What happens when schools facing similar challenges work together—rather than alone—to improve outcomes for students? What becomes possible when they share not just their goals but also what they are learning along the way, helping one another build a clearer map of the path to improvement?
This question drove the Gates Foundation’s Networks for School Improvement (NSI) initiative. From 2018 to 2026, educators in nearly 800 schools, with support from intermediary organizations, worked across classrooms, schools, and districts to improve outcomes for students, especially those furthest from opportunity, such as Black and Latino students, and students experiencing poverty. They focused on critical transition points—entering and progressing through high school, and moving into postsecondary education—where intentional changes can have significant effects on students’ trajectories.
The NSI initiative was ambitious in both scale and purpose. It aimed not only to improve student outcomes and experiences, but also to build the capacity of schools and systems to keep improving over time.
How NSI worked
At the heart of NSI was a simple idea: schools can make more progress when they work in structured networks, supported by disciplined methods for learning what works. When schools share what they are seeing and testing, they can create a fuller map of the path to improvement—helping one another spot obstacles, avoid dead ends, and build on what is working.
Schools were organized into networks led by intermediary organizations that provided structure, coaching, and regular opportunities for educators to learn from one another while working toward shared outcomes.
Rather than trying to solve every challenge at once, school teams focused on specific problems connected to key student transition points. For some, that meant reducing 9th-grade course failures or chronic absenteeism. For others, it meant increasing FAFSA completion or enrollment in postsecondary institutions that meet students’ needs and goals.
Teams used disaggregated data and student feedback to understand why, and for whom, a problem was happening. They set clear aims, tested changes in practice on a small scale, studied short-cycle results, and then refined, abandoned, or spread those changes based on what they learned.
That approach matters. NSI encouraged educators to test bold ideas first on a small scale, creating opportunities to learn quickly and adapt.
What the evaluation found
NSI leaves behind robust and causal research to help inform future decisions about improvement at scale. A just-released evaluation from RAND, AIR, and Mathematica found that networked continuous improvement can help schools and systems build real capacity to solve problems together. The results showed significant improvements for some student outcomes based on a rigorous analysis that included a randomized controlled trial. Across many sites, educators strengthened their ability to use data, student voice, and disciplined inquiry to improve practice. Schools developed shared resources, tested change ideas, and learned from one another in more structured ways than is typical in traditional school improvement efforts.
Researchers also identified conditions that are related to success: engaged leadership, coaching focused on building local capacity, and clear network design and expectations. Certain practices—such as developing an aim statement, defining a theory of practice, and adopting a change idea—were significantly related to positive impact on student outcomes.
Another important lesson is the value of both strong facilitation and protected time during the school day. When educators had dedicated time to work together—and skilled support to guide that work—they were better positioned to use improvement methods consistently and effectively. This insight has informed the foundation’s work on scheduling and staffing as part of our math strategy.
In some types of networks, the results were especially promising. These sites saw meaningful improvements in outcomes such as 9th-grade GPA, courses passed, credit accumulation, and FAFSA completion. That progress is especially notable given that much of NSI unfolded during the disruption of COVID-19. Even as the pandemic interrupted early implementation, some networks continued to make gains, underscoring the value of responsive, data-informed improvement methods.
A promising approach with room to grow
It is equally valuable to be clear about what NSI did not show.
The initiative did not produce the same level of impact everywhere. Some networks saw impressive gains, while others showed no measurable effects. That variation matters. It reminds us that while networked continuous improvement is a promising approach, its impact depends on how well it’s implemented, and on its own, it does not guarantee results.
Success depends on more than adopting a process. It requires the conditions, supports, and leadership needed to make that process meaningful and sustainable. It also requires clarity about the problem being addressed, discipline in how improvement work is carried out, and the ability and willingness to learn from both progress and setbacks.
NSI reinforced for the foundation the importance of coherence across curriculum, instruction, and professional learning. Those lessons have informed the foundation’s math strategy and its focus on high-quality instructional materials, aligned supports for teachers, and the system-level conditions that enable sustained improvement in math teaching and learning.
What comes next
As the foundation’s formal support for NSI ends, the work does not end here.
This initiative leaves the field with tested methods, practical lessons, and a stronger evidence base for future efforts. Most importantly, it reinforces something educators have long known: meaningful improvement rarely comes from schools working in isolation. It comes from shared commitment, disciplined learning, and sustained attention to the outcomes that matter most for students.
That is the real legacy of NSI: not a single route or formula, but a deeper understanding of how schools and systems can learn together, share knowledge, and build a clearer map for improving outcomes for students.