Sketch your evaluation framework

As you develop your logic model, identify what your measures of success are. These metrics will help you monitor the initiative and communicate the outcomes to the broader ecosystem. It is helpful to define indicators at all the different levels of your logic model, i.e. for the inputs, outputs, outcomes and impact.

Think about:

• identifying a set of measures that will be used to establish the success of the programme. For example:

  • the number of organisations submitting data using your standard

  • the number of new products and datasets created thanks to your initiative

  • an improvement in the data infrastructure your initiative relies upon

  • the number of people that use your tools or read your use cases

  • the number of programs aligned with your initiative that receive new funding

• establishing targets which define success

For example, the number of companies adopting your standard is a measure of success, so depending on the wider context, 10 companies adopting the standard by the end of the year could be considered a success.

We recommend looking at the work that the Open Contracting Partnership has been doing to openly monitor and evaluate their progress on a quarterly basis. To help your team understand the value of monitoring and evaluating progress and the ways in which this can be done, they have created a set of guidance to help better understand how to monitor your progress, and develop useful indicators.

Note that measuring success against outcomes and impacts might involve proactive pieces of work, such as carrying out surveys or targeted research. The results of this evaluation can also be useful to motivate stakeholders, so consider including publishing adoption reports, worknotes and case studies as part of your programme design, as well as your monitoring and evaluation process. Over time, you can refine the measures you use alongside your logic model, as you build up a better understanding of the data infrastructure needed to improve access to data.