Startups Scaling Fast in 2026 Face a Hidden Culture Problem

New research links rapid growth to declining trust between managers and teams

Fast-growing startups in 2026 are running into a problem that rarely shows up in a board deck: trust between managers and teams breaks down exactly when headcount takes off. New workplace research from the O.C. Tanner Institute, Emtrain, and an April 2026 interview with an organizational psychologist all point to the same pattern. As companies scale from a dozen employees to hundreds, psychological safety, the sense that people can speak up without punishment, does not scale with them.

Alexander Kopylkov breaks it down into first principles: small teams build trust through daily repetition, growth adds structural complexity, and structural complexity opens a trust gap.

A twelve-person team builds trust through constant, face-to-face contact. Add layers of managers, time zones, and reporting lines, and that repetition disappears.

The result is a workforce that still nods in meetings but stops raising the problems that matter. Kopylkov compares this to municipal water pressure: a system built for ten houses does not fail because the pipes broke, it fails because nobody redesigned the pressure system once five hundred houses were added.

De Jong Schouwenburg, who studies this at scale, put it directly in her April 2026 conversation with InfoQ: "You communicate well with 12 people. You know each other. It doesn't happen with 500." She distinguishes trust, which is personal and earned between two people, from psychological safety, which is collective and enabled by the group's norms.

As she put it, "As teams scale, you can't copy-paste trust from one group to another. Each new team has to rebuild it; it has to grow." A ten-person team with high trust does not pass that trust down to the next fifty hires automatically. The work starts over.

Citing data from the O.C. Tanner Institute, Kopylkov notes that 72% of workplace culture initiatives, per Harvard Business Review research the Institute cites, produced no measurable improvement, and 57% of employees actually felt worse after their company rolled out a culture-building perk, treating it as a band-aid rather than a fix.

The one intervention that did move the needle was cheap and unglamorous: when senior leaders changed their own daily behavior, without a formal program, trust scores rose 26%, and employees became 18 times more likely to describe the workplace as having both high expectations and high support.

"Trust is between people; psychological safety is among people. Trust is earned; psychological safety is enabled."

The organizational psychologist said.

According to Kopylkov, this is why so many founders misdiagnose the symptom. A slowdown in decision-making, a rise in silent disagreement, or sudden turnover on a team that seemed fine six months ago usually gets treated as a hiring problem or a compensation problem.

A 2025 study in the Journal of Business and Psychology adds a useful correction: psychological safety is a perishable resource that rises and falls based on active team behaviors, connecting, clarifying, supporting, and performing, not on how long people have worked together. It does not build up automatically with tenure, and it does not survive scaling on its own either.

For founders, Kopylkov recommends treating psychological safety like a metric, not a mood. Track how long decisions take at each layer of the org, how often people raise disagreement in retrospectives or postmortems, and how much drift there is between how one team operates and how another does. Those are observable signals, unlike culture in the abstract, and they catch the erosion before it shows up in attrition data.

From an investor's perspective, Kopylkov evaluates a scaling team by watching how decision speed and open disagreement behave as headcount grows, not by reading the org chart itself. A flat chart with slow, guarded decisions is a worse sign than a layered one where people still argue openly in front of leadership.

Not every friction point during a growth phase is a psychological-safety failure. Emtrain's 2026 research found a genuine bright spot alongside the erosion: workplace accountability, meaning clear expectations and consistent follow-through, is actually improving at many companies. That is a reminder that some of what looks like a trust problem is really a structure problem, and the fix is clearer decision rights, not another all-hands meeting about values.

In his view, the startups that handle scaling well by 2027 will be the ones that design for trust on purpose, through habits like cross-team retrospectives, rotating meeting facilitators, and shared rewards for transparency, rather than assuming culture will hold just because it always has.