Evidence of Impact

Research-backed outcomes for student success

Mainstay helps institutions turn timely conversations into measurable gains across enrollment, persistence, financial aid, and course performance.

Peer-reviewed studies Working papers A/B tests Ongoing evaluations

Landmark RCT

21.4%

reduction in summer melt in the Georgia State University study.

+3.3pp on-time enrollment in GSU summer melt RCT
+4pp earning A or B in the GSU randomized trials
-50% withdrawals due to balances in GSU retention outreach
+14% early fall registration in GSU Perimeter RCT

Confirmed research

Research that follows students from intention to action

Explore rigorous studies, working papers, and partner evaluations showing how personalized outreach helps students take the next right step.

Showing all 8 studies and research partnerships.

Study 01 Peer-reviewed

GSU Summer Melt: Pounce virtual assistant

Page & Gehlbach, AERA Open, 2017. n = 7,489.

Georgia State used conversational outreach to help admitted students clear the final hurdles to enrollment, from forms and financial aid to deadline-sensitive next steps.

-21.4% summer melt reduction
+3.3pp on-time enrollment
Study 02 Peer-reviewed

ECU Summer Melt: first-generation student outcomes

Nurshatayeva, Page, White & Gehlbach, Research in Higher Education, 2021. n = 4,442.

East Carolina University extended summer melt support with personalized guidance, helping first-generation students make progress on financial aid and enrollment steps.

+8pp loan acceptance for first-gen students
+3pp registration and fall enrollment for first-gen students
View study
Study 03 Peer-reviewed

GSU Retention: text-based nudging conditions

Page, Meyer, Lee & Gehlbach, Journal of Research on Educational Effectiveness, 2025. GSU Atlanta n = 13,657; GSU Perimeter n = 11,561.

Georgia State used timely, targeted outreach to help students resolve the administrative barriers that can quietly derail persistence.

-50% withdrawals tied to balances
+8pp students resolving registration holds
Study 04 Case study / A-B test

West Texas A&M FAFSA: behavioral design test

Irrational Labs + AdmitHub case study, 2018. n = 2,000+.

Behaviorally designed messages helped students see FAFSA completion as a natural part of enrollment, increasing on-time completion for students who had not applied the prior year.

3x odds of applying on time for prior non-applicants
+3pp on-time completion in reported subgroup
View case study
Study 05 Peer-reviewed

GSU Course Performance: "Let's Chat"

Meyer et al., Economics of Education Review 115 (2026), 102816.

Across two large, asynchronous online courses at Georgia State, 2,483 students were randomized to receive proactive, course-specific outreach or standard course communications.

+4pp more likely to earn A or B
2,483 students randomized across two courses
Study 06 Working paper

CSU Northridge: scaling conversational AI support

Mata, Russell & Page, EdWorkingPaper 26-1409, 2026. n ~ 8,700.

Across four years, CSU Northridge scaled conversational support for high-volume student needs, improving completion of time-sensitive tasks while adding nuance to where AI support has the greatest impact.

4 years randomized evaluation window
Administrative action where the clearest gains appeared
View paper DOI
Study 07 Active research

WSAC / MDRC OtterBot: FAFSA support optimization

MDRC + WSAC + Mainstay, OTTERS Project. Report published March 2025.

Washington partners are improving OtterBot through research-backed message design, helping more students engage with financial aid guidance when deadlines matter most.

Engagement alternative messages increased OtterBot engagement
Male students significant FAFSA completion lift in A/B test
Study 08 Evaluation partnership

Common App / College Advising Corps: Oli support campaign

Common App + AdmitHub/Mainstay + College Advising Corps, launched 2020.

Common App, College Advising Corps, and Mainstay brought Oli to students at national scale, expanding access to timely college-going support during a uniquely challenging admissions season.

173k students reached in 2020 campaign
23M+ student-support messages

Why it matters

Better student support starts with better evidence.

Mainstay partners with institutions and researchers to understand where conversations move students forward, and where support needs to be sharper.

Rigorous research Peer-reviewed studies and randomized evaluations show measurable impact in real student journeys.
Practical outcomes The strongest gains appear when outreach helps students complete time-sensitive, high-friction tasks.
Designed for action Behavioral science and conversational AI work together to make the next step clear, personal, and doable.
Always learning Active research keeps improving how institutions reach students and measure what actually changes.

Research library

Go deeper into the evidence

Read the studies, working papers, reports, and partner stories behind the outcomes above.

Book a call with us:

Get a guided tour of Mainstay’s Engagement Platform

Mainstay’s Engagement Platform uses Behavioral Intelligence to help businesses build meaningful relationships, optimize insights, and drive action.

Sign up for your personalized demo to find out how Mainstay makes it easy to spark conversations with your employees and guide them toward positive results.

Channels

Instant support — anytime, anywhere

Mainstay’s behaviorally intelligent chatbots make it easy to engage with employees on the channels they prefer to use:

SMS text messaging
Live chat
Web chat
Social media

Let's talk

Get a guided tour of Mainstay’s Engagement Platform

Mainstay’s Engagement Platform uses Behavioral Intelligence to help schools and businesses build meaningful relationships, optimize insights, and drive action.

Sign up for your personalized demo to find out how Mainstay makes it easy to spark conversations with your students or employees and guide them toward positive results.

Channels

Instant support — anytime, anywhere

Mainstay’s behaviorally intelligent chatbots make it easy to engage with students and employees on the channels they prefer to use:

SMS text messaging
Live chat
Web chat
Social media

Let's talk