> For the complete documentation index, see [llms.txt](https://urd.gitbook.io/compass/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://urd.gitbook.io/compass/learning-and-adapting/practising-adaptive-management/adaptive-management-in-practice-implementation-and-monitoring-phases.md).
# Adaptive Management in Practice — Implementation and Monitoring phases
The design phase builds the conditions for adaptation. The implementation phase is where adaptation actually happens or not. The most common failure at this stage is not a lack of information. It is a failure to use information to make decisions, and to make those decisions at the right level, at the right moment.
This page covers four interconnected dimensions of adaptive implementation:
* **Experimenting and prioritising**;
* **Monitoring for learning;**
* **Deciding and adapting;**
* **Communicating change**.
***
### **Try — experimenting and prioritising in short cycles**
An adaptive intervention does not wait for the perfect solution before acting. It starts from a working hypothesis and tests it through short, repeated cycles of action and observation.
**Work in short cycles.** Rather than planning implementation in long sequences, adaptive management works in short iterations based on the project's milestones, at the end of which the team reviews what was done, what was learned, and what the priorities are for the next cycle. Each cycle produces a short list of priority actions, reviewed and updated continuously. This keeps the plan connected to reality rather than to what was assumed at the start.
**Prioritise relentlessly.** In complex contexts, not everything can be done. Choosing what matters most, and also what to drop or defer, is one of the most important management skills in adaptive implementation. A few criteria that help prioritise:
* **Criticality** — is this action essential to achieving the purpose? What happens if it is not done?
* **Immediate feasibility** — can this be done now, or does it depend on something else being completed first?
* **Expected return** — will this action move the intervention meaningfully forward, or is it peripheral?
* **Risk** — what is the likelihood of negative consequences if this action is poorly designed or implemented?
* **Impact on the target outcome** — how directly does this action contribute to the change being sought?
> *"What is important is rarely urgent and what is urgent is rarely important."* — Dwight D. Eisenhower
**Address access challenges proactively.** In complex or insecure contexts, the ability to implement often depends on managing access, both physical access to areas of intervention and the acceptance of the team by communities and authorities. Neither can be improvised. Acceptance strategies require sustained relationship-building investment, and remote management configurations (from remote control to genuine remote partnership) each carry different implications for where decision-making authority lies and how adaptation actually happens.
| Configuration | Decision-making | Implementation |
| ------------------------ | ---------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------- |
| Remote control | Majority of decisions taken by international staff relocated to a safer environment | National/local staff |
| Remote management | Temporary and partial delegation of authority to national staff | National/local staff, partners |
| Remote support/oversight | Long-term strategy to transfer decision-making to national/local actors, while maintaining financial and strategic oversight | National/local staff, partners, contractors |
| Remote partnership | Significant transfer of responsibility to local actors | Local partner or contractor |
The important point is not which configuration is used, but whether the decision-making arrangements that come with it are explicit, understood by all parties, and genuinely aligned with the level of capacity and trust that exists.
***
### **Measure — monitoring for decisions, not for reports**
Adaptive management places different demands on monitoring systems than traditional project management does. The purpose of monitoring in an adaptive intervention is not primarily to report on what was done but it is to generate the information needed to decide what to do next.
**Prioritise effective changes over planned activities.** Traditional monitoring tends to track what was done (activities completed, inputs delivered, outputs produced) because these are easier to measure. Adaptive monitoring tracks what is changing in relation to the overall objective: changes in the lives of communities, in the dynamics of the context, in the assumptions underlying the project strategy. The question is not "did we implement?" but "are we contributing to the change we set out to achieve?"
**Use flexible and empirical measurement.** In complex interventions, pre-defined indicators quickly become obsolete. An adaptive monitoring system needs to be able to find and share priority information even when that information has changed since the project began. This means:
* Reviewing indicators regularly (i.e. at each programme cycle) and adjusting them when the situation changes
* Privileging empirical measures over predictive ones: tracking what is actually being achieved, not what was planned
* Integrating monitoring, evaluation and learning into a single continuous cycle rather than treating them as separate activities
**Distinguish learning from lessons learning (or capitalisation).** These two concepts are often confused, but they serve different purposes:
* **Learning** is the process of acquiring new knowledge through reflection on experience, with the immediate purpose of improving current practice. It happens continuously, in short cycles, throughout implementation.
* **Capitalisation** is the process of formalising and sharing knowledge derived from experience, intending to make it usable for others (i.e. other projects, other organisations, future interventions). It requires more structured processes and more time.
Both matter. Confusing them, treating capitalisation as the only form of learning, or assuming that learning automatically produces transferable knowledge, leads to either missed opportunities for adaptation during implementation, or poorly documented experience that cannot be used by others.
