Shared research study link
Poshmark UX Study: Resale Shopping Experience
Understand how US consumers perceive Poshmark's shopping experience, social features, and trust mechanisms for buying secondhand fashion
Study Overview
Updated Jan 18, 2026
Research question: How US consumers perceive Poshmark’s shopping experience, social features, and trust mechanisms for buying secondhand fashion. Research group: 18 responses from 6 US consumers (ages 23–45) who buy/sell secondhand on Poshmark/Depop/ThredUp. What they said: Trust is photo-first and forensic-buyers want clear, unedited multi-angle photos, readable brand/care tags and hardware, in-frame measurements, explicit flaw disclosure, and calm, responsive sellers with history; stock/filtered images, evasiveness, and off-platform asks are red flags. Social features generally reduce use: people follow a few trusted sellers and use comments/likes functionally, while Posh Parties/Shows are seen as noisy, FOMO-heavy, and low-trust; a small minority tolerates calm, detail-focused lives and some request higher provenance (receipts/video).
Main insights: Smart List AI helps as a draft for titles/categories but does not build buyer trust and raises risk of factual errors, inflated condition language, search noise, privacy concerns, and murky accountability. Net-net, confidence comes from verifiable proof and human tone, not hype or templated/AI copy; broadcast social layers erode utility for these buyers. Takeaways: Prioritize proof capture/display (timestamp overlay, short video, real zoom, in-frame measurement guidance), add search filters/badges for proof-complete and “AI-assisted, seller-confirmed” listings, and default to a quiet, utility-first experience that foregrounds search/follows/comments while de-emphasizing Parties/Shows.
Main insights: Smart List AI helps as a draft for titles/categories but does not build buyer trust and raises risk of factual errors, inflated condition language, search noise, privacy concerns, and murky accountability. Net-net, confidence comes from verifiable proof and human tone, not hype or templated/AI copy; broadcast social layers erode utility for these buyers. Takeaways: Prioritize proof capture/display (timestamp overlay, short video, real zoom, in-frame measurement guidance), add search filters/badges for proof-complete and “AI-assisted, seller-confirmed” listings, and default to a quiet, utility-first experience that foregrounds search/follows/comments while de-emphasizing Parties/Shows.
Participant Snapshots
6 profiles
Dietrich Manley
Dietrich Manley, 43, is a married renter in Salem, OR. Living mainly on a spouse’s income, exploring re-entry into IT support. Practical and budget-conscious, volunteers, cooks at home, and prefers durable, privacy-minded, mid-tier products and transparent…
Andrew Lucero
Andrew Lucero, 26, is a bilingual city permits technician in Greensboro, NC. Budget-conscious renter who e-bikes to work, meal-preps, and favors durable, offline-friendly gear. Free time: DIY projects, fitness, gaming, local films; community-oriented helper.
Kevin Dominguez
Kevin Dominguez, 33, is a male Evansville homeowner pausing traditional work to renovate his fixer-upper and grow a resale/crafts side business. Frugal, uninsured, Spanish-at-home; budget-minded, repair-first, privacy- and transparency-focused, leaning Repu…
Austin Wal
Austin Wal is a Charlotte-based 27-year-old Black sales pro in footwear retail. He is a homeowner with a motorcycle commute, faith-driven, sneaker-savvy, and community-minded. Pragmatic, tech-forward, and value-focused, he balances hustle with mentoring,…
Alfonso Phillips
Alfonso Phillips is a 23-year-old Jamaican-born construction cleanup crew lead in Boynton Beach. Single, uninsured, cash-first, and faith-driven. Pragmatic, budget-conscious, and mobile-first, saving for a truck and credentials while supporting family back…
Brandi Castellanos
Brandi Castellanos is a bilingual, faith-centered 45-year-old mom in Springfield, MO. Not working; spouse’s high income. Practical, organized, and community-minded. Chooses durable, value-forward products, avoids contracts, and prefers respectful, bilingual…
Dietrich Manley
Dietrich Manley, 43, is a married renter in Salem, OR. Living mainly on a spouse’s income, exploring re-entry into IT support. Practical and budget-conscious, volunteers, cooks at home, and prefers durable, privacy-minded, mid-tier products and transparent…
Andrew Lucero
Andrew Lucero, 26, is a bilingual city permits technician in Greensboro, NC. Budget-conscious renter who e-bikes to work, meal-preps, and favors durable, offline-friendly gear. Free time: DIY projects, fitness, gaming, local films; community-oriented helper.
