The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-2664
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T16:58:01.332502Z
status: published
lens: g_space
sector: business
headline: Striking the Right Pitch: The Inverted U-Shaped Effect of AI Anchor Pitch Variability on Consumer Engagement
dek: As artificial intelligence and digital human technologies become increasingly integrated into livestream commerce, AI anchors are becoming important marketing agents. Yet prior research has focused primarily on their visual characteristics, leaving dynamic vocal cues largely unexplored. Drawing on social response theory and perceived authenticity research, this study examines the nonlinear association between AI anchor pitch variability and consumer engagement, together with a proposed psychological pathway and…
gain_title: In 4,322 segments from 330 Douyin AI-anchored livestreams, moderate pitch variability went with the most live comment engagement, and a preregistered experiment found it raised perceived authenticity and engagement intentions versus the extremes.
problem_title: (none)
trace_subject: (none)
gain_reading: In 4,322 segments from 330 Douyin AI-anchored livestreams, moderate pitch variability went with the most live comment engagement, and a preregistered experiment found it raised perceived authenticity and engagement intentions versus the extremes.
gain_evidence: robust evidence of an inverted U-shaped association between pitch variability and real-time danmaku engagement | moderate pitch variability produced greater perceived authenticity and engagement intentions than the average of the two endpoint conditions
problem_reading: (none)
problem_evidence: (none)
quick_read: A two-part study of AI anchors in livestream shopping found that engagement peaks when the anchor's pitch varies by a moderate amount. In real Douyin data covering 4,322 product-presentation segments from 330 livestreams, engagement measured by real-time danmaku comments followed an inverted U-shape against pitch variability.

A preregistered experiment with four AI anchors, four products and three pitch-variability conditions backed this up. Moderate variability produced greater perceived authenticity and engagement intentions than the average of the low and high conditions, and an analysis was consistent with authenticity helping explain the effect.

The work shifts livestream-commerce research from human streamers and visual looks to the voice of AI-mediated sellers. Some parts of the picture remain unsettled: the role of how realistic the anchor looks depended on how pitch and exposure were measured, and the moderate-pitch advantage varied in size across anchor-product pairs.
limitation: The appearance-realism moderation result was conditional and specification-sensitive, and the authors say stimulus-level generalization and causal mediation require further replication.
tag: Evidence-backed gain
key_points: Study 1 analysed 4,322 product-presentation segments nested within 330 AI-anchored livestreams from 85 independent Douyin accounts. | Negative binomial models, with account and session fixed effects and a viewer-minutes exposure measure, showed an inverted U-shaped link between pitch variability and danmaku engagement. | Study 2 was a preregistered mixed design using four AI anchors, four products and three between-participants pitch conditions. | Planned contrasts showed moderate pitch variability beat the average of the two endpoint conditions on perceived authenticity and engagement intentions. | Across the 16 anchor-product combinations the average moderate-pitch advantage was positive, though its size varied by stimulus.
rundown: The field analysis drew on Douyin livestreams in which AI anchors presented products. Researchers modelled real-time danmaku engagement per segment, and checked the inverted U against alternative specifications including account and livestream-session fixed effects, a correlated-random-effects decomposition and viewer-minutes exposure.

In the experiment, participants were assigned to one of three pitch-variability conditions across four AI anchors and four products. A 2-1-1 multilevel analysis gave an indirect-effect pattern consistent with perceived authenticity as the pathway from pitch to engagement intentions.

The authors tested whether appearance realism changes the pitch effect. Results on that boundary condition were conditional and sensitive to how pitch, exposure and within-account effects were specified.
sources:
- peer_reviewed | Journal of Theoretical and Applied Electronic Commerce Research | https://doi.org/10.3390/jtaer21080273 | 2026-08-14
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

cedbdd8cabfb3d428bf7c67538b92419ce448f94e9a3db55811e606dfe9df5ab

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000