The Empathy Project tests whether four one-to-one audio calls across cultures, framed by a Brief, champion check-ins, guided reflection and a team Debrief, help employees listen more carefully and understand other people more accurately.
Figures combine an August 19, 2026 law-firm snapshot with a September 9, 2026 Dolby update; the September 2026 pilot learning report publishes them for a law firm and a technology company. Hours were reported by Acquaint on September 18 and count both sides of each conversation.
Most employees spend their working days among people who share their language and much of their outlook. Intergroup contact research gives a reason to take structured one-to-one encounters seriously, and the Empathy Project asks whether a repeated, supported version of that contact, placed inside a company’s volunteering program, helps people listen more carefully and read other people more accurately.
The project is designed as a learning program. Each cohort is meant to contribute anonymized aggregate learning about empathy development, behavioural practice and intergroup contact, with the aim of learning what shifts, for whom, and under what conditions. RW Institute holds the overall design, theory of change, champion model and research governance; Acquaint runs the conversations, and a dedicated reflection space from Inqli is coming soon.
Both people in every conversation are volunteers. An employee gives about seven hours across six weeks, and the conversation partner, recruited through Acquaint’s community and networks such as Volunteer Groups Alliance and Idealist, gives about four. The design counts the partner as a participant in the exchange, so partner-side outcomes are a research question in their own right.
The design follows the Brief, Check-in, Debrief cycle RW Institute uses across its volunteering work. Select an element to see what happens, the theoretical role it is meant to play, and what the founding cohorts received.
One-to-one audio calls, arranged by Acquaint, with a volunteer from another culture. Light prompts move past small talk without scripting what either person should discover.
Repeated, cooperative contact in which both people take part as volunteers, with institutional support from the company. The design aims at several of Allport’s optimal conditions; whether it achieves them is an empirical question.
34 of 41 employees joined at least one conversation and 14 reached four or more across the two founding cohorts.
Employee time is a one-hour Brief, four one-hour conversations, an hour of guided reflection and a one-hour Debrief. Counting both sides matters for the research as well as the accounting, since the partner also gives time, stories and attention.
The seat price, set in October 2026 after pilots at US$35, is paid as a charitable contribution and published with its full breakdown. Whether cohorts reach the size a well-powered comparison needs is one of the questions for research partners.
The project’s canonical logic model sets its expectations against meta-analytic evidence. It treats anxiety reduction as the primary mechanism, expects small to medium effects that decay without reinforcement, and labels several outcomes as aspirational.
Adapted from the logic model, which was written when the design was six hours at US$35 per seat. Inputs show the current seven-hour, US$70 design, and “more accurate understanding” is added from the pilot report’s research priorities.
These are the logic model’s working estimates, informed by the literature but not taken from any single study. The pilot did not measure them. Select an outcome to see the expected effect, its likely durability, and what it implies for measurement.
Assess anxiety about intercultural interaction at baseline, immediately after the program and at follow-up. The expected decay makes timing as important as the endpoint.
Eight bodies of work shape what the project expects and how it plans to measure. Some support the design directly; others, such as the corporate volunteering meta-analysis, argue for restraint.
Across 515 studies and more than 250,000 participants, intergroup contact typically reduces prejudice. In structured programs meeting Allport’s conditions the mean was about r = −.29, though the conditions were not essential, and effects typically extended to the whole outgroup.
A consistent but modest prior, which points to incremental change.
Reduced anxiety and increased empathy and perspective-taking mediate contact effects more strongly than knowledge about the other group. The authors caution against direct causal inference.
Conversations prioritize connection and personal stories over cultural instruction.
Across 18 randomized trials, empathy training showed a medium effect, about 0.51 after adjusting for publication bias. Training length was not linked to effect size, and objective measures showed larger effects than self-report.
A medium effect is achievable in trials; whether a seven-hour, conversation-based design reaches it is untested.
Across 27 controlled workplace trials training emotional competencies, empathy included, the effect was moderate and lasted beyond three months, though study quality was low.
Workplace training effects can last, but the evidence base is uneven.
Across 57 sources, individual participation in corporate volunteering showed small or nonsignificant relations with well-being, commitment, job satisfaction and positive behaviours. Organizational participation related to all four, and more strongly for two.
Engagement and retention are not treated as primary outcomes of individual participation.
Across 25 experiments, imagining another person’s perspective did not improve accuracy about their thoughts and sometimes reduced it. In a final experiment, asking them directly did.
Measure accuracy, and whether participants ask and check what others mean.
Across 29 effect sizes and 6,576 participants, online contact was linked to less prejudice, more so when people were asked to cooperate. A 2024 meta-analysis found a smaller digital-contact effect (g = 0.25), with chat and video calls doing best among digital formats.
