How do Bing Shopping personalized results pick products?
Quick answer: Bing is testing a Shopping tab that asks shoppers three questions and then tailors results to their answers. Microsoft hasn't published the questions or how it ranks the answers. The products most likely to show up are the ones whose feed attributes and product pages clearly state what they're for, what they cost and details like size, material or style. A complete feed and plain, direct product copy are your best bet for free visibility.

Put your organic marketing on autopilot
artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.
What exactly is Bing testing in its Shopping tab?
Bing is testing a version of its Shopping tab that asks shoppers three questions before showing any products, then tailors the results to their answers. The test was spotted in early November 2026, and it looks a lot like the question-based shopping tests that have recently appeared in Google AI Mode and ChatGPT. When a shopper switches to the Shopping tab, the questions appear. Once they submit their answers, the product grid reflects what they said, not just the words they typed. Microsoft hasn't announced the feature, published the questions or explained how answers are weighted, and Bing runs plenty of interface experiments that never ship. Treat this as a sign of where things are heading, not a finished product. And the direction is clear: search engines want to shrink a shopping query to a short list before anyone starts scrolling.
For store owners, the interesting part isn't the interface. It's what gets filtered out. A query like “running shoes” used to return a broad grid ranked by relevance, price and merchant signals. Add three answers, say trail running, under $120 and wide fit, and the pool shrinks sharply. Any product whose data doesn't clearly confirm those details risks dropping out, however good it actually is. Bing's free product listings draw on the same Microsoft Merchant Center feed that powers paid product ads and product cards in Copilot. The data you submit there is your answer sheet. If a field is blank, vague or contradicts your product page, the system has nothing to match against. A personalized filter can only work with the attributes it's able to read.
Here's our take: the same shift is happening across AI-assisted shopping, and it rewards boring, thorough product data more than clever copywriting. Google AI Mode asks follow-up questions and builds comparison lists, and Google is even testing a website button that sends AI Mode shoppers to the retailer's own page. ChatGPT's shopping answers ask what the shopper needs before recommending anything. Each of these flows turns a vague query into a set of specific requirements, then looks for products that clearly meet them. Stores that spent years packing titles with high-volume keywords will find that a stuffed title doesn't help when the engine is checking whether a product is waterproof, what it costs and who it's for. The upside is that you can fix this without a bigger ad budget. Mostly it means stating clearly, in feed fields and in plain sentences, what each product is and who it suits.
Which product details are the questions likely to filter on?
Bing hasn't disclosed its questions, but shopping question flows almost always narrow things down by three factors: what the product is for, how much the shopper wants to spend, and a personal preference such as size, color, material, brand or style. That's how a good salesperson sizes up a customer, and it's how clarifying questions in other AI shopping tools usually work. Use case is the hardest of the three for stores, because it's rarely a feed field. Phrases like “for a home office,” “for a toddler” or “for sensitive skin” live in descriptions, reviews and buying guides, if they appear anywhere at all. Budget is the easiest, since price is a required attribute. Preferences sit in between: many have their own attributes, but stores leave them empty more often than you'd expect.
This is where product attributes do real SEO work. Each answer a shopper gives probably maps to data the engine can check. Budget maps to the price and sale price fields. Preference maps to color, size, material, gender, age group and pattern. Use case maps to product type, category and description. Product identifiers matter too. A correct GTIN lets the engine connect your item to what's already known about it elsewhere, such as reviews, specs and comparisons, which makes it more confident about whether you're a match. A missing identifier won't automatically disqualify you, but it leaves the system guessing, and in a filtered list a guess usually loses to a confirmed fact. If two stores sell the same kettle and only one says it's 1.7 liters, stainless steel and cordless, that's the one most likely to survive a three-question filter.
