aceModerate evidence
More
Chatbots won back shoppers after small errors; humans did better on complex tasks, studies find
Study title: Self-driven or human-assisted? The impact of chatbot service recovery on consumer purchase intentions
- Evidence type
- Controlled or validated comparison
- Sources
- 1 sourcePeer-reviewed
- Published
- 2 min read
- Truvace record
- TRV-2026-2588cite or check it
Study title: Self-driven or human-assisted? The impact of chatbot service recovery on consumer purchase intentions
Four experiments tested how consumers respond when a customer-service chatbot makes a minor mistake and either fixes it itself or hands the repair to a human. Chatbot self-recovery using psychological strategies significantly raised purchase intentions, while human-mediated recovery worked better when paired with tangible strategies.
Read the study abstract
Purpose This study aims to investigate the interaction effect between service recovery providers (AI chatbot vs human) and recovery strategies (psychological vs tangible) on consumers’ purchase intentions in the context of chatbot minor service failures. The research further seeks to provide companies with insights to enhance chatbot-based service recovery. Design/methodology/approach Four experiments were conducted to test the proposed hypotheses. Findings The results indicate that when chatbots encounter minor…
57Gain side
How the two sides scoreClose call
55Problem side
The gain
In four experiments, a chatbot that recovered from its own minor failure with psychological strategies significantly raised consumers' purchase intentions.
The problem
In the same experiments, when the task was complex, human-mediated recovery worked better than chatbot self-recovery whichever strategy was used.
A fixed formula sets both scores from the evidence and its sources. Reader votes don’t move them.
What the scores mean
G is the gain side, out of 100. It goes up when the gain is specific, when it measurably helped people and when the sources behind it are strong.
P is the problem side, out of 100. It goes up when the problem is specific, when it measurably hurt people and when the sources behind it are strong.
In brief
The results matter for companies deciding when to let a chatbot repair its own errors and when to bring in staff. Consumers' judgments of the chatbot's competence shaped their reactions more strongly than judgments of its warmth.
The pattern shifted with task complexity and chatbot intelligence. On complex tasks, human recovery was more effective whatever strategy was used, while a highly intelligent chatbot's self-recovery worked well with either strategy.
Main points
- Researchers ran four experiments comparing AI chatbot recovery with human recovery after minor chatbot service failures.
- Chatbot self-recovery with psychological strategies significantly raised consumer purchase intentions.
- Human-mediated recovery was more effective when paired with tangible strategies.
- Perceived competence mediated consumer responses more strongly than perceived warmth.
- For high task complexity, human-mediated recovery was more effective with either a psychological or a tangible strategy.
- When a highly intelligent chatbot failed in a minor way, its self-recovery was highly effective with either strategy type.
The full story
The study crossed two factors: who carries out the recovery (the AI chatbot itself or a human) and which strategy is used (psychological or tangible). It measured how each combination affected consumers' purchase intentions after a minor chatbot service failure.
Results split by pairing. The chatbot repairing its own mistake did best with psychological strategies, while human intervention did best with tangible ones. Perceived competence played a stronger mediating role than warmth in explaining why consumers responded as they did.
Two conditions changed the picture. High task complexity favoured human recovery under both strategies, while a highly intelligent chatbot could recover effectively on its own with either strategy. The authors say the findings can help companies improve chatbot-based service recovery.
What this doesn’t fix
The findings cover minor chatbot service failures tested in experiments, so they may not extend to serious failures or real-world customer behaviour.
Sources
1 on file- Journal of Services Marketingdoi:10.1108/jsm-09-2024-0488
Talk it over
This record found a gain and a problem in the same place. Which matters more to you, and who might answer differently?
Chatbots won back shoppers after small errors; humans did better on complex tasks, studies findhttps://truvace.com/item/self-driven-or-human-assisted-the-impact-of-chatbot-service-recovery-on-consumer
This record found a gain and a problem in the same place. Which matters more to you, and who might answer differently?
The gain In four experiments, a chatbot that recovered from its own minor failure with psychological strategies significantly raised consumers' purchase intentions.
The problem In the same experiments, when the task was complex, human-mediated recovery worked better than chatbot self-recovery whichever strategy was used.
Answer in the threadHow do we know? What can this kind of evidence show, and what can’t it?
The evidence This rests on a controlled comparison, from Journal of Services Marketing.
Answer in the threadWhat does it leave open? What else would you want to know before acting on it?
The record’s own caveat The findings cover minor chatbot service failures tested in experiments, so they may not extend to serious failures or real-world customer behaviour.
Answer in the threadHave you seen AI used this way where you study, work or live? Did it go the same way?
Answer in the thread
The questions are the same on every record. The details under them are taken from this record as written.
What readers say