Kevin Dominguez
Kevin Dominguez, 33, is a male Evansville homeowner pausing traditional work to renovate his fixer-upper and grow a resale/crafts side business. Frugal, uninsured, Spanish-at-home; budget-minded, repair-first, privacy- and transparency-focused, leaning Repu…
Austin Wal
Austin Wal is a Charlotte-based 27-year-old Black sales pro in footwear retail. He is a homeowner with a motorcycle commute, faith-driven, sneaker-savvy, and community-minded. Pragmatic, tech-forward, and value-focused, he balances hustle with mentoring,…
Alfonso Phillips
Alfonso Phillips is a 23-year-old Jamaican-born construction cleanup crew lead in Boynton Beach. Single, uninsured, cash-first, and faith-driven. Pragmatic, budget-conscious, and mobile-first, saving for a truck and credentials while supporting family back…
Brandi Castellanos
Brandi Castellanos is a bilingual, faith-centered 45-year-old mom in Springfield, MO. Not working; spouse’s high income. Practical, organized, and community-minded. Chooses durable, value-forward products, avoids contracts, and prefers respectful, bilingual…
Overview
0 participants
Sex / Gender
Race / Ethnicity
Locale (Top)
Occupations (Top)
Demographic Overview
No agents selected
| Age bucket | Male count | Female count |
|---|
Participant locations
No agents selected
Participant Incomes
US benchmark scaled to group size
| Income bucket | Participants | US households |
|---|
Source: U.S. Census Bureau, 2022 ACS 1-year (Table B19001; >$200k evenly distributed for comparison)
Media Ingestion
Connections appear when personas follow many of the same sources, highlighting overlapping media diets.
Questions and Responses
3 questions
Response Summaries
3 questions
Word Cloud
Question Summary
Preparing summary…
Summary
Themes
| Theme | Count | Example Participant | Example Quote |
|---|
Outliers
| Agent | Snippet | Reason |
|---|
Analyzing correlations…
Generating correlations…
Taking longer than usual
Persona Correlations
Analyzing correlations…
Overview
Across 18 respondents, trust in Poshmark purchases is overwhelmingly photo-first and forensic: natural light, multiple angles, readable tags/hardware, in-frame measurements, and explicit flaw photos are the dominant trust builders. Social features are used narrowly (following trusted sellers, saving items, requesting proof) and broadly rejected when noisy (Posh Parties, live shows). Auto-generated/templated copy or marketing-style photos reduce trust; buyers want human tone and verifiable visual evidence. Occupational background, income, and apparel knowledge shift which secondary cues matter most: apparel/sneaker experience raises provenance and receipt/video demands, younger working professionals favor process-driven triage, and caregivers prefer narrative context and calm seller voice. Friction tolerance (time, climate, patience) modulates willingness to chase additional proof or engage with social features.