Digital estimates run below the general contact literature, so expectations for audio calls are set lower.
Effects of mostly educational interventions commonly fade after treatment ends, often alongside some persistence. Pathways to persistence include building habits, changing what and how people think, changing future costs and harnessing external reinforcement.
Returning cohorts, councils and mentoring are the candidate reinforcement routes.
Between June and September 2026 the first cohorts ran with an international law firm and Dolby. The data are participation snapshots and appreciation notes, so they describe engagement and what people valued, and were not designed to measure change.
Percentages are computed here from the published counts, relative to sign-ups. The cohorts were at different points in delivery, so the snapshots should not be read as a comparison between companies. Four conversations do not establish completion of the Brief, reflection or Debrief.
Reported by Acquaint on September 18, 2026, counting employees and partners together. A one-hour call between two volunteers adds two hours. Not independently recomputed.
54 notes are associated with 9 law-firm participants and 31 with 6 Dolby participants. They are written to say thank you, so they show what people valued and cannot stand for everyone’s experience.
Partners appreciated employees who remembered details and showed patience; employees valued partners who helped them through first calls and the platform.
Some accounts describe familiar interests emerging alongside very different lives. The accounts resemble parts of the contact processes, though anxiety was not measured.
Participants discussed assumptions through the stories they exchanged. The notes rarely explain an original belief and how it changed.
One employee reported sharing something learned with others afterward, and other notes express intentions to explore further. Both give follow-up something concrete to ask about.
“The experience with your platform and connecting with people around the world is amazing.”
Conversations can give people experiences worth returning to and thinking about. Participants describe reciprocal attention, nervousness giving way to comfort, and curiosity, and one employee reported sharing what was learned with others afterward.
Lasting empathy, more accurate understanding, changed assumptions or prosocial identity change. Anxiety and prejudice were not measured, and the pilot cannot isolate the contribution of the full design from the conversations alone.
Each limit below is named in the project’s own logic model or pilot report. Open one to see why it matters and how the next phase proposes to respond.
The project’s planning rule assumes about a quarter of employees are inclined to volunteer. That group may differ from colleagues in baseline empathy, openness and prosocial motivation, so effects may be smallest where they would matter most.
Proposed: describe who enrolls against the wider workforce, and treat generalization to non-volunteers as an open question.
RW Institute is exploring three research strands, and academic partnerships are in development. The steps that follow are proposals for research partners to revise and agree.
Whether and how individual experience travels into team practice, the question the logic model warns against assuming.
Behavioural design of invitation, scheduling and return, given that 22 active participants were below four conversations at the latest reporting.
What volunteering does for employees, studied with the restraint the corporate volunteering meta-analysis calls for.
For every participant, track whether they actually received the Brief, champion check-ins, reflection prompts and Debrief, alongside how many conversations they had.
A short assessment at the start and a follow-up later, so change over time can be examined.
A comparison design, agreed with research partners, to estimate what the full design adds beyond the conversations alone.
Follow returning participants to see whether listening becomes more deliberate and unfamiliar encounters become easier to enter.
Ask whether people check what the other person means, including when they believe they already understand.
Include conversation partners and people who stop partway, while keeping private reflections private and sharing only approved, combined results with companies.
Unresolved design choices where academic input would change what the project does.
Private reflections stay in participants’ own notes. People choose whether to share a question with their cohort.
The project is designed to protect individual responses and share only approved, combined results with participating companies.
Research questions and consent are planned to be agreed with academic partners in the first quarter of 2027, before starting assessments.
Pilot participant material is summarized as broad, unattributed paraphrase. Only a testimonial offered for use is quoted by name.
Dates from the fourth quarter of 2026 onward are plans and may move as research partners and companies join.
At a salon on October 14, 2026, researchers in management, volunteering and organizational behaviour discussed what these early signals can actually tell us. The conversation continues with anyone willing to help shape the next phase, and to say what it should not claim.
Help choose constructs, instruments and timepoints, and name what the project should not claim from them.
Advise on a comparison design that respects voluntary participation and company constraints.
Contribute to work on spread through teams, enrollment and persistence, or well-being and retention.
