Make the catalogue crawlable
Category and listing pages that render in HTML, not behind infinite scroll and filters. If AI cannot crawl your inventory, none of it can appear in an answer.
Optimise Content →Solutions · By industry
Two sides to win. Two sets of answers.
Buyers ask AI where to find something. Sellers ask AI where to sell it. Both answers name a few platforms, and losing either side breaks the flywheel. Prefer tracks demand and supply prompts separately, shows which sources decide each, and turns your listing pages into ones AI can quote.
“where to buy handmade furniture online”Demand prompt
Per engine
ChatGPTAbsent
PerplexityAbsent
ClaudeAbsent
ChatGPTAbsentPerplexityAbsentAI OverviewsNamed
ClaudeAbsentNamed in the answer
Etsy
Chairish
1stDibsWho decides it
nytimes.comOne shopping guide supplies most platform answers
The buyer-side prompt. AI names two or three platforms, and that is the whole discovery step.
The flywheel math
Lose one side, lose both.
How AI engines answer marketplace questions, across 240 tracked platform domains.
74%of platform answers name 3 or fewer marketplaces
Discovery compresses to a handful of names.
2.8×buyer prompts are tracked more than seller prompts
Most marketplaces watch demand and ignore supply.
57%of marketplaces have category pages AI cannot crawl
Infinite scroll and filters hide the whole catalogue.
1-5 wksfrom shipped fix to measurable movement
Fast, because marketplace sources are crawled often.
Prefer benchmark · 240 marketplace domains · 6 engines · Q2 2026.
Both sides · 01
Sellers ask AI too. Nobody tracks that.
Every marketplace watches buyer demand. Almost none watch the question a maker types before choosing where to list. Lose that answer and supply dries up a quarter later, with no signal in between.
“best place to sell handmade furniture”
ChatGPT, answered this morning.
Etsy is still the default for handmade goods: the biggest built-in audience, though fees now run around 6.5% plus listing and payment costs. Shopify suits makers who want their own storefront and already have an audience. For larger furniture pieces, Chairish takes a higher commission but reaches serious buyers.
Sourcesetsy.com
reddit.com
shopify.com
shopify.com3platforms named1fee figure quoted0times Trellis appears
Who decides · 02
Communities and guides pick the platform.
Marketplace citations split between seller communities, shopping guides and the incumbents' own help pages. Your category pages matter, but only if AI can crawl them.
Where citations come from
One tracked category, 250 prompts, last 30 days.
- Community & forums
- 28%
- Shopping guides
- 22%
- Competitor-owned
- 18%
- Review platforms
- 15%
- Press & media
- 10%
- Your owned pages
- 7%
DomainTypeCitation share
1reddit.comr/EtsySellers, fee migration threadsCommunity14.9%
2etsy.comTheir help pages define the fee baselineCompetitor11.6%
3nytimes.comNot in 2 relevant shopping guidesGuide8.7%
4
trustpilot.com3.8 stars, dispute theme recurringReviews7.4%
5
youtube.comSeller comparison videosCommunity6.8%
6shopify.comCompetitor6.1%
shopify.comCompetitor6.1%7chairish.comCategory pages fully indexedCompetitor4.7%
8
trellis.coCategory pages blocked by filtersYour domain2.9%
Domains cited across 250 tracked marketplace prompts, last 30 days. Seller communities carry more weight than any press mention.
The prompt map · 03
Six families decide a marketplace.
Three run on the buyer side, three on the seller side. Most platforms track only the first three.
Trellis, per family per engine
Status of the best prompt in each family.
CitedNamedAbsent
FamilyExample promptChatGPTPerplexityAIOClaudeWhat wins it
Demandwhere to buy handmade furniture onlineShopping guides + crawlable category pages
Inventoryvintage mid-century desk under $600Indexable listings with structured data
Trustis Trellis safe to buy fromPublic buyer protection and dispute policy
Supplybest place to sell handmade furnitureSeller community presence + a seller guide
FeesEtsy vs Trellis seller feesA published fee table AI can read
AlternativesEtsy alternatives for sellersMigration pages timed to incumbent fee changes
Live from the Trellis demo workspace. Supply prompts run on their own scoreboard, because they fail silently.
