Conversion deep-dive · Trade Me + Website

Fix the channel that's selling. Then fix the one that isn't.

Trade Me is almost certainly where the money is made today — so its listings are the immediate CRO win. The website is search-invisible and likely converting near zero — so it's the bigger multiple, once it's measured and rebuilt. Both come down to one thing for refurbished goods: de-risking the buyer.
Trade Me = today's revenue Website = the 10×+ once measured Trust is the conversion lever Success fee just dropped

1 Two channels, and the data blind spot

Trade Me
Where sales actually happen today (the site is search-invisible)
<0.1%?
Estimated website conversion — vs the ~1.5–2.5% ecommerce norm
Step 0
Measure before you optimise — no analytics installed today
You can't optimise what you can't see. Before any website work: install GA4 + Microsoft Clarity (free heatmaps, session recordings, funnel drop-off) behind Stape sGTM for clean server-side data. This isn't a side quest — the same event data powers the cart-abandonment and dead-stock email flows, so it unlocks three workstreams at once. Until it's in, "below 0.1%" is an educated guess, not a number.

Part A — Trade Me CRO (the channel selling now)

2 Rank in Best Match, convert with trust

Trade Me doesn't publish its Best Match weightings, but one factor is confirmed and the rest are well-founded. Ranked by impact:

🔍 Live store audit — stores/Tydi (super-scraper, Jun 2026)

2007
Member since — a 19-year account
99.16%
Positive feedback · 10,586 reviews
Top Seller
Badge held · authenticated (red star)
2,370
Active listings · ~16,600 listed / 2mo
The basics are already done — and they're a moat. Every listing runs Buy Now + Ping + Afterpay. A 19-year, ~99%-feedback, Top-Seller account is something a new entrant simply can't buy — arguably one of the most valuable assets in the deal, and the IM undersold it badly. (A full 20-listing process audit — single photos, zero attributes, no Sponsored Listings — is in the Trade Me deep-dive →)

The real CRO gaps — where the wins are:

