As organizations grow across geographies, one thing becomes disproportionately important. Culture. We, engineers, often dismiss culture as soft and cushy. This is until you see the hard costs of ignoring it. Culture isn’t just how we feel about work; it is the distributed operating system for how decisions are made when leadership isn't in the room. There is company culture, the one leadership defines, articulates, and is supposed to live by. And then there is engineering culture, the rigor of how systems are built, reviewed, and operated. They overlap, but they are not identical.
This piece focuses on the former. I would like to talk about what actually happens when you set up teams in new locations, introduce new managers, and hire people who never experienced the early days. How do you preserve what worked? More importantly, how do you avoid freezing it in time and instead improve it as you scale?
As teams expand across time zones and contexts, shared interpretation weakens. What was once obvious becomes negotiable. What was once assumed becomes redefined by whoever speaks first. If that redefinition is not deliberate, it will still happen. Just not in your favor. Scale introduces entropy. The question isn't whether your culture will mutate as you grow. It absolutely will. The only choice you have is whether it mutates by design or by drift.
Top:
“SO YOU’RE TELLING ME…”
Bottom:
“…WE HAVE A SLACK CHANNEL…
AND THAT COUNTS AS CULTURE?”
Culture Scales By Systems
When teams are small, culture travels through proximity. People observe how decisions are made. They see how trade-offs are handled. They learn what good means by watching others work. No one needs to write it down because it is visible every day.
As organisations grow, that transmission mechanism weakens. Distance increases. Context fragments. What was once absorbed through shared experience now has to be made explicit. At scale, culture no longer spreads through exposure. It spreads through systems.
Why Culture Degrades as Organisations Grow
As organisations expand across geographies, three structural shifts occur: more people, more distance, and less shared context. Shared context is what keeps standards aligned. When it declines, interpretation diverges. What one team considers acceptable, another considers risky. What used to be obvious becomes negotiable.
Ownership also changes. In early stages, engineers operate fluidly across boundaries. As headcount increases, roles become rigid and specialization intensifies. So, you need to clearly define roles and responsibilities. Otherwise, people will define those on their own. Or worse, nobody will take any responsibility.
As the growth happens, you want stability. But additional structure does not automatically create alignment. It can just as easily create overhead without shared understanding.
Culture Follows Optimisation
Culture is the emergent result of what people optimise for under pressure. If individual KPIs dominate, behaviour becomes competitive. If metrics become targets, delivery turns into theatre.
Systems define what winning looks like, which trade-offs are tolerated, and whether raising problems early is safe or risky. As shared context declines, those systems become the primary carrier of culture.
Goal Design is The Primary Culture Lever
If culture follows optimisation, then goal design is the primary culture lever. Because goals define what winning means. And once winning is defined, behaviour adjusts around it. If success is ambiguous, people optimize locally. If success is shared, optimization aligns.
One Team KPI
Each team should have one outcome that’s part of the overall vision. Improve X by Y% within six months. Not five competing priorities. One outcome that defines success. When multiple metrics compete at the same level, trade-offs become implicit. People choose what to optimize for. Alignment becomes interpretive. That’s why the concept of the north star exists. It makes it easy to apply strategy and planning.
One shared KPI forces trade-offs into the open. It forces prioritisation. It makes delivery plans transparent. It results in disagreement productive instead of political. Culture weakens fastest when success can be interpreted differently by different people.
Individual goals support the team outcome
If individuals are measured independently, optimization shifts. Behaviour becomes positional. Risk becomes something to protect against. Information sharing becomes selective. In engineering environments especially, individual metrics tend to be gamed. People can perform visible activity over real impact. The fastest way to kill a culture is to design a game where an individual can win while the boat is sinking. If I can hit my personal goals while the team fails, we are no longer teammates; we are competitors sharing a budget.
Individual goals should exist, but they serve a different function. They describe contributions. They clarify ownership. They evolve as work evolves. They are directional, not comparative. Individual goals exist to strengthen the team outcome, not to create parallel definitions of success.
Raising What Good Looks Like
A KPI defines direction. It does not define quality. This is where leadership often underestimates the problem. As teams scale, complexity and cognitive load increases. Dependencies multiply. Coordination overhead rises. What counted as strong execution at 15 engineers becomes baseline at 80.
As teams scale, good enough must be redefined deliberately, otherwise it will be redefined accidentally. This is about acknowledging that scale changes the threshold of acceptable execution. Yesterday’s success becomes today’s baseline.
Bad Goal Design to Bad Culture
Bad goal design does not fail immediately. It shapes behaviour gradually. Once metrics become targets rather than signals, behaviour bends around them.
The fastest way to damage culture is to let people win while the team loses. When that pattern stabilises, trust erodes. Collaboration becomes selective. Optimisation becomes defensive.
Bonding Actually Scales Culture
If goal design aligns optimisation, bonding aligns interpretation. Systems can define a KPI. They cannot ensure people interpret it the same way.