Total responses: 18
Key Segments
| Segment | Attributes | Insight | Supporting Agents |
|---|---|---|---|
| Younger working professionals (early–late 20s) |
|
Process-driven, evidence-oriented buyers who maintain low-volume trusted-seller networks and use quick heuristics (checklist/triage) to evaluate listings; they request provenance (videos/receipts) when stakes are higher and will abandon listings that fail simple forensic checks. | Austin Wal, Alfonso Phillips |
| Middle-aged caregivers / family-focused buyers (mid-40s), Spanish speakers |
|
Places more weight on seller narrative and human tone (why item is sold, number of wears) alongside basic forensic signals; prefers calm, neighborly copy and practical proof rather than gamified or hype-driven features. | Brandi Castellanos |
| Lower-income pragmatic buyers (mid-20s to early-40s) |
|
Highly intolerant of noise and marketing polish; demands plain, forensic listings (tape-measure photos, tag close-ups, flaw shots) and avoids social/live features that add friction or pressure-time and effort costs outweigh marginal price savings. | Andrew Lucero, Kevin Dominguez, Dietrich Manley |
| Apparel / sneaker-experienced consumers |
|
Possess a higher provenance bar and sensitivity to small authenticity details; they ask for specialized visual evidence (outsole wear patterns, hardware close-ups) and are more likely to require receipts or short videos to complete purchase confidence. | Austin Wal, Dietrich Manley, Alfonso Phillips |
Shared Mindsets
| Trait | Signal | Agents |
|---|---|---|
| Photo-first forensic verification | Natural/unfiltered photos in daylight, multiple angles, readable tags/hardware, in-frame tape-measure dimensions, and explicit flaw photos are the dominant trust signals across segments. | Kevin Dominguez, Alfonso Phillips, Austin Wal, Andrew Lucero, Dietrich Manley, Brandi Castellanos |
| Distrust of polished/templated marketing content | Stock images, heavy filters, studio-style collages, and AI or templated copy lower perceived authenticity and prompt abandonment. | Kevin Dominguez, Dietrich Manley, Andrew Lucero, Brandi Castellanos, Austin Wal |
| Selective, low-volume social feature use | Following a few trusted sellers, bookmarking for price-watches, and commenting to request proof are common uses; live shows and Posh Parties are widely seen as noisy and pressure-oriented and are ignored by most. | Austin Wal, Alfonso Phillips, Brandi Castellanos, Kevin Dominguez, Dietrich Manley, Andrew Lucero |
| Seller responsiveness and human tone increase trust | Calm, specific DM replies and human-sounding listings improve confidence; defensive or evasive seller behavior terminates interest quickly. | Brandi Castellanos, Austin Wal, Andrew Lucero, Kevin Dominguez |
| Value vs. friction calculus governs conversion | Buyers rapidly balance price against extra effort (requests for more photos, returns friction). Even lower prices won't convert if the listing requires chasing proof or coordination. | Andrew Lucero, Austin Wal, Alfonso Phillips |
Divergences
| Segment | Contrast | Agents |
|---|---|---|
| Younger working professionals vs. Middle-aged caregivers | Younger professionals favor process-driven, evidence-heavy triage (checklists, provenance requests), while caregivers prioritize narrative and calm seller tone alongside basic proof-evidence needs differ in formality and social framing. | Austin Wal, Alfonso Phillips, Brandi Castellanos |
| Lower-income pragmatic buyers vs. Apparel-experienced consumers | Both demand forensic photos, but lower-income buyers reject social features and extra friction at all costs, whereas apparel-experienced buyers are willing to chase provenance (receipts, detailed outsole photos) when item value warrants it. | Andrew Lucero, Kevin Dominguez, Dietrich Manley, Austin Wal |
| General buyer population vs. tolerance for AI/smart-listing copy | Most respondents view AI/templated copy as neutral or negative for trust; sellers with auto-generated descriptions must still supply robust photographic proof to convert-copy convenience does not substitute for visual evidence. | Kevin Dominguez, Dietrich Manley, Andrew Lucero, Austin Wal, Brandi Castellanos, Alfonso Phillips |
| Social-feature utility vs. situational/contextual friction | Some buyers use social features (follow/save/comment) as low-friction tools for curated tracking, but in contexts of low patience or cold weather, any added social noise or required engagement sharply reduces usage and conversion. | Alfonso Phillips, Andrew Lucero, Kevin Dominguez |
Creating recommendations…
Generating recommendations…
Taking longer than usual
Recommendations & Next Steps
Preparing recommendations…
Overview
Buyers use a forensic, photo-first logic: natural light, many angles, readable tags/hardware, in-frame measurements, explicit flaw shots, and a simple proof-of-ownership cue (username/date). Social layers (Parties/Shows) feel like noise; users keep only functional signals (follow a few sellers, comments for verification, likes for price alerts). AI-generated copy is tolerated only as a draft; trust still comes from photos and human specifics. Action: make transparency effortless, de-noise social, and constrain AI to assist + validate.
