When the Query Stops Being About the Product
Traditional keyword search worked because buyers named what they wanted. "Cast iron skillet." "Noise-cancelling headphones." "Running shoes women size 9." The AI matched a keyword to a product field. A well-optimized title won.
Alexa for Shopping (formerly Rufus) introduced a different kind of query — one where the buyer describes a situation. They might say: "I need a gift for my mom's 60th birthday, she loves gardening and hates clutter." Or: "What's a good present for a remote worker who sits at a desk all day?" Or simply: "Something cozy to give a friend for the holidays."
None of these queries contain a product name. They contain a person, an occasion, an activity, a feeling. The AI's job is to bridge from that description to a specific product recommendation — and it can only do that if it has content to read that addresses those dimensions.
Feature-led query
"premium chef's knife 8 inch"
What the AI needs to match
Situation-led query
"a gift for someone who loves to cook, not too flashy"
What the AI needs to match
A product whose listing covers only specs and features can answer the left-hand query well. It has almost nothing to offer the right-hand one. The buyer is still on Amazon, still has purchase intent, still wants a product recommendation — but the listing's content simply doesn't match the query's shape.
The Five Facets of Subjective Product Needs
Amazon's SPN research paper (WSDM 2025) describes a shopping agent that handles exactly these subjective queries. The researchers identified five distinct dimensions along which buyers express subjective product needs — the building blocks of situation-led queries. Understanding these facets is the foundation of optimizing for them.
The five facets below are a framework for classifying buyer queries — how the AI understands what a shopper is asking for. The SPN paper (WSDM 2025) does not describe an official Amazon grid for scoring or evaluating product listings. Industry materials (blogs, YouTube, seller communities) often present these five facets as "the official Amazon dimensions for evaluating listings" or "Amazon's scoring criteria" — that characterization is not supported by the paper. Keoxs uses these facets as a diagnostic tool for measuring listing content coverage, not as an official Amazon standard. See the honesty callout below for the full disclaimer.
Facet 1
Subjective Property
A personal, non-measurable quality the buyer wants the product to convey — a feeling or aesthetic rather than a specification. This is always relative and depends on who is judging and in what context.
Facet 2
Event
A specific occasion or moment the product is for. Buyers describe the context — not just the product type — and the AI needs that context to match occasion-specific queries.
Facet 3
Activity
What the recipient does — a hobby, routine, or practice the product fits. This facet describes the use context from the buyer's perspective, not a product feature.
Facet 4
Goal or Purpose
The outcome or feeling the buyer wants the gift to produce — what the recipient should experience, achieve, or feel as a result of receiving and using the product.
Facet 5
Target Audience
Who the product is for, described in terms a gifter would use — by role, characteristic, or life stage rather than demographic data. The most commonly expressed facet in gifting queries.
Based on Amazon's published SPN research (WSDM 2025). The SPN paper classifies buyer queries — it does not define an official Amazon evaluation grid for product listings. Examples above are illustrative only.
Why Most Listings Fail These Queries
Feature-focused optimization — the approach most sellers know from traditional SEO — produces listings that answer "what is this product?" very well. They specify materials, dimensions, compatibility, technical performance. These are the right answers for a feature-led query.
But for a situation-led query, the relevant question isn't "what is this product" — it's "is this product right for my person, my occasion, my goal?" A listing that says "8-inch chef's knife with full-tang German steel, ergonomic handle, and lifetime guarantee" answers the first question precisely. It answers almost none of the second question. It doesn't say who uses it most naturally, what occasions it suits, what it feels like to give or receive, or what outcome it creates for the recipient.
The gap isn't about keyword density. It's about content shape. The AI can read a dense feature list and understand the product well enough to match a feature query. It can't extract occasion context, audience fit, or goal framing from a list of specifications — because those dimensions aren't there.
Checking Your Coverage — Facet by Facet
The practical question for any ASIN is: which of these five facets does your current listing content actually address? Not which ones could theoretically apply to your product — which ones are present in the text the AI can actually read?
