# Adaptive capacity | Envisioning

Source: https://www.envisioning.com/capability

---

---
[The program](/capability)[Countries](/capability/national)[Organizations](/capability/organizations)[Forecasts](/hindsight)
---

# Adaptive capacity

Why are some institutions able to perceive change, make sense of it, coordinate around it, experiment and adapt, while others are not?

Can that capacity be measured and strengthened? This is the research program behind Envisioning’s work.

The problem

AI and other discontinuities raise the speed and ambiguity of change. Institutions are slower systems. Most organizations now have more intelligence, data and analysis than they can use. The constraint is their ability to absorb it and act.

In practice: recognizing which changes matter, separating signal from noise, translating an external development into implications, deciding before the evidence is complete, coordinating across departments, experimenting before certainty, learning from the experiment, moving resources, and making a successful adaptation normal practice. Institutions fail at different points on that list. Where they fail determines the intervention.

Seven dimensions

Adaptive capacity has seven dimensions. Each is a question about what an institution did in the past twelve months. Questions about attitude are excluded because the desirable answer is obvious.

1.  01
    
    ### Sensing
    
    Can the organization notice relevant change early?
    
2.  02
    
    ### Sensemaking
    
    Can it understand what that change means?
    
3.  03
    
    ### Decision
    
    Can it make consequential decisions under uncertainty?
    
4.  04
    
    ### Coordination
    
    Can different actors align around a response?
    
5.  05
    
    ### Experimentation
    
    Can it test responses cheaply and quickly?
    
6.  06
    
    ### Learning
    
    Does evidence change beliefs and behavior?
    
7.  07
    
    ### Scaling
    
    Can successful responses become normal institutional practice?
    

These are working categories. Benchmark data will confirm or revise them.

How the three parts connect

**Understand the environment.** [Signals](/signals) and the [open research](/research) map what is changing outside the institution: thousands of technologies, checked against sources.

**Understand the institution.** Two benchmarks measure adaptive capacity: the [National Capability Benchmark](/capability/national) for countries, 53 of them published at [ncb.envisioning.com](https://ncb.envisioning.com), and the [Organizational Capability Benchmark](/capability/organizations) for companies, agencies, universities and NGOs.

**Check anticipation against outcomes.** [Hindsight](/hindsight) takes the forecasts that institutions publish, from the IMF to Gartner to our own early technology posters, and grades each one against what happened. It tests the first dimension with the record, not with a survey.

**Strengthen adaptive capacity.** Each service works on named dimensions:

-   [Signals Session](/services/research-sessions): sensing and sensemaking.
-   [Build Sessions](/services/build-sessions): experimentation and learning.
-   [Bespoke Projects](/services/bespoke-projects): coordination and decision.

Data from Signals, the benchmarks and the interventions feeds back into the model. A Build Session also tests whether rapid experimentation moves one dimension. A benchmark also tests whether the seven dimensions are the right ones.

What would disprove it

The model makes three claims that data can contradict.

-   The seven dimensions are distinct. If they load on a single factor that tracks organization size or budget, the benchmark is measuring resources.
-   The profile predicts adaptation. An organization with a weak dimension should show worse outcomes on the responses that depend on it, measured over time. If it does not, the dimension does not describe a capability.
-   An intervention moves the dimension it targets. A Build Session should raise experimentation scores on the next assessment. If it does not, the service improves skills without changing capacity, and it will be described as training.

The national benchmark runs the first test on every release: each dimension score is compared with the score its income level predicts, and correlation and redundancy between dimensions are reported. The organizational benchmark will run the same tests once the sample supports them.

[Measure an organization](/capability/organizations)[Measure a country](/capability/national)