Following on from this information-gathering phase, a strategy should be developed. This strategy should choose between a confrontational or a more collaborative approach. The choice should be based on an assessment of whether there is some potential within the police to effectuate change at all. If there is none, if there are no points of common interest, if the police is involved in grave human rights violations without there being internal support to stop these, engagement is out of the question.
It can be helpful to organize an internal workshop in which strategy and methodology can be further discussed with those involved. Inviting other partners, most notably fellow NGOs, can be useful. However, this can also be postponed until after an internal strategy has been developed. During a workshop, the following issues need to be addressed: 

1. Identify main concerns 
2. Identify overall goals 
3. Are there points of mutual interest with the police? If so, specify 
4. What is the national strategy for police engagement work (within the context of the Work On Own Country policy & guidelines)? 
5. Draft a risk assessment of potential engagement work 
6. What further expertise does the section need? 
7. Are there sufficient resources (finance, time)? 
8. Are there any established police contacts that may support reform and which could help to identify areas of intervention? As a rule, always verify these with other organizations affected (political opposition, other NGO’s, journalists, etc.) or with individuals that are familiar with the situation (academics, prosecution, magistrates)
9. Decide: Is engagement viable?

It can be helpful to include an assessment of the main pitfalls and seek ways to deal with them. For example, if a lack of commitment from police top management is indicated, a strategy needs to be developed to address this. Making use of ‘friendly’ senior police officers from other countries may be useful, as they may be able to open doors that would otherwise stay closed.

The result of this analysis may be that a collaborative effort is either too ambitious or simply inappropriate, e.g. because the police seem to be systematically and institutionally involved in human rights violations. It can also result in the conclusion that other NGOs may be better equipped to engage with the police.