Quick Wins (next 2–4 weeks)
| # | Action | Why | Owner | Effort | Impact |
|---|---|---|---|---|---|
| 1 | Proof-of-ownership prompt with timestamp overlay | Respondents explicitly trust an in-frame username/date; adding an easy overlay lowers seller effort and raises buyer confidence fast. | Product (Marketplace) + Mobile Eng | Low | High |
| 2 | Enable 10–15s video and true pinch-to-zoom on media | Short video and zoomable close-ups were strong trust boosters (stitching, outsole, fabric). Improves conversion and reduces disputes. | Mobile Eng | Med | High |
| 3 | In-frame measurement helper + buyer comment templates | Tape-in-photo measurements are a top trust cue; templates (‘please add pit-to-pit with tape’) reduce friction for both sides. | Design + CX | Low | High |
| 4 | Trust filters in search (tags, tape, video, timestamp) | Users want to surface ‘boring, honest’ listings; a Show only: tag photo, measurements, video, timestamp filter increases signal. | Search/Ranking + Design | Med | High |
| 5 | Default Quiet Mode for Parties/Shows notifications | Most mute or ignore broadcasts; a calm, utility-first default reduces churn from perceived ‘carnival’ noise. | Product (Engagement) + Notifications | Low | Med |
| 6 | Label ‘AI-assisted’ and require seller confirmation | Buyers distrust templated hype; an AI-assisted, seller-confirmed label plus mandatory human review reduces mislabel risk. | ML/AI + Trust & Safety | Low | Med |
Initiatives (30–90 days)
| # | Initiative | Description | Owner | Timeline | Dependencies |
|---|---|---|---|---|---|
| 1 | Guided Proof Capture & Validation | A seller flow that nudges and verifies core trust assets:
|
Product (Marketplace) + Computer Vision/ML | 8–10 weeks to MVP | Mobile Eng, Design, CV/OCR, Trust & Safety, Data Science |
| 2 | Trust Filters + Forensic Ranking | Introduce filters and boost signals for listings containing tag photos, in-frame measurements, flaw shots, video, timestamp. Rerank search to favor proof-complete listings; expose a ‘Trust Boost’ badge. | Search/Ranking + Data Science | 6–8 weeks | Design, Analytics, Marketplace Product |
| 3 | Social Noise Redesign (Quiet Mode by default) | De-emphasize Parties/Shows in default UI, consolidate notifications, and provide a single Quiet Mode toggle. Keep functional social (follows, comments, price alerts) prominent. | Product (Engagement) + Design | 1 quarter for rollout + A/B | Notifications Platform, Growth/CRM, User Research |
| 4 | Forensic Live Mode for Posh Shows | A live template optimized for trust: split-view ‘measure cam’, required tag/outsole demo, on-screen measurements, slower pace, auto-clip replays per item. | Live Commerce | 8–12 week pilot with 50 hosts | Live Ops, Mobile Eng (streaming), Seller Education, Trust & Safety |
| 5 | Smart List AI 2.0: Assist + Validate | Shift AI from ‘write’ to ‘verify’:
|
ML/AI | 1 quarter | CV/OCR, Legal/Policy, Design, Trust & Safety |
| 6 | Seller Education + Incentives (Trust Boost) | Micro-tutorials and nudges showing that ‘proof’ lifts sales; award a Trust Boost badge and small fee credits for proof-complete listings; dashboards with conversion deltas. | Seller Ops + Marketing | 6 weeks to launch toolkit | Content/CRM, Design, Analytics, Finance |
KPIs to Track
| # | KPI | Definition | Target | Frequency |
|---|---|---|---|---|
| 1 | Proof completeness rate | Share of new listings with tag close-up, in-frame measurements, flaw photos, and timestamp | 60%+ within 90 days of rollout | Weekly |
| 2 | Conversion lift from proof-complete listings | Relative purchase conversion of proof-complete vs. proof-light listings (matched by category/price) | +20% lift within 60 days | Weekly |
| 3 | Not-as-described return rate | Percent of orders returned due to misdescription/misattribution | -25% in 6 months | Monthly |
| 4 | Quiet Mode adoption | Percent of active buyers with Parties/Shows notifications off or in Quiet Mode | 70% adoption without engagement decline on core search | Monthly |