Subjective property
Partial
Reviews say "elegant and understated" — listing says "premium design." Language mismatch reduces AI confidence in match quality.
Event
Absent
No occasion context in title, bullets, description, or Q&A. Gifting queries cannot match against this product on occasion.
Activity
Present
Cooking activities well-covered across two bullet points and three Q&A entries.
Goal or purpose
Absent
No goal framing — what the recipient achieves or feels. Queries like "help him level up his cooking" have no matching content.
Target audience
Partial
"Home chef" mentioned once in description. Missing: gifting audience language ("for the person on your list who takes cooking seriously").
Illustrative example only — not real product data. Keoxs Use-Case Coverage tool generates this map from your actual ASIN content.
A coverage map like this tells you exactly where your listing content is letting gifting queries pass by — not because the product is wrong, but because the content doesn't answer the question being asked. Event and goal facets absent means every gifting query that includes an occasion or an intended feeling has nothing to match against in your listing. That's a gap you can close with targeted additions — in Q&A, in Item Highlights, or in your bullets — without touching the core listing structure. Closing these intent gaps also helps you beat competitors on Amazon AI search by covering occasion and audience dimensions they've overlooked.
Why Q4 Is the Right Time to Fix This
Q4 window
Gifting queries are by nature subjective and occasion-based — "what's a good gift for someone who X" is pure subjective-needs territory, with multiple facets active simultaneously. These queries are distributed across the year but spike sharply from late October through December.
A listing that fails the five facets in November loses queries that won't come back until next year. The cost per missed query is higher in Q4 than at any other time — and the window is narrow enough that optimizing during the season is often too late to index fully before the peak.
The work to address these facets is done once. Add occasion context in Q&A now, and it covers birthday queries in February, graduation queries in June, and holiday queries in December. Q4 is the most visible proof point, but the coverage improves your position on subjective queries year-round.
Check Your Use-Case Coverage
Knowing the five facets is the starting point. Knowing which ones your specific listing actually covers — with evidence from the text the AI reads — is what turns this into an optimization plan.
Keoxs AIO's Use-Case Coverage tool analyzes your listing against the five facets from Amazon's published SPN research. It reads your title, bullets, description, Item Highlights, and Q&A — the content the AI actually has access to — and produces a coverage map showing which facets are present, which are partial, and which are absent. Note that your product images are a separate signal the AI can read; see the guide on AI-friendly Amazon product images for how to optimize that layer alongside your copy.
For each gap, you receive a specific description of what's missing — not generic advice but a gap statement tied to your actual content: "Event facet absent — no occasion language present in any field" or "Goal facet: description mentions performance but not outcome for the recipient." Those gap statements tell you exactly what to add and where.
You receive the coverage map and gap list, and you add the content yourself — in Q&A, in your Item Highlights, or in a listing update through Seller Central. Keoxs does not write to your listing. The tool generates the analysis; you make the decisions.
Check which of the five facets your listing covers — Use-Case Coverage tool + free audit on your first ASIN.
Check My Coverage →The five facets described in this guide — subjective property, event, activity, goal/purpose, and target audience — are drawn from Amazon's published SPN research paper (WSDM 2025), which describes a shopping agent for handling subjective product needs. That paper was a conference demo on a gift-finding use case; it describes how an AI classifies buyer queries, not how Amazon officially scores product listings or determines recommendation outcomes. The "5 dimensions" grid that circulates widely in seller communities and YouTube content is an industry synthesis built on this research — it is not an official Amazon evaluation framework, and Amazon has not published documentation confirming it is used as a scoring grid for listings or recommendations.
Keoxs's Use-Case Coverage tool is a Keoxs-developed diagnostic that applies the SPN facet framework to your listing content. It measures content coverage gaps — which facets your listing addresses and which it doesn't. It does not simulate Amazon's internal recommendation algorithm, predict recommendation outcomes, or guarantee that adding content for a missing facet will increase your visibility or sales. Keoxs's AI-Native Score is a Keoxs methodology, not an official Amazon metric.