When the shopping assistant chooses
Purchase intention, purchase decisions and consumer choice appear on 178 recent papers, mostly in management and economics, but only 14 of them carry a generative-AI, conversational-AI or recommendation keyword. Shoppers are starting to let AI assistants search, compare and buy. Who gets recommended, and what happens to brands and prices when a model makes the choice?
What prompted this#
Since October 2025, 178 papers carry a purchase-intention, purchase-decision, consumer-choice or online-shopping keyword (88 in management journals, 48 in economics, 24 in operations research): "purchase intention" on 48, "purchase decision" on 38, "consumer choice" on 37, "online shopping" on 19. Only 14 of them also carry a generative-AI, ChatGPT, language-model, AI-assistant, conversational or recommendation keyword, while "conversational AI" appears on 20 papers in the archive.
Questions#
- When an assistant picks products, how concentrated are its choices across brands compared with human shoppers, and do the choices favour the assistant's owner or advertisers?
- Do consumers who delegate to assistants pay less, buy more, or buy different things?
- How do sellers optimize for assistants rather than for human eyes, and does that change product descriptions and prices?
Data that could help#
Logs of assistant-mediated shopping sessions, experiments comparing assistant and human choices for the same shopping tasks, and retailer price and description data over time.
Next step#
Ask several shopping assistants to buy the same 100 everyday products and compare brand concentration and prices with what human panels chose.