At a smaller scale, interpretation synchronises through proximity. People observe how disagreement happens. They see whether risk is surfaced early or hidden. They learn what good execution means by watching it happen.
At scale, you cannot rely on telepathy. You need calibration. Bonding here is not about social cohesion or happy hours; it is the high-bandwidth synchronization of judgment.
Bonding is Interpretive
Bonding is where shared mental models form. Teams with aligned interpretation resolve conflict earlier. They escalate faster. They debate trade-offs instead of personalities. Documentation cannot solve this.
A document can define a metric. It cannot define how aggressively to pursue it. It cannot define whether pushing back is safe. It cannot define whether deadlines are flexible under risk. Those rules are inferred. Bonding is where people observe:
- how leadership reacts to bad news
- whether disagreement is punished or welcomed
- whether trade-offs are explicit or hidden
- what quality actually means in practice
Without interpretive bonding, systems exist on paper while behaviour diverges underneath.
Bonding for New Members
New hires arrive without shared interpretation. If bonding is weak, they optimize for self-protection. Silence becomes the rational strategy. And silence is the earliest signal of cultural dilution. Early involvement in goal discussions, planning, and retrospectives accelerates interpretive alignment. It exposes how the team thinks.
Without this exposure, new members form local interpretations. In distributed organizations, those interpretations fragment quickly across locations.
Bonding for Existing Members
Bonding is not only onboarding. As organisations scale, subcultures form. Assumptions diverge, Context Fragments. Even when the KPI remains shared, interpretation drifts. One location may optimize aggressively for speed. Another for risk avoidance. Both believe they are aligned. Neither is explicitly misaligned.
Bonding re-synchronises interpretation. Its surfaces drift before it becomes visible in outcomes. It refreshes shared standards and principles. It prevents local optimisation from hardening into silo behaviour.
Bonding is how culture is re-synchronised at scale. Systems align direction. Bonding aligns meaning. Together, they make culture flourish despite the scale.
Execution is Proof Of Culture Holding
Execution is where culture becomes measurable. If goals are aligned and interpretation is synchronised, execution goes silent. If they are not, execution becomes volatile.
Coordination cost increases when shared mental models weaken. Teams begin clarifying assumptions repeatedly. Decisions are revisited. Rework increases. These are not productivity issues. They are interpretation issues.
Healthy Execution Has Clear Patterns
Healthy teams disagree early. They surface risk before it becomes expensive. They debate trade-offs when there is still time to adjust. Once alignment is reached, execution becomes quiet. They disagree and commit to the outcome.
When you have loud tension early and then later calm, the work is happening in the way it should happen. Because when psychological safety is present, escalation happens sooner. When it is absent, risk is delayed. Delayed risk converts into deadline pressure.
Execution Problems Signal
Most execution breakdowns are treated as effort problems. They are rarely effort problems.
- Missed deadlines often indicate unclear goal hierarchy.
- Repeated rework often indicates unresolved trade-offs.
- Last-minute escalations indicate suppressed risk.
- High meeting volume often indicates fragmented interpretation.
When evaluation pressure is attached to narrow metrics, behavior bends around it. Execution failure is usually upstream optimization failure. If you only address delivery symptoms, you reinforce the same incentive structure.
Reviews As Culture Reinforcement
Review systems shape behaviour. If reviews reward visibility, people optimize for visibility.
If they reward narrative defence, people prepare narratives. If they reward early risk sharing, people escalate earlier.
Reviews are calibration events. They surface trade-offs, clarify expectations, and reinforce what is rewarded. A review is just around calibrating personal outcomes against clearly defined expectations.
A Diagnostic Lens
Cultural drift begins with hesitation. Fewer hard questions in meetings. Fewer challenges to assumptions. Fewer early escalations. We need friction in a healthy environment to get to a better place. When you feel there’s no friction, it either means no motivation or fear of being judged.
Patrick Lencioni describes this very well in The Five Dysfunctions of a Team. Trust erodes first, then conflict is avoided, then commitment weakens, then accountability softens, and only at the end do results visibly drift. By the time results come, the damage is already done.
If growth makes conversations softer but delivery heavier, you know that the drift has started. If growth makes debate sharper but execution calmer, the system is still aligned. The dysfunctions are not causes. They are lagging indicators.
In Consequence
Culture does not scale through intention. It scales through repeated behaviour shaped by systems. As organisations grow, shared context weakens and coordination cost rises. If goals fragment, optimisation fragments. If interpretive alignment weakens, friction moves downstream and shows up as rework, defensiveness, and slower decisions.
Stop looking at your values statement and start looking at your calendar. Culture is how work actually moves through the pipe. It is the visible manifestation of who gets heard, who gets blocked, and what gets shipped. When growth sharpens judgment instead of diluting it, the system is holding. When growth introduces hesitation and defensive optimisation, something structural needs redesign.