Does limiting product choices increase trust in voice assistants?

A single-session online experiment (n = 484) using a custom Alexa shopping app finds that when consumers shop for low-involvement, utilitarian products through a voice assistant (VA), evaluating fewer alternatives — a mark of greater efficiency — is associated with higher decision satisfaction and a greater intention to cede decision autonomy to the VA. The authors (Mari, Mandelli, Algesheimer, in the Journal of Interactive Marketing, 2026-03-04) report this as reversing the usual pattern in other shopping environments: here "the number of alternatives evaluated is negatively associated with consumers' intentions to delegate tasks to the VA and trust its product recommendations, both directly and indirectly, through decision satisfaction." The effect "persists across brand types and price levels."

To isolate the role of the first recommended option, the study varied its brand (private label vs. national) and price (cheap vs. premium) while measuring how many alternatives shoppers went on to evaluate. The abstract frames the result as an efficiency–autonomy trade-off: fewer alternatives evaluated reads as a sign of efficiency, and that efficiency is what drives the downstream effect on delegation and trust, "both directly and indirectly, through decision satisfaction." That the effect "persists across brand types and price levels" is the authors' evidence that the trade-off is a structural feature of VA-mediated shopping rather than an artifact of which product happened to be recommended first — which is also what gives the finding, in their words, "strategic implications for national brands operating in voice commerce environments."

This narrows the delegation question that Could AI assistants become gatekeepers of what people see? poses at the level of policy and societal scale: where that report treats assistants that are "easy to use, highly personalised and personable" as liable to attract trust and dependency in general, this experiment measures one concrete channel for it — a utilitarian shopping task where streamlined choice itself, not persuasive language, is what buys the assistant trust and delegated authority. It also runs differently from Can sycophantic AI advice still push people away from polarized views?, where AI advice moved people away from their initial leanings in a deliberative, information-rich setting across 30 decision environments; here the mechanism runs through narrowing rather than informing, in a single low-involvement, utilitarian product category. Together they suggest that ceding ground to an AI's framing of a decision can arise from opposite proximate causes: more informative input in one case, less choice to evaluate in the other.

The excerpt is a publisher abstract, not the full paper, so it gives no information on effect sizes, how "decision satisfaction," "trust," and "intention to delegate" were operationalized, what the custom Alexa app actually showed participants, or whether the single-session design captures anything beyond immediate post-purchase sentiment. It also does not establish that the pattern generalizes beyond low-involvement, utilitarian purchases or beyond voice interfaces — the claim that this is "contrary to other shopping environments" is the authors' framing, not a comparison measured within this study. The defensible reading is narrow: in this one experimental setting, a voice assistant that shows fewer options was not experienced as a loss of agency but as a service worth rewarding with more autonomy, which is a testable boundary condition for autonomy-ceding claims elsewhere in the literature, not yet evidence that it holds beyond voice commerce.