| 5 | AI discrepancy rate | Percent of AI-assisted listings where OCR/tag data conflicts with AI-filled attributes at publish | <3% after AI 2.0 | Weekly |
| 6 | Seller median first-response time | Median hours from buyer comment to seller reply on trust-critical asks (measurements, tag shots) | <12 hours | Weekly |
Risks & Mitigations
| # | Risk | Mitigation | Owner |
|---|---|---|---|
| 1 | Added listing steps increase seller friction and reduce supply | Make proof steps optional but rewarded via ranking and badges; streamline with overlays/OCR; A/B test thresholds before enforcing. | Product (Marketplace) |
| 2 | Live Show constraints reduce short-term GMV | Pilot ‘Forensic Live’ with trusted hosts; offer fee credits and better discovery; measure dwell, trust, and conversion before wider rollout. | Live Commerce |
| 3 | AI mislabels create liability and distrust | Force human confirmation; show AI confidence; block hype terms; cross-check with OCR; add an ‘AI-assisted’ label. | ML/AI + Trust & Safety |
| 4 | Privacy concerns around receipts/OCR/timestamps | Mask PII by default, strip EXIF, provide clear consent and opt-outs; store derived text, not raw docs where possible. | Legal/Policy + Security |
| 5 | Bad actors spoof proof assets (fake timestamp, staged measurements) | Heuristics on inconsistencies (EXIF/time drift, repeated templates), random spot checks, buyer-report fast path, penalties for abuse. | Trust & Safety |
| 6 | Engineering complexity slows delivery | Stage delivery: UI nudges first, then OCR checks; reuse existing media components; keep pilots small and metrics-driven. | Engineering |
Timeline
- Weeks 0–4: Quick wins live (timestamp overlay, comment templates, Quiet Mode default, media zoom/video)
- Weeks 5–10: Guided Proof Capture MVP + Trust Filters/Ranking A/B
- Weeks 8–12: Seller education + Trust Boost incentives
- Quarter 2: Smart List AI 2.0 (assist+validate) + Forensic Live pilot; iterate social de-noise
- Quarter 3: Scale successful pilots; consider graduated requirements in high-risk categories (sneakers, luxury)
Research Study Narrative
Objective and context
Claude commissioned a qualitative study to understand how US consumers perceive Poshmark’s shopping experience, social features, and trust mechanisms when buying secondhand fashion. Across our sample (18 respondents across three prompts), buyers converge on a “forensic, photo-first” logic for trust, use social features narrowly and functionally, and treat AI-generated listing copy as draft-only rather than a credibility signal.
What builds (and breaks) trust in listings
- Primary trust signals (consistently cited): clear, unedited photos in natural light; multiple angles (front/back/interior/soles); readable brand and care tags plus hardware close-ups; in-frame tape measurements; explicit, “warts-and-all” flaw photos; a simple proof-of-ownership cue (handwritten username/date in-frame). Buyers also value short videos and true zoom for stitch/fabric inspection, and receipts/provenance when available.
- Seller behavior cues: calm, specific descriptions; consistent photo setup; responsive, straightforward replies; established closet history increase confidence. Evasive replies or off-platform requests are red flags.
- Red flags (inverse of above): stock or studio images, heavy filters/collages, cropped/blurry tags, hype-y generic copy (“vintage,” “elevated” without evidence), prices/claims that feel too-good-to-be-true.
Social features: narrow utility, broad aversion to noise
- What people use: follow a small set of trusted sellers (often ~5–6), comment to request proof (extra photos, measurements, timestamp), and “heart” items as functional bookmarks/price-drop alerts.
- What people ignore/mute: Posh Parties and Posh Shows are widely perceived as noisy, hype-driven, and low-trust; many mute notifications or avoid broadcast features entirely. A minority will watch calm, detail-focused lives hosted by trusted sellers.
