Are You Interested in A/B Testing? Partner with Youth Impact.
Are you interested in carrying out A/B testing at your organization but are unsure of how to get started?
We can help.
Fill out this brief, 10-question survey about your interest and we will reach out about next steps. Youth Impact has funding to support a limited number of organizations to start or expand their A/B testing practice.
What is A/B testing?
A/B testing is a rigorous evaluation methodology used to optimize a program. A/B testing typically involves making a tweak or variation to an existing program. It uses randomization (i.e., randomly allocating students or classes or schools or groups to program versions A or B), which allows us to be confident that any differences in impact are causal, meaning they can be attributed to the differences in the program versions, rather than other contextual factors. Understanding the differences in impact of different program versions allows implementers to adopt the best-performing, most cost-effective, and/or most scalable program model. Using A/B testing as an ongoing part of an organization’s M&E practice allows for continuous learning and program improvement.
How does Youth Impact support organizations in carrying out A/B testing?
We offer technical support from Youth Impact researchers and implementers to carry out A/B testing and offer financial support that allows you to get started with A/B testing.
We take several steps to support fellow NGO partners in getting started:
-
Assess suitability and fit: we outline the necessary conditions for A/B testing (e.g. indicators, number of beneficiaries, data systems, use case, etc) and help organizations decide if they are ready to carry out A/B testing.
-
Follow a three-stage orientation: we help organizations ease into the A/B testing process by asking them to carry out three steps
We recommend a phased approach to help prime organizations for A/B testing
Pilot implementation tweak
Implement a small program variation and collect outcome data.
Main focus is implementing an easy tweak to gear up for the larger-scale next phase.
Data flow
Collect high-frequency data on a useful indicator from a large number of sites (similar number as would use for A/B testing).
This is the backbone of any ongoing M&E learning infrastructure.
A/B testing process
Successfully execute a first A/B test.
Adapt the program based on results. Plan for the next test.
Tip: First question doesn’t need to be the best. Try an easy one to get things moving and build the system and momentum.
How do I work with Youth Impact?
Fill out this brief, 10-question survey and we will follow up to get started on the suitability and fit step above.