Truace
Subscribe
68/100

Moderate evidence

Trace: a gain and a problemPublished recordBusinessThe Trace

Full story
More
Export record JSONEmailLinkedInX

When and What Conversational AI Should Ask: Extending the Persuasion Knowledge Model to Explain User Engagement and Purchase Intention

Truvace record
TRV-2026-2563raw data (JSON)
Version
v1Published
First published
Evidence type
Controlled or validated comparison
Sources
1 sourcePeer-reviewed
sha256
335f5fd66c17…check it

Conversational AI is rapidly expanding in e-commerce, where chatbot questions critically shape users’ cognition and engagement. Yet, the effects of question timing and type on persuasion remain underexplored. Building on the persuasion knowledge model, this study examines how question timing (in-conversation vs. follow-up) and type (informational vs. experience-evoking) influence users’ knowledge structures, flow experience, and purchase intention. We developed a retrieval-augmented generation-based recommendati…

59Gain side

How the two sides scoreClose call

59Problem side

The gain

In an online experiment with 335 participants, informational chatbot questions asked during the conversation strengthened shoppers' topic knowledge, which positively predicted flow experience and purchase intention.

The problem

In the same experiment, experience-evoking questions asked during the conversation most strongly activated persuasion knowledge, which negatively predicted flow experience and purchase intention.

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.

What this doesn’t fix

The findings come from a single online experiment with 335 participants using one purpose-built recommendation chatbot, and measure purchase intention rather than actual purchases.

Sources

1 on file
  1. International Journal of Human–Computer Interactiondoi:10.1080/10447318.2026.2728933
    Peer-reviewed2026-09-11

Cite this record

v1

Truvace Impact Record TRV-2026-2563, v1: “When and What Conversational AI Should Ask: Extending the Persuasion Knowledge Model to Explain User Engagement and Purchase Intention.” Truvace, 2026-11-12. /record/TRV-2026-2563 (accessed on the day you copy it). sha256 335f5fd66c17ca8b…

Change history

1 version

Every change made to this record since it was first published, and why. Nothing has changed since it went up.

  1. Publishedv1335f5fd66c17…

    Certified into the record

Check this record hasn’t been altered

Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-2563 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.