is an Us, Them closely follows). By adding a second specialist, you have inadvertently given them an in group, an Us identity in the workplace, that goes beyond just belonging to your startup - they now belong to [X] team at your startup. You have started dividing your startup into sub-groups, separate civilizations, whether you realize it or not. We call these "silos."
Now add in that startups are stressful environments with constant tension over scare resources. Our ancestors fought for food, water, land. In startups we contest for scarce dollars from the seed round, for the founder's time and attention, and for less flattering reasons like promotions (and our egos). Amid this stress, we look for comfort. Say a pitch for an initiative was shot down in favor of focusing on something else, advocated by [Y] team. And say that made your life harder. Maybe you have to continue to work longer hours because of that lack of funding. You don't feel great, so you turn to your N+1 and say something innocuous, like "I can't believe [Y] team got their budget approved. We ([X] team) are working our asses off, while they get all these extra perks and support."
Soon the friction becomes obvious to the founder, so they solve it by hiring a manger or leader for that team. But early on, that's often a first time manager or leader. And that leader really wants their team to like them. And what's the quickest, cheapest way to create affinity and emotional bonds? Tell unflattering stories and express frustration about [Y] team (Them) to make [X] team (Us) feel closer and more emotionally secure. The leader feels good because their team likes them - maybe even loves them - but no one is tallying the expense of this approach. Silos are being built, and silos with thick walls.
This is the startup variant of Tupac's Changes lyrics: "You gotta operate the easy way / I bonded the team today / but you did it in a sleazy way."
Silos are not quite inevitable - more on how to mitigate them below - but the natural course is for silos to form, and earlier and more quickly than founders suspect (and with more staying power than anticipated). How dangerous and damaging these silos can be to your startup's prospects depends almost entirely on how they interact.
There are typically three types of silo or micro-civilization ecosystems in a startup:
Peaceful agrarian
Warring militarized
Colonial imperialism
(is my poli-sci undergrad showing? Sorry, not sorry)
Let's examine Peaceful Agrarian ecosystems first. This a sustainable silo-ed system, when times are good. Each function is peacefully doing their own farming thing. They harbor no animosity to other functions, but also exhibit zero curiosity. This works reasonably well when business is humming along, no major changes are needed, everyone is just executing in their silo and it feels good. But, the good times always end, or the pressure amps up for more performance. So sales starts tweaking the positioning of the product to close more deals. And it works! Sales is killing it! But no one told the product team or CS team, so they're doing the same thing they were doing before. Retention starts dropping, because the promise made to the customer starts diverging from reality. Product and CS now feel pressure to "fix" retention. So they start making changes (in their own silos). The divergence gets worse. Now the silos are interested and curious about one another ... but with hostile intentions, seeking to blame others for their woes.
That's how Warring Militarized ecosystems start. These are ugly. No one is hitting their goals (if goals were even well architected in the first place). No one can fix it alone. But because there were no healthy behaviors learned on how to collaborate across teams, it becomes a doom loop. Blame, sabotage, politics of the worst sort. Often, the founder / CEO will attempt to fix it via central planning -- everything coordinated, highly top down. But a single person lacks the expertise and granularity to orchestrate it all. And it's highly demotivating to the teams themselves.
It also can introduce the temptation to become a "[X] function led company" eg a Sales Led company or a Product Led company or an Engineering Led company. This is Colonial Imperialism, where one function dictates how all the other work. This can be highly successful. It's way better than a Warring Militarized system. But two things have to be true for it to hold. First, the function doing the leading has to be absolutely world class at what they do. They cannot have been granted that power because of politics, or because the founder understands that function best, but because of some level of real market beating durable performance. It needs to be a differentiated core competency for the company. Second, everyone in the company needs to know what they signed up for. If you're in a Sales Led company, and you think you're going to have autonomy as a product person and be able to do unshackled customer discovery - think again. You'll be executing projects based on what Sales deems necessary to win.
How do you avoid the natural (d)evolution of scaling startups into disastrous silos?
Emphasize a "one team" mentality at very top of the org, with the executive leadership team. I've tried a number of different methods, but still believe Patrick Lencioni's classic The Five Dysfunctions of The Team is the best.
Build systematic ways to integrate teams around a single unified mission, using clear goals and thoughtful, high repetition communication. These systems dive deep, and are fairly intricate. Both One Mission: How Leaders Build a Team of Teams and Traction work well.
Orient externally around the customer instead of internally around the org chart. Customer journey maps, as well as service blueprints illustrating how various functions support that journey, are invaluable alignment tools, and typically create strong cross functional curiosity.
A few footnotes:
Blitzscaling includes a useful taxonomy for different company sizes, by employee count, and the key changes that happen at each transition point.
Over time, AI may radically alter some of these silo dynamics. When fewer people can do more and create more value, or where less deep specialization is required because generalists can be augmented by AI, these tricky N+1 team building and bonding dynamics may become less complex, or perhaps less of a threat to company success. We'll see.
