---
title: Spring's Broken Control Loops (VSM Analysis)
author: IBEAR Toolkit
cohort: Classes 40-49
language: en
date: 2026-08-20
paths: [stayer, returner, pivoteer, climber]
---

# Spring's Broken Control Loops: A VSM Lens

**Current thinking:** Spring breaks the program's ability to sense and adapt. Here's why, and what that means.

## What Control Loops Are

In systems thinking (VSM — Viable System Model), a control loop needs three things:
1. **Input** (the program does something)
2. **Feedback** (the cohort responds, program sees it)
3. **Adaptation** (program adjusts based on feedback)

Fall has all three: program publishes calendar → cohort responds (visible in chat, attendance) → program adjusts (changes sprint focus, reschedules events).

Spring? The loop breaks.

## Why Spring's Feedback Vanishes

**Recruiting timelines fracture:**
- Stayers: recruiting is live (month 4-6 of search)
- Returners: recruiting done (home-country hiring cycles closed in Feb)
- Pivoters: no recruiting (they're building proof)
- Climbers: recruiting locked (already have offers)

One calendar assumes everyone has the same stakes. They don't. Program can't see which person is in which timeline.

**Chat drops.** (This is VSM's key signal loss.) Cohort goes quiet not because they're disengaged but because:
- Stayers anxious about recruiting (working privately)
- Returners prepping departure (low social energy)
- Pivoters deep in portfolio work (invisible)
- Climbers negotiating offers (confidential)

Program observes silence and assumes stasis. Wrong. System is *highly active but invisible*.

## What This Breaks

**Generic support fails.** Mentorship that works in Winter (recruiting advice for stayers) becomes noise for returners (already have jobs), pivoters (need technical feedback, not career advice), climbers (need offer negotiation, not encouragement).

**Cohort metrics collapse.** How many people are "on track"? The program doesn't know because "on track" means different things for four people.

**Adaptation cascades fail.** Program can't respond to what it can't see. So program continues one cadence (recruiting focus) even though half the cohort isn't recruiting.

## What We're Testing

What if Spring explicitly *expects* fragmentation instead of fighting it?

- **Path-specific signals:** Different check-ins for each persona (recruiting pipeline for stayers, home-opportunity monitoring for returners, portfolio reviews for pivoters, offer advisors for climbers).
- **Parallel standups:** Standup by path, not one cohort standup. Feedback flows through four channels instead of one.
- **Visible dispersal:** Post "who's recruiting, who's departing, who's in portfolio mode" weekly. Make the four timelines visible.
- **Closure as output:** Departures aren't losses; they're transitions the program actively enables. Ceremony around exit.

This trades "one cohort in sync" for "four paths with real support." Feedback loops reactivate because program now has signal from each path.

## Risks We See

- **Isolation:** If paths are too separate, cohort fragmenting (already happening, but program shouldn't accelerate it).
- **Unfairness:** If some paths get more support than others, resentment forms.
- **Coordination loss:** Some collaboration requires cross-path energy. All-parallel might kill it.

How to navigate? Cross-path projects (returners mentor pivoters, climbers give talks), but voluntary, not mandatory.

## Open Questions

1. Does "path-specific support" actually reduce isolation, or does it enable it?
2. How do we maintain cohort identity if people are on completely different timelines?
3. Should Part-Time MBA students (3-year track) have a different Spring altogether (they never fully disperse)?

---

**Status:** Current thinking, open for testing.
**Related:** season-landings, VSM-lens, collaboration-by-season