**Use appropriate tools for learning in complex contexts.** Some evaluation and monitoring approaches are better suited to adaptive management than others:
* **Outcome mapping** and **outcome harvesting** are designed to track behavioural changes in people and organisations, which are the kind of changes that matter most in complex interventions, and to identify unexpected as well as intended outcomes
{% file src="/files/rmK9a9YasSlEgZzQFoJz" %}
{% file src="/files/NauNlKoin13u4aVwU3s0" %}
* EIMS (Iterative Evaluations with Mini-Seminars) combines short learning cycles with structured collective analysis, integrating evaluation into ongoing project management rather than treating it as an end-of-project exercise
{% file src="/files/tDoSsCsaNRo6d6stI2ZI" %}
{% file src="/files/qnVufjYygCechIavCbXv" %}
* Most Significant Change collects and analyses stories of change, allowing teams to surface important outcomes that are difficult to capture through pre-defined indicators
{% content-ref url="/pages/OpoV422kttqtU04MDQXQ" %}
[Learning and adapting](/compass/learning-and-adapting/learning-and-adapting.md)
{% endcontent-ref %}
**Use informal information that falls below the radar of traditional monitoring systems**. A great deal of valuable information is sometimes difficult to collect and make use of in traditional monitoring systems, as it does not fit into quantifiable indicators and is therefore overlooked along the way. This is a shame, particularly as such information is often highly qualitative and focuses on the stakeholders themselves. One way to capture this information on an ongoing basis is to monitor what are known as ‘weak signals’ – subtle indicators of emerging changes. These can take various forms: verbal feedback, information in an email, elements of a speech, a passage in a document, direct observations, and so on.\
The value of monitoring weak signals lies in tracking all the small, sometimes insignificant indicators that suggest a change is underway or has been achieved, and recording them continuously in a designated matrix.
After a certain period of time, weak signals can be used to: 1) carry out a learning exercise; 2) inform an evaluation process (internal or external); 3) enter a cycle of measuring the impact.
***
### **Adapt — making the right decision at the right level**
Good information is necessary but not sufficient for adaptive management. The information needs to translate into decisions, and decisions need to be made at the right level, by the right people, at the right moment.
**Three levels of adaptive decision-making.** Oxfam's comparative analysis of adaptive programmes identifies three distinct levels at which adaptation happens, each with different implications:
* **Adaptive delivery** — the day-to-day work of field teams, navigating constantly shifting conditions. Adjustments at this level are tactical: changing opening hours, adjusting meeting locations, modifying data collection methods. These decisions should be delegated to the field team as far as possible, with minimal approval requirements.
* **Adaptive programming** — changes to what is being done or how resources are allocated, while remaining within the agreed strategy. These decisions require programme manager involvement and may require donor notification.
* **Adaptive governance** — changes to the fundamental strategy or purpose of the intervention, triggered when evidence suggests that the original theory of change is no longer valid. These decisions require senior leadership and donor engagement, and cannot be made unilaterally.
> For more information, see
**Apply the principle of "optimal ignorance".** Robert Chambers introduced the concept forty years ago: the question is not how much information is needed before acting, but how much information is needed to make a responsible decision at a given level. In adaptive management, this means clarifying, for each level of the organisation, what the minimum information threshold is for making a decision, rather than waiting for certainty that will never come.
**Make decisions empirically.** Adaptive decisions are based on two postulates: knowledge comes from experience, and decisions are made based on what is known at the time. This is not a licence to act without evidence but rather a recognition that evidence in complex situations is often incomplete, contradictory and contested. The response to this is not paralysis but structured empiricism: seeking multiple perspectives, triangulating information sources, and being transparent about the uncertainties that remain.
***
### **Communicate — making adaptation understood and accepted**
Making the right decision is not enough. Adaptation only works if it is understood and accepted by the people affected by it, whether they are communities, partners, donors and internal teams.
**Communicate proactively, not reactively.** Changes should be anticipated and communicated before they happen, not explained after the fact. When a significant adaptation is likely (i.e. a change of target population, a shift in geographic focus, a modification of an activity), the relevant stakeholders should be involved in the discussion before the decision is finalised. This is not just good communication practice. It is an expression of *CHS Commitment 4: communities and people affected by the intervention have a right to be informed about decisions that affect them, and to influence those decisions*.
**Involve support teams early.** Adaptation decisions have consequences not just for implementing teams but for finance, logistics, HR and procurement. These teams are often designed for stable, predictable contexts and their procedures may not flex easily when the programme changes. Involving them early, explaining decisions clearly, and working with them to find creative solutions is more effective than presenting them with a fait accompli.
**Build and maintain trust with all stakeholders.** In volatile contexts, trust is the most important enabling condition for adaptation. When stakeholders (i.e. communities, partners, donors) trust that adaptations are made for the right reasons, with their interests in mind, they are far more likely to accept them. Building this trust requires consistent, honest communication and genuine participation throughout the project, not just at moments of crisis or change. It requires time, energy and endurance: it needs to be maintained constantly, and rebuilt every time a new person joins the team or a new partner comes on board.
**Document adaptations and the reasoning behind them.** Every significant adaptation should be recorded: what changed, why, who decided, and what the expected effect is. This documentation serves three purposes: it supports accountability to donors and partners, it enables the team to track whether adaptations are having the intended effect, and it provides the raw material for continuous learning.
***
### **Key points**
* Adaptive implementation works in short cycles of action, observation and adjustment, not long sequences of planned activities
* Monitoring in an adaptive intervention tracks results / effective changes and context change, not just activity completion
* Learning and capitalisation are distinct: learning drives adaptation during implementation; capitalisation makes experience usable for others
* Three levels of adaptive decision-making: delivery, programming, governance. Each require different authority and different processes
* Communication is not an afterthought: it is an integral part of adaptive management, requiring proactive engagement with all stakeholders, not just reporting after the fact
### **Links and toolbox**
{% content-ref url="/spaces/8VQvJtfuTFNUeSv1dIbV/pages/r0ZYgBncaIZwoiOnEFjd" %}
[Adaptive management](/compass/documentation/adaptive-management.md)
{% endcontent-ref %}