Below is a step-by-step approach to guide you. The examples used refer to gender-based violence, though the same reasoning could apply equally well to other areas.

1) Applicants should provide a clear description of the issue to be tackled by the proposed measures. Taking gender-based violence as an example, the description should specify the type(s) of violence concerned (physical and/or emotional, for example), its characteristics (such as prevalence and severity) and the specific context(s) (such as university, workplace or household) to be covered by the proposed measures. It should also clearly define the measures’ objective(s).

2) The proposed measures should identify the underlying determinants (causes) of the issue at stake. These could include such factors as bystanders’ passivity or victims’ under-reporting of violence. As the issues identified under point 1 are likely to have multiple underlying determinants, the proposed measures could focus on several of these or on a single one. The proposal should clearly state each of the underlying determinants the project is designed to tackle.

3) The proposed measures should encourage a specific target group to engage in target behaviour or discourage them from doing so. Measures may, for instance, be designed to:
- encourage victims of violence to report incidences of violence to the relevant authorities and/or to seek help in shelters;
- change professionals’ work practices, to make it easier to report incidences of violence;
- motivate journalists to avoid stereotypes, by changing the way they report on and depict violence against women; or
- change men's perception of traditional gender roles, which support the acceptability of violence against women.

4) Once the target group, the underlying determinants and the target behaviour are identified, applicants should describe the solutions that will be implemented. For example, if the proposed measures are designed to reduce social acceptance of gender-based violence, the project could focus on changing social norms (for example, signalling the appropriate behaviour by comparing an individual’s conduct with that of the majority of his/her peers/neighbours/friends), or make use of 'commitment devices’ (persuading the target group to make a public pledge to engage in a certain type of behaviour or supporting the target group in developing a specific plan to achieve a certain goal) as part of the intervention.

5) The proposed measures should specify the outcome that the proposed solution is designed to achieve and how the outcome will be measured. Outcomes refer to impact, change in attitudes or behavioural change resulting from the measures, and should not be confused with project outputs (e.g. number of leaflets published, number of people reached). If possible, projects should propose a measurable objective that can be compared with a baseline. For example, if the goal is to increase reporting of violence, a valuable outcome measure would be the number of cases reported to the police following the implementation of the project. If the goal is to change the attitudes of a particular group of professionals, projects should not only measure how many attended training, but also ask them about their attitudes before and after training. All interventions should include collection of relevant data, to be able to compare the situation before and after the project.

6) To check whether the proposed measures will have the desired effects, the project should ideally include a pre-testing phase before solutions are implemented full-scale. Such prior assessment of the effect the intervention is expected to produce will provide valuable insights into whether the proposed action has the intended effects, thereby helping to ensure that, once the measure is fully deployed, it focuses on what works. One possibility would be to use a randomised controlled trial.

7) The proposed measures should include an impact evaluation. This should be part of the initial project plan, to make sure that the most appropriate evaluation design is used and to set up valid measures of outcomes that will enable the measures’ actual impact to be ascertained. Several evaluation methods can be used, such as after-only designs, before-after or pre-post designs, randomised control trials and field experiments.