Org diagnosis / strategy-organization fit
Nadler-Tushman Congruence Model
An input-transformation-output model of the organization where performance depends on congruence: how well four core components, the work, the people, the formal organization, and the informal organization, fit each other and the strategy.
Problems are never in one component; they live in the fit between two.
- Problem
- Org diagnosis / strategy-organization fit
- Altitude
- Enterprise
- Effort to run
- Heavy
- Evidence base
- Established
Theory & origin
David Nadler and Michael Tushman published the model in 1980 (A Model for Diagnosing Organizational Behavior), importing open-systems theory into practical diagnosis. The organization takes inputs (environment, resources, history), passes them through a strategy, and transforms them into outputs through four interacting components: the work (the actual tasks), the people (skills, needs, expectations), the formal organization (structure, processes, systems), and the informal organization (culture, norms, networks, politics). The central hypothesis is the congruence hypothesis: the greater the fit between each pair of components, the higher the performance. The model deliberately has no preferred design: a design is only good or bad relative to what it must fit, which is what separates it from best-practice frameworks.
Key components
The parts at a glance. Click any term for the full definition, a field example, and the common failure, in the model below.
Explore the model
How a consultant runs it
- 01 Frame the inputs first: what do the environment, the resource base, and the organizational history fix in place? History matters most and is most often skipped: past bets constrain present options.
- 02 Describe each of the four components as it actually is, not as the operating-model deck says: observe the work, profile the people, read the formal design, and map the informal organization through interviews.
- 03 Score the six pairwise fits (work-people, work-formal, work-informal, people-formal, people-informal, formal-informal), plus each component against the strategy. Evidence per pair, not impressions.
- 04 Locate the performance gap in the two or three weakest fits and design interventions that repair the fit, which sometimes means changing the component you were not asked to touch.
- 05 Re-test congruence after the change: fixing one fit routinely breaks another (a new structure that suits the work but violates the informal networks), so the diagnosis is a loop, not a report.
When to use
- 01 Diagnosing why a sound strategy underperforms: locating which fit is broken
- 02 Pressure-testing a redesign before launch: score the six fits under the future state
- 03 Post-merger work: two informal organizations colliding under one formal design is a congruence problem by definition
When not to use
- 01 When you need a designed answer: the model diagnoses fit, it prescribes no structure (pair with Galbraith Star for the design step)
- 02 Quick single-issue problems where one component is obviously at fault and cheap to fix
- 03 As a culture-change program by itself: it will locate the informal-organization misfit but offers no change mechanics (pair with Kotter or Burke-Litwin)
Worked example
A multifinance lender pivots to digital origination. Strategy is clear, funding is cheap, and a new credit workflow (formal organization) is live, yet approvals are slower than before. A congruence diagnostic scores the six fits and finds two broken seams: work-people at 2/5 (the work now demands statistical judgment the surveyor-heavy workforce does not have) and formal-informal at 1.5/5 (the new workflow criminalized the informal exception network that used to clear edge cases in hours). The formal-organization component itself scores fine, which is why every workflow audit kept coming back clean. The intervention is aimed at the seams: a capability bridge (redeploy-and-train plan for surveyors) and formalizing the exception network into a documented fast lane, rather than another round of process redesign.
Common pitfalls
- 01 Auditing components instead of fits: every box can score well while the seams between them fail
- 02 Skipping history: designs keep dying because they violate lessons the organization is still obeying
- 03 Reading the informal organization as resistance rather than as a load-bearing component
- 04 Stopping at diagnosis: congruence names the broken fit but someone still has to choose the design and run the change
Sample deliverable
One real engagement, start to finish. Watch the numbers travel from raw input, onto the chart, into the finished artifact.
Input
- Formal-informal1.5 / 5
- Work-people2.0 / 5
- Strategy-work3.9 / 5
- Work-formal4.1 / 5
- People-informal4.3 / 5
Process
Evidence-scored fit per component pair, weakest seams flagged for intervention
Congruence scan: digital origination pivot
- Weakest seamformal-informal at 1.5, the new workflow broke the exception network
- Second seamwork-people at 2.0, surveyor skills vs statistical work
- Interventionformalize the fast lane + capability bridge, not more process redesign