How to Unite the Right People

“Upstream interventions require cooperation across systems that were not designed to cooperate.” — Dan Heath

The Cross-Boundary Problem

Almost every significant upstream problem crosses organizational boundaries. Chronic student absenteeism requires cooperation between schools, social workers, housing agencies, and healthcare providers. Teen drug abuse requires coordination between families, schools, law enforcement, community organizations, and the healthcare system. Traffic fatalities require collaboration between road designers, car manufacturers, police departments, hospitals, and urban planners.

These organizations don’t normally work together. They have different missions, different cultures, different budgets, different metrics, and often compete for funding. They may not even recognize that they are working on the same underlying problem from different angles.

The first challenge of upstream work is not technical — it is social. It requires building a coalition of organizations that have never cooperated before, around a problem that no single organization can solve alone.

The Chicago Schools Story

Heath opens the chapter with the story of Chicago Public Schools and chronic absenteeism — defined as missing 10% or more of school days. The number is devastating: in some Chicago neighborhoods, over 30% of students were chronically absent. Academic outcomes for these students were dire.

The school district’s initial response was downstream and siloed: teachers noted absences, attendance officers made calls, parents received letters. None of it worked.

The breakthrough came when a working group brought together all the organizations that touched chronically absent students: schools, but also child protective services, healthcare providers, the housing authority, and social service agencies. When they looked at their combined data — something that had never been done before — they discovered that many of the chronically absent students shared specific circumstances: recent housing moves, untreated health problems, siblings in foster care, parents with mental health challenges. The problem wasn’t school disengagement. It was life instability.

Building Unusual Coalitions

Solving upstream problems requires bringing together people who don’t normally work together — and who may not immediately see why they should. Heath offers a framework for building these “unusual coalitions.”

Who Needs to Be at the Table

The right coalition for an upstream problem typically includes:

1. Those who touch the problem — organizations that encounter the people affected, even if the problem isn’t their primary focus
2. Those who own relevant data — agencies that collect information that could reveal patterns invisible to any single organization
3. Those who control relevant resources — organizations with funding, staffing, or infrastructure that could be deployed differently
4. Those with legitimacy — community members, community organizations, or government entities whose involvement confers credibility and builds trust
5. Those with authority to change systems — policymakers, administrators, or funders who can authorize systemic changes rather than just individual interventions

Note what’s often missing from upstream coalitions: the people most directly affected by the problem. Effective upstream work brings in community members not just as data sources or beneficiaries, but as genuine partners in diagnosis and solution.

System Leadership vs. Point Leadership

Traditional leadership means solving problems within your own organization, with resources you control, according to your own strategy. This is point leadership — effective within one system, helpless across multiple systems.

Upstream problems require what Heath calls system leadership: the ability to coordinate action across organizations you don’t control, using influence rather than authority, toward goals that none of the partners could achieve alone.

The Skills of System Leadership

System leaders need different skills than traditional managers:

• Translation: explaining each organization’s concerns and constraints in terms the other partners can understand
• Trust-building: demonstrating through small commitments that cooperation will be honored and that data shared won’t be weaponized
• Patience: accepting that cross-system collaboration moves slowly, with setbacks, and requires sustained relationship-building over months and years
• Shared metric-creation: working with diverse partners to agree on how success will be measured — a genuinely difficult task when each organization has its own reporting requirements

Reflection

Think of a problem in your community or organization that spans multiple systems or stakeholders. Who would need to be at the table to solve it upstream? What would it take to convene that coalition — and who would be willing to lead it?

Key Takeaways

• Upstream problems almost always cross organizational boundaries — this is structural, not accidental, and must be addressed directly
• Chronic absenteeism in Chicago was solved only when schools, healthcare, housing, and social services looked at their combined data for the first time
• The right coalition includes those who touch the problem, own relevant data, control resources, have community legitimacy, and hold authority to change systems
• People most affected by the problem should be genuine partners in the coalition, not just beneficiaries of it
• System leadership — coordinating across organizations you don’t control — requires different skills than traditional management: translation, trust-building, and patience
• Cross-system collaboration moves slowly and requires sustained investment in relationships before any solutions can be implemented