- Impact: Most respondents said social features make them less likely to use the app when they add clutter or pressure (“I want a clean marketplace, not a carnival”).
Smart List AI: convenience, not credibility
- Seller stance: acceptable as a draft for basics (title, category, color), but sellers plan to rewrite details and add proof (tape-on-garment measurements, tag photos, flaws).
- Buyer stance: trust remains anchored to photos/tags/measurements, not AI text. Concerns include factual errors (fabric/size/brand), condition inflation, generic boilerplate, and search pollution from low-effort listings. Some request an “AI-assisted” badge and filters; others want AI to verify against tag OCR rather than generate hype.
Persona nuances
- Younger working professionals (23–27): process-driven triage (checklists, provenance requests), small trusted-seller networks; will abandon listings that fail simple forensic checks.
- Middle-aged caregivers (bilingual): value human, calm tone and practical proof (why selling, number of wears) alongside core photo evidence.
- Lower-income pragmatists: intolerant of noise; demand plain, proof-rich listings; avoid lives/Parties that add friction.
- Apparel/sneaker-experienced: higher provenance bar (outsole/hardware close-ups, receipts, videos) and sensitivity to authenticity details.
Recommendations
- Proof-of-ownership overlay (username/date) and true pinch-to-zoom + 10–15s video to strengthen visual evidence.
- In-frame measurement helper and buyer comment templates (“Please add pit-to-pit with tape”) to reduce back-and-forth.
- Trust filters in search (show only listings with tag photo, measurements, flaw shots, video, timestamp) and a Trust Boost badge/rerank for proof-complete listings.
- Quiet Mode default to de-emphasize Parties/Shows while keeping follows, comments, and price alerts prominent.
- Smart List AI 2.0: Assist + Validate (draft with confidence labels, tag OCR cross-checks, block hype terms, require human confirmation; optional “AI-assisted” label and filters).
- Forensic Live Mode for shows (required tag/outsole demo, on-screen measurements, slower pace, auto-clip replays).
Risks and guardrails
- Added listing friction: keep proof steps optional but rewarded via ranking/badges; streamline with overlays/OCR.
- AI mislabels/privacy: force human confirmation, show AI confidence, mask PII on receipts/timestamps, and add an “AI-assisted” label.
- Spoofed proof: apply heuristics (EXIF/time drift, template reuse), spot checks, and fast-path enforcement.
Next steps and measurement
- Weeks 0–4: Ship timestamp overlay, comment templates, Quiet Mode default, zoom/video.
- Weeks 5–10: Launch Guided Proof Capture MVP and Trust Filters; A/B “Trust Boost.”
- Quarter 2: Pilot Forensic Live and Smart List AI 2.0 (assist+validate); expand social de-noise.
- Quarter 3: Scale successful pilots; consider graduated proof requirements for high-risk categories.
- KPIs: Proof completeness rate (target 60%+ in 90 days); Conversion lift of proof-complete vs. proof-light (+20% in 60 days); Not-as-described return rate (−25% in 6 months); Quiet Mode adoption (70% without core search decline); AI discrepancy rate (<3% after AI 2.0).
Recommended Follow-up Questions
Updated Jan 18, 2026
-
Rank the following discovery features by how much they would help you quickly find trustworthy items on Poshmark (most helpful to least helpful): Exclude listings with stock/heavily filtered photos; Show only listings with in‑frame measurements; Show only listings with a short video; Filter by verified timestamp/watermark on photos; Filter by seller rating 4.8+ and low cancellation rate; Exclude AI‑generated descriptions; Hide sellers you’ve muted/blocked; Filter by standardized, verified condit...rank Prioritize discovery/filter investments that best reduce noise and surface credible listings.
-
For each buyer‑protection element, how much would it increase your willingness to purchase on Poshmark? Rate 1 (not at all) to 5 (a lot): Payment held in escrow until you confirm item condition; Easy 'not as described' returns with prepaid label; Platform authenticity check for high‑value items; Clear dispute resolution timeline with status updates; Default shipping insurance for full order value; Guaranteed delivery window or automatic refund; Seller penalty/visibility reduction for repeated mi...matrix Identify which protections most shift purchase intent to guide policy and UI prioritization.