Bailey, D. H., Duncan, G. J., Cunha, F., Foorman, B. R., & Yeager, D. S. (2020). Persistence and fade-out of educational-intervention effects: Mechanisms and potential solutions. Psychological Science in the Public Interest, 21(2), 55–97. https://doi.org/10.1177/1529100620915848
Barlow, F. K., Paolini, S., Pedersen, A., Hornsey, M. J., Radke, H. R. M., Harwood, J., Rubin, M., & Sibley, C. G. (2012). The contact caveat: Negative contact predicts increased prejudice more than positive contact predicts reduced prejudice. Personality and Social Psychology Bulletin, 38(12), 1629–1643. https://doi.org/10.1177/0146167212457953
Eyal, T., Steffel, M., & Epley, N. (2018). Perspective mistaking: Accurately understanding the mind of another requires getting perspective, not taking perspective. Journal of Personality and Social Psychology, 114(4), 547–571. https://doi.org/10.1037/pspa0000115
Frey, E., & Rogers, T. (2014). Persistence: How treatment effects persist after interventions stop. Policy Insights from the Behavioral and Brain Sciences, 1(1), 172–179. https://doi.org/10.1177/2372732214550405
Howard, M. C., & Serviss, E. (2022). The antecedents and outcomes of corporate volunteering: An employee- and organizational-level meta-analysis. Journal of Managerial Psychology, 37(2), 93–110. https://doi.org/10.1108/JMP-01-2021-0018
Imperato, C., Schneider, B. H., Caricati, L., Amichai-Hamburger, Y., & Mancini, T. (2021). Allport meets internet: A meta-analytical investigation of online intergroup contact and prejudice reduction. International Journal of Intercultural Relations, 81, 131–141. https://doi.org/10.1016/j.ijintrel.2021.01.006
Mehler, M., Balint, E., Gralla, M., Pößnecker, T., Gast, M., Hölzer, M., Kösters, M., & Gündel, H. (2024). Training emotional competencies at the workplace: A systematic review and metaanalysis. BMC Psychology, 12, Article 718. https://doi.org/10.1186/s40359-024-02198-3
Mezirow, J. (1997). Transformative learning: Theory to practice. New Directions for Adult and Continuing Education, 1997(74), 5–12. https://doi.org/10.1002/ace.7401
Newman, A., Donohue, R., & Eva, N. (2017). Psychological safety: A systematic review of the literature. Human Resource Management Review, 27(3), 521–535. https://doi.org/10.1016/j.hrmr.2017.01.001
Pereira da Costa, L., Bierwiaczonek, K., & Bianchi, M. (2024). Does digital intergroup contact reduce prejudice? A meta-analysis. Cyberpsychology, Behavior, and Social Networking, 27(7), 440–451. https://doi.org/10.1089/cyber.2023.0591
Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90(5), 751–783. https://doi.org/10.1037/0022-3514.90.5.751
Pettigrew, T. F., & Tropp, L. R. (2008). How does intergroup contact reduce prejudice? Meta-analytic tests of three mediators. European Journal of Social Psychology, 38(6), 922–934. https://doi.org/10.1002/ejsp.504
Schäfer, S. J., Kauff, M., Prati, F., Kros, M., Lang, T., & Christ, O. (2021). Does negative contact undermine attempts to improve intergroup relations? Deepening the understanding of negative contact and its consequences for intergroup contact research and interventions. Journal of Social Issues, 77(1), 197–216. https://doi.org/10.1111/josi.12422
Teding van Berkhout, E., & Malouff, J. M. (2016). The efficacy of empathy training: A meta-analysis of randomized controlled trials. Journal of Counseling Psychology, 63(1), 32–41. https://doi.org/10.1037/cou0000093
RW Institute (2026). The Empathy Project: Pilot Learning Report, updated September 21, 2026. Participation figures, appreciation-note counts and the three 2027 learning priorities.
RW Institute (2026). The Empathy Project overview. Program design, seat price and breakdown, partner roles.
The Empathy Project: A Critical, Evidence-Based Logic Model. Internal canonical document, RW Institute, corrected October 14, 2026; available to research partners on request. Source of the core hypothesis, outcome priors and threats to validity shown here.
Acquaint participation snapshots (August 19 and September 9, 2026) and appreciation-note collections. Unpublished; summarized in the pilot report without identifying participants. Conversation hours: Acquaint correspondence, September 18, 2026.
The Empathy Project tests whether four one-to-one audio calls across cultures, framed by a Brief, champion check-ins, guided reflection and a team Debrief, help employees listen more carefully and understand other people more accurately.
Figures combine an August 19, 2026 law-firm snapshot with a September 9, 2026 Dolby update; the September 2026 pilot learning report publishes them for a law firm and a technology company. Hours were reported by Acquaint on September 18 and count both sides of each conversation.
Most employees spend their working days among people who share their language and much of their outlook. Intergroup contact research gives a reason to take structured one-to-one encounters seriously, and the Empathy Project asks whether a repeated, supported version of that contact, placed inside a company’s volunteering program, helps people listen more carefully and read other people more accurately.