A practical way to prepare is to write down, for each of your top categories, the three questions you'd ask a customer in a physical shop. A bike store might ask about terrain, budget and frame size. A skincare brand might ask about skin type, main concern and price. A furniture store might ask about the room, the dimensions and the style. Then check whether each answer is written somewhere a machine can read it: in a feed attribute, in structured data on the page, or in a clear sentence of copy. If an answer only shows up in a photo, lives in your head or hides behind a vague adjective like “versatile,” you've found a gap. Those gaps are where tailored results will pass you over, so they're the first things to fix.
| Likely question | Example answer | Where to state it in the feed | Where to state it on the page |
|---|---|---|---|
| What will you use it for? | Trail running | product_type, category, description | Opening lines of copy, FAQ, buying guides |
| What's your budget? | Under $120 | price, sale_price, availability | Visible price, Product structured data offers |
| Any preferences? | Wide fit, waterproof | size, color, material, gender, age_group | Spec table, Product structured data |
| (Matching the exact product) | n/a | brand, gtin, mpn | Product structured data identifiers |
How do you set up your Bing Merchant Center feed to answer those questions?
Start by getting every product into Microsoft Merchant Center with a complete, accurate feed, then fill in the optional attributes that describe use and preference. To appear in free listings on the Bing Shopping tab, you need an approved Merchant Center store, a verified website and a valid feed. You don't need to run paid campaigns. The core fields are the familiar ones: ID, title, description, link, image link, price, availability and condition, plus brand and a GTIN or MPN for most products. Microsoft's feed format closely follows Google's, so if you already maintain a Google Merchant Center feed, you can usually submit the same file or a lightly adjusted version. That's an easy win many stores still skip because Bing feels like an afterthought.
The required fields get you listed. The optional ones help you get through the filter. Fill in color, size, material, pattern, gender and age group wherever they apply, and keep values consistent: “Navy” in one row and “dark blue” in another just splits your catalog. Build product_type into a real hierarchy that reflects how customers shop, such as “Shoes > Running > Trail,” rather than your internal warehouse codes. Write titles that start with the brand, the product type and the attribute shoppers filter on most, followed by one key differentiator. Open descriptions with who the product is for and what it does, in plain language. Keep prices and stock status in sync with your site. A mismatch can get items disapproved, and a disapproved product won't show up in any results, filtered or not.
Keep updating the feed instead of uploading it once and forgetting it. Schedule automatic fetches at least daily, and check Merchant Center for warnings after every catalog change. Merchant Center also reports impressions and clicks for free listings, so review them to see which categories are gaining or losing visibility. If you can only fix part of the catalog this month, start with your best sellers and highest-margin lines, because that's where a spot in a tailored result is worth the most. Supplemental feeds, where your setup supports them, let you add or override attributes without rebuilding the main file. For seasonal lines the same logic applies, and our pre-holiday checklist for getting into AI Mode gift ideas covers the timing. And resist the urge to cram every possible use case into titles. A title written for a keyword tool reads worse to shoppers and machines than one that simply names the product precisely.

Put your organic marketing on autopilot
artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.
How should product pages and buying guides answer shopper questions?
Your product pages should answer the three likely questions in visible text near the top, and your buying guides should match use cases to specific products. The feed tells Bing what a product is. The page and the content around it tell Bing, Google and AI assistants who it's for and why. That second layer matters more as search engines combine product data with the rest of the web. Bing Shopping already produces AI-generated buying guides and review summaries, and those features have to pull from somewhere. Say your page states plainly that a backpack fits a 16-inch laptop, handles light rain and suits daily commuting. That product has a much better chance of matching “commuting, under $80, waterproof” than one whose page only says it's “built for adventure.”
On each product page, add Product structured data with offers, price, availability, brand, identifiers and ratings where you genuinely have them, and make sure it matches the feed exactly. Below the main image and price, include a short spec table that spells out size, materials and dimensions in words rather than icons. Then add two or three sentences that answer the use-case question directly: who it's for, what it's best at, and what it isn't suitable for. That last part feels counterintuitive, but honest limits improve matching. A product that clearly says “not designed for trail use” is less likely to be shown to the wrong shopper and then returned. A short FAQ on fit, care and compatibility covers the remaining preference questions in the shopper's own words.
