BALTIMORE, Sept. 23, 2026 — Current thinking increasingly treats artificial intelligence (AI) as a scalable substitute for human coaching in digital health. New research published in the INFORMS journal Marketing Science challenges that assumption, finding that access to human coaches in addition to an AI coach leads to significantly greater weight loss than AI coaching alone.
The study is, “Does Access to Human Coaches Lead to More Weight Loss Than With AI Coaches Alone?” from Anuj Kapoor of the University of Missouri, Sridhar Narayanan of Stanford University and Puneet Manchanda of the University of Michigan.
The researchers analyzed roughly 65,000 consumers on HealthifyMe, a major fitness app in India with two paid plans: AI coaching only and AI coaching plus access to human coaches.
Using large-scale activity, food consumption and weight logs spanning 2018 to 2020, the researchers compared adopters of the two plans. They did this after matching them on demographics and pre-subscription behavior to address self-selection.
What they found was a robust advantage for the hybrid plan. Over a typical three-month goal period, users with AI coaches alone lost 1.22 kilograms of weight on average. Users with AI plus human coaches lost about 2.12 kilograms, a difference of about 74%.
Relative to an average starting weight of about 79 kilograms, that amounted to 1.5% of body weight with AI alone versus 2.7% when you add human coaching.
“Our findings challenge the idea that AI coaching is already a full substitute for human support in weight loss,” said Kapoor. “Access to a human coach still produced a meaningful incremental gain, and that gain appears to operate through more ambitious goals and much more consistent tracking.”
The study found important differences in the behaviors associated with that extra weight loss. Users with human coaches set higher weight-loss goals: 16.6 kilograms on average, compared with 15.2 kilograms for AI-only users, a 9.6% increase. They also logged their weight more often, 1.35 times a week versus 1.1 times.
“The largest behavioral gap was in food logging,” said Narayanan. “AI-only users logged food intake an average of 49.7 times a week, while AI-plus-human users logged food 102.0 times a week, an increase of more than 105%.”
The researchers also documented heterogeneity, or the sameness and differences of the demographics of study participants.
“The incremental effect was larger in relative terms for women than for men, for older users than for younger users, and for consumers with lower starting body mass index than for those with higher starting BMI,” added Manchanda. “The patterns suggest that empathy, accountability and comfort with AI are not uniform across user demographic groups or individual users themselves.”
The findings carry certain implications for consumers, digital-health firms and policymakers. Human coaches are costly and difficult to scale, which is why certain weight loss organizations and platforms have invested in AI alternatives. Still, the study suggests that hybrid designs still outperform AI-only coaching.
To help with scalability, while maintaining some human coaching involvement, the key will be to create an understanding of which users will benefit from human coaching, and which behaviors respond best to human intervention, the study authors say.
Read the full study here.
About INFORMS and Marketing Science
INFORMS is the world’s largest association for professionals and students in operations research, AI, analytics, data science and related disciplines, serving as a global authority in advancing cutting-edge practices and fostering an interdisciplinary community of innovation. Marketing Science, a leading journal published by INFORMS, publishes research on quantitative marketing, consumer behavior, pricing, and strategy that informs managerial and policy decisions. INFORMS empowers its community to improve organizational performance and drive data-driven decision-making through its journals, conferences and resources. Learn more at www.informs.org or @informs.
# # #
Media contact
Rebecca Seel
Public Affairs Specialist
(443) 757-3578
Media Contact
Jeff Cohen
Chief Strategy Officer
INFORMS
Catonsville, MD
[email protected]
443-757-3565