The project is designed as a learning program. Each cohort is meant to contribute anonymized aggregate learning about empathy development, behavioural practice and intergroup contact, with the aim of learning what shifts, for whom, and under what conditions. RW Institute holds the overall design, theory of change, champion model and research governance; Acquaint runs the conversations, and a dedicated reflection space from Inqli is coming soon.
Both people in every conversation are volunteers. An employee gives about seven hours across six weeks, and the conversation partner, recruited through Acquaint’s community and networks such as Volunteer Groups Alliance and Idealist, gives about four. The design counts the partner as a participant in the exchange, so partner-side outcomes are a research question in their own right.
The design follows the Brief, Check-in, Debrief cycle RW Institute uses across its volunteering work. Select an element to see what happens, the theoretical role it is meant to play, and what the founding cohorts received.
{{current.what}}
{{current.theory}}
{{current.pilot}}
Employee time is a one-hour Brief, four one-hour conversations, an hour of guided reflection and a one-hour Debrief. Counting both sides matters for the research as well as the accounting, since the partner also gives time, stories and attention.
The seat price, set in October 2026 after pilots at US$35, is paid as a charitable contribution and published with its full breakdown. Whether cohorts reach the size a well-powered comparison needs is one of the questions for research partners.
The project’s canonical logic model sets its expectations against meta-analytic evidence. It treats anxiety reduction as the primary mechanism, expects small to medium effects that decay without reinforcement, and labels several outcomes as aspirational.
Adapted from the logic model, which was written when the design was six hours at US$35 per seat. Inputs show the current seven-hour, US$70 design, and “more accurate understanding” is added from the pilot report’s research priorities.
These are the logic model’s working estimates, informed by the literature but not taken from any single study. The pilot did not measure them. Select an outcome to see the expected effect, its likely durability, and what it implies for measurement.
{{outcome.measure}}
Eight bodies of work shape what the project expects and how it plans to measure. Some support the design directly; others, such as the corporate volunteering meta-analysis, argue for restraint.
{{e.finding}}
{{e.implication}}
Between June and September 2026 the first cohorts ran with an international law firm and Dolby. The data are participation snapshots and appreciation notes, so they describe engagement and what people valued, and were not designed to measure change.
Percentages are computed here from the published counts, relative to sign-ups. The cohorts were at different points in delivery, so the snapshots should not be read as a comparison between companies. Four conversations do not establish completion of the Brief, reflection or Debrief.
Reported by Acquaint on September 18, 2026, counting employees and partners together. A one-hour call between two volunteers adds two hours. Not independently recomputed.
54 notes are associated with 9 law-firm participants and 31 with 6 Dolby participants. They are written to say thank you, so they show what people valued and cannot stand for everyone’s experience.
{{t.body}}
“The experience with your platform and connecting with people around the world is amazing.”
Conversations can give people experiences worth returning to and thinking about. Participants describe reciprocal attention, nervousness giving way to comfort, and curiosity, and one employee reported sharing what was learned with others afterward.
Lasting empathy, more accurate understanding, changed assumptions or prosocial identity change. Anxiety and prejudice were not measured, and the pilot cannot isolate the contribution of the full design from the conversations alone.
Each limit below is named in the project’s own logic model or pilot report. Open one to see why it matters and how the next phase proposes to respond.
{{t.why}}
{{t.response}}
RW Institute is exploring three research strands, and academic partnerships are in development. The steps that follow are proposals for research partners to revise and agree.
{{s.body}}
{{d.body}}
Unresolved design choices where academic input would change what the project does.
{{g.body}}
Dates from the fourth quarter of 2026 onward are plans and may move as research partners and companies join.
At a salon on October 14, 2026, researchers in management, volunteering and organizational behaviour discussed what these early signals can actually tell us. The conversation continues with anyone willing to help shape the next phase, and to say what it should not claim.
{{a.body}}
{{rf.text}} {{rf.doi}}
RW Institute (2026). The Empathy Project: Pilot Learning Report, updated September 21, 2026. Participation figures, appreciation-note counts and the three 2027 learning priorities.
RW Institute (2026). The Empathy Project overview. Program design, seat price and breakdown, partner roles.
The Empathy Project: A Critical, Evidence-Based Logic Model. Internal canonical document, RW Institute, corrected October 14, 2026; available to research partners on request. Source of the core hypothesis, outcome priors and threats to validity shown here.
Acquaint participation snapshots (August 19 and September 9, 2026) and appreciation-note collections. Unpublished; summarized in the pilot report without identifying participants. Conversation hours: Acquaint correspondence, September 18, 2026.