Category-level buying guides do most of the work for use case, which feeds handle worst. Title a guide after a real search, such as “best running shoes for wide feet” or “which espresso machine for a small kitchen.” Open with a direct answer and link to specific products. That gives every engine the context it needs to connect what shoppers want with what you sell. These guides can also be cited in AI answers in their own right, which is a separate source of free traffic. We've covered what makes a page likely to be chosen as an AI Mode source in more detail. Before you publish, a quick GEO score check on your main category and product pages will show where your copy buries the answer instead of leading with it.
When is your catalog too thin for this to matter?
If you sell a small range in a narrow niche, or mostly items shoppers search for by exact name, a three-question filter will change little for you. Personalization matters when there's a crowded pool to narrow down. A store selling one patented gadget, a handful of made-to-order pieces or replacement parts identified by model number doesn't win or lose on use-case filtering, because its shoppers already know what they want. Likewise, if Bing sends you a tiny share of your traffic today, rebuilding your feed around a test that may never launch isn't the best use of a busy week. The basics still pay off, but your priority should be ranking for the exact product names and part numbers your buyers actually search for. Clean titles and correct identifiers matter more to you than detailed preference attributes.
The stores that should act now are those in crowded categories where shoppers compare on specs and details: apparel, footwear, furniture, beauty, electronics accessories, outdoor gear and home goods. There, dozens of similar products compete for every query, and the tie-breaker is increasingly which one the engine can confidently match to a shopper's stated needs. The good news is that the same work pays off three times. A complete feed can power Bing free listings and, with little adjustment, Google's free listings too. Clear, detailed product pages rank better in regular organic search. And buying guides that lead with the answer get quoted by AI assistants. None of it switches off when a budget runs out, which is the real difference from paid ads. If you're weighing that trade-off, it's worth looking at how SEO compares with Google Ads on cost over a year.
Our view is that question-led shopping is here to stay, on Bing or elsewhere, because it simply works better for shoppers than a wall of near-identical product tiles. The stores that benefit will be the ones whose product data and content already answer real questions. That's the work Artiql automates. It produces product and buying-guide content that leads with the answer, built around the attributes and needs your customers search for, and publishes it consistently so your organic visibility builds over time across Google, Bing and AI assistants. Pricing is tailored to each store after a short discovery call, because a boutique with a few dozen products and a catalog of thousands need very different plans. If you'd like to see where your products would fall out of a three-question filter today, book a short demo and we'll go through your catalog together.
Frequently asked questions
Has Bing officially launched question-based Shopping results?
No. As of November 2026 it's a test that some users are seeing, not an announced feature. Microsoft hasn't published the questions it asks or how the answers affect ranking, and Bing regularly trials Shopping layouts that never roll out widely. Still, Google AI Mode and ChatGPT are moving toward similar clarifying questions, so improving your product data is a safe bet whichever version eventually ships.
Do I need to pay for Microsoft Advertising to appear in Bing Shopping?
No. Free product listings can show eligible products in the Bing Shopping tab without any ad spend. You need an approved Microsoft Merchant Center store, a verified website and a valid product feed that meets Microsoft's policies. Paid product ads run on the same feed, but they're optional. Merchant Center reports impressions and clicks for free listings, so you can track what they deliver.
Can I reuse my Google Merchant Center feed for Bing?
Usually, yes. Microsoft's feed format closely follows Google's, so most stores can import or submit the same file, then fix any warnings Merchant Center flags. Check that attribute values are consistent, that prices and availability match your site, and that identifiers like GTINs are correct. Optional attributes such as color, size and material matter more now that results may be filtered by shopper preferences.
How do I optimize products for AI shopping if I can't see the questions?
Focus on the questions any good salesperson would ask: what the product is for, what it costs and which preferences matter most, such as size, material, color or style. State each answer in feed attributes, in Product structured data and in plain sentences near the top of the page. Then cover use cases with buying guides that link to specific products. That approach works for Bing, Google AI Mode and ChatGPT alike.

Put your organic marketing on autopilot
artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.