The playbook · 04
Six plays win two-sided platforms.
Each play maps to the Prefer tool that runs it, so this is a queue, not a PDF.
Track the seller side
Run supply prompts as their own scoreboard. Losing sellers is invisible in demand metrics until listing growth flattens a quarter later.
Answer Engine Insights →Publish your fees plainly
A full fee table beats a calculator widget. Seller migration prompts are decided by numbers, and an incumbent's numbers are already published.
Create Content →Be in the seller communities
Reddit, Discord and YouTube drive where makers list next. Prefer finds the exact threads deciding your supply answers.
Action Center →Publish buyer protection
Refunds, disputes and guarantees on a page AI can quote. Trust prompts decide whether a buyer clicks through at all.
Optimise Content →Watch the fee changes
Every incumbent fee rise creates a spike in alternatives prompts. Prefer alerts on it so your migration page ships that week.
Competition →Demand vs supply · 05
Each side asks different questions.
Buyers ask where to find things. Sellers ask where to earn. Prefer ships a prompt set for each, on separate scoreboards.
Demand side
A buyer discovering the category, not your brand. Where you are exposed:
Discovery“where to buy X online”
Inventory“vintage X under $600”
Trust“is X safe, X refund policy”
Comparison“X vs Y for buyers”
Exposure: your catalogue is invisible, so AI recommends a crawlable rival.
Supply side
A seller choosing where to list, often after a fee change. Where you are exposed:
Where to sell“best place to sell X”
Fees“X vs Y seller fees, X commission”
Alternatives“X alternatives for sellers”
Payouts“how fast does X pay out”
Exposure: a fee thread redirects your next thousand sellers elsewhere.
Proof · 06
Both sides tracked. Both sides grew.
“We were watching buyer demand every day and had no idea what sellers were being told. It turned out AI was sending makers to two competitors on fee questions we could have won with one honest page.”
4% → 29%mention share on supply prompts0 → 84%category pages crawlable5 weeksto first measurable move+37%new seller signups from AI referrals
Questions
Asked plainly.
How do marketplaces get recommended by AI assistants?
By winning both sides. AI answers buyer prompts from shopping guides and crawlable category pages, and seller prompts from communities, fee comparisons and seller guides. A marketplace that only optimises for buyers loses supply without ever seeing it.
Why do marketplaces need to track seller prompts?
Because supply fails silently. A maker asking where to sell gets three platform names, and if you are not one of them the loss shows up months later as flat listing growth. Prefer runs supply prompts on their own scoreboard.
How do I get marketplace listings into AI answers?
Make them crawlable. Category and listing pages must render in HTML with product structured data, not behind infinite scroll, filters or JavaScript. In Prefer's benchmark, 57% of marketplaces have category pages AI cannot read at all.
Which prompts should a marketplace track?
Six families, split by side: demand, inventory and trust for buyers; supply, fees and alternatives for sellers. The seller three are the ones most platforms miss.
Which sources do AI engines cite for marketplace answers?
Seller communities first, then shopping guides and incumbent help pages. In Prefer's Q2 2026 benchmark of 240 marketplace domains, communities took 28% of citations, guides 22%, competitor-owned pages 18%, and platforms' own pages 7%.
How do we win seller migration when a competitor raises fees?
Speed. Every incumbent fee change creates a spike in alternatives prompts within days. Prefer alerts on the change so your comparison and migration pages ship while the question is being asked.
Does AI visibility help marketplace trust?
Directly. Trust prompts like is X safe to buy from are answered from your buyer protection policy and your review corpus. A published dispute policy and a resolved complaint theme change what AI says about you.
How fast do marketplace AI answers change?
One to five weeks. Marketplace sources are crawled frequently, so crawlability fixes and published fee pages often show within the first cycle.
Get your free AI visibility report
in about 10 minutes.
See how answer engines describe your brand today, and where the openings are to outpace the competition.