-
Above what item price (USD) would you require third‑party authentication before buying on Poshmark? Enter a number; enter 0 if you would never require it.numeric Set authentication eligibility thresholds and size demand for verification services.
-
Which seller reputation signals most increase your confidence to buy? You will see sets of items-select the MOST and LEAST impactful in each set. Signals: Government ID‑verified seller badge; Number of completed sales; 12‑month cancellation rate; Average response time to buyer questions; Percentage of repeat buyers; Dispute/return rate for 'not as described'; Years active on Poshmark; Buyer reviews with photos; Verified address/payment on file; On‑platform video walkthroughs on recent sales.maxdiff Decide which reputation metrics to surface prominently in profiles, search, and badges.
-
Which changes would meaningfully increase your likelihood to purchase from Posh Shows (live streams)? Select all that apply: Mandatory unedited close‑ups of tags/hardware/flaws; On‑screen measurements taken live; Slower pace with time for angle requests; Independent moderator verifying claims in real time; Ability to pause/rewind and watch replay; Display buyer protection/return terms on‑screen; Host identity and seller history verified/badged; Transparent price/comps (retail and recent solds);...multi select Define MVP requirements to make live shopping usable and trustworthy for skeptics.
-
Please indicate your agreement with these potential listing policies on Poshmark (1 strongly disagree to 5 strongly agree): Require at least six unedited photos per listing; Require a 10–15 second video per listing; Require in‑frame measurements for apparel/shoes; Block stock or heavily filtered images; In‑app camera adds timestamp/username watermark; Flag listings when AI‑detected attributes conflict with tags/photos.likert Gauge buyer tolerance for stricter media requirements to balance trust gains vs seller friction.
Randomize item order within rank/maxdiff/matrix to reduce order bias. Consider oversampling high‑value buyers to validate authentication threshold.
Study Overview
Updated Jan 18, 2026
Research question: How US consumers perceive Poshmark’s shopping experience, social features, and trust mechanisms for buying secondhand fashion. Research group: 18 responses from 6 US consumers (ages 23–45) who buy/sell secondhand on Poshmark/Depop/ThredUp. What they said: Trust is photo-first and forensic-buyers want clear, unedited multi-angle photos, readable brand/care tags and hardware, in-frame measurements, explicit flaw disclosure, and calm, responsive sellers with history; stock/filtered images, evasiveness, and off-platform asks are red flags. Social features generally reduce use: people follow a few trusted sellers and use comments/likes functionally, while Posh Parties/Shows are seen as noisy, FOMO-heavy, and low-trust; a small minority tolerates calm, detail-focused lives and some request higher provenance (receipts/video).
Main insights: Smart List AI helps as a draft for titles/categories but does not build buyer trust and raises risk of factual errors, inflated condition language, search noise, privacy concerns, and murky accountability. Net-net, confidence comes from verifiable proof and human tone, not hype or templated/AI copy; broadcast social layers erode utility for these buyers. Takeaways: Prioritize proof capture/display (timestamp overlay, short video, real zoom, in-frame measurement guidance), add search filters/badges for proof-complete and “AI-assisted, seller-confirmed” listings, and default to a quiet, utility-first experience that foregrounds search/follows/comments while de-emphasizing Parties/Shows.
Main insights: Smart List AI helps as a draft for titles/categories but does not build buyer trust and raises risk of factual errors, inflated condition language, search noise, privacy concerns, and murky accountability. Net-net, confidence comes from verifiable proof and human tone, not hype or templated/AI copy; broadcast social layers erode utility for these buyers. Takeaways: Prioritize proof capture/display (timestamp overlay, short video, real zoom, in-frame measurement guidance), add search filters/badges for proof-complete and “AI-assisted, seller-confirmed” listings, and default to a quiet, utility-first experience that foregrounds search/follows/comments while de-emphasizing Parties/Shows.
Awaiting responses…
Usually takes — seconds
Taking longer than usual
Question
0 responses
Waiting…
| Name | Response | Info |
|---|