  • Titles don't lead with the formula, and rarely say "Refurbished". A refurb specialist whose titles don't signal refurbished/tested loses both trust and keyword relevance. Front-load Brand → Item → Model → condition; templatise per category (an AI-first cataloguing win).
  • Subtitles wasted on "CHECK OUT OUR 2700+ LISTINGS". Only ~half the sample used the 50-char subtitle, and those that did shout cross-promo instead of a conversion hook. Swap to 12-mo warranty · Tested · Fast ship.
  • Zero free shipping in the sample (it showed ~$9.50 ship). Free shipping on higher-value units unlocks the Free Shipping search filter (visibility) and lifts conversion.
  • Title hygiene — typos, ALL-CAPS, truncation ("SONUS PALBASE", "Lisitings"). A templated, AI-generated title/subtitle system fixes this across 2,370 listings at once.
  • Unfocused even on Trade Me — listing across Clothing (178), Motors (30), Baby gear, Art… not just refurb electronics. Reinforces the brand-specialist focus argument.
Already in practice: they run $1-reserve auctions for faulty/dead stock — so the dead-stock-auction play is live. The upside is routing slow-but-sellable stock to the email engine (free-shipping add-ons) rather than auctioning it for cents.
The throttle — protect the 99.16% feedback above all else. The ≥98% Top-Seller account is the single most valuable asset in the deal, and both the $1-auction velocity (dead-stock clearance) and the automated-listing volume (AI-generated titles/attributes at scale) can quietly erode it if run flat-out. So the velocity is explicitly gated to the feedback score: (1) ring-fence the $1 dead-stock auctions — faulty/clearance lots only, clearly graded and described, never the slow-but-sellable stock (that goes to the email engine); (2) QA-sample the AI-generated listings before and after publish, so a templating error never ships across thousands of listings; and (3) pace volume scaling to hold feedback ≥98% — if the score dips, throttle the auction batch size and listing throughput back until it recovers. Reputation is the asset; velocity serves it, never the reverse.
LeverThe moveTier
Title formulaBrand → Item → Model/Colour → "Refurbished" in the first 3–4 words. The one confirmed ranking signal — first 3 words carry most weight. e.g. Dyson V11 Cordless Vacuum — RefurbishedDo first
Sponsored ListingsPPC on Buy Now stock — pay-per-click, daily budget, scales catalogue-wide. The best paid-ROI lever; concentrate budget on high-margin/high-demand SKUs.Do first
Optimise the live StoreTYDI already runs a Trade Me Store (trademe.co.nz/…/stores/Tydi) — so this is an audit-and-tighten, not a setup: store branding/header, authenticated account (red star), and using the free-listing allowance + bulk tools well.Audit first
Fill 100% of attributesBrand, model, condition, every spec — powers filtered electronics search (how many buyers shop).Do first
Subtitle (50 chars)Conversion hook under the title: 12-mo warranty · Free shipping · TestedConvert
Free shippingWhere margin allows — unlocks the Free Shipping search filter (visibility) and lifts conversion.Convert
Refurb trust blockIn every description (up to 2,048 chars): grade, what was tested/replaced, real flaw photos, warranty + returns. Transparency is THE conversion driver for second-hand electronics.Convert
Ping + Afterpay + Make-an-offerAlready on every listing ✓ (audit confirmed Buy Now + Ping + Afterpay). Maintain — Afterpay ASP ~24% higher; Ping = Buyer Protection to $5,000.Done — keep
6+ own photosItem-only main image, all angles, flaws shown, plain + in-situ. (Your AI photo studio feeds this.)Convert
Top Seller + feedbackAlready Top Seller ✓ — 99.16% / 10,586 reviews since 2007. Protect it; it's a trust moat a new entrant can't buy. Guard ≥98%.Done — protect
Gallery / Was-Now / Q&AGallery Plus broadly (cheap thumbnail prominence + free relists); Was/Now strike-through to frame refurb savings; answer Q&A fast; keep listings fresh.Compound
The economics just improved. Trade Me scrapped its ~7.9% success fee (≈10 March 2026 cutover — verify), replacing it with a 2.19% Ping seller fee + a buyer-side service fee (99c–$4.99). Re-model unit economics on listing + Ping 2.19% + Sponsored spend, not the old 7.9% — refurb margins on the dominant channel just got materially better. (One watch-out: the new buyer-side fee nudges total basket cost, so factor it into price competitiveness vs new.)
Verify before quoting: the exact success-fee change date, current General Items promo prices (the $59–$179 figures floating around are Property-category, not General Items), and Store cost (~$399 setup / ~$79/mo) — all need confirming on Trade Me's live fees pages.

Part B — Website CRO (the bigger multiple, once measured)

3 Three surfaces, one principle: de-risk the refurb buyer

A refurbished buyer's first question is "is this dodgy?" — so trust and transparency are the conversion levers, on every page.
SurfaceThe jobKey levers
Collection pages
(where SEO/ads land)		Browser → product clickFilter by brand / condition-grade / price; trust bar up top (NZ warranty, returns, "refurbished explained"); "from $X" + savings-vs-new; strong product-card imagery; social proof; clear hierarchy
Product pages
(the money page)		Click → buyMultiple real photos + lifestyle + UGC video; condition-grade transparency + testing process; warranty & returns prominent; price-vs-new saving; scannable specs; Afterpay/Laybuy; cross-sell parts/accessories; sticky CTA; free-shipping threshold (feeds the email engine); reviews/Q&A
Advertorial LPs
(hero / high-value SKUs)		Cold paid traffic → saleLong-form direct-response: hook → cost-of-new / e-waste problem → certified-refurb solution → proof (process, warranty, reviews) → offer → single CTA. Run paid traffic here for controlled, measurable conversion
The compounding move: the same trust assets do double duty — the condition-grading system, real product media (the AI photo/video studio), and review content (mined by voice-of-customer) lift conversion on the website and Trade Me and make the catalog legible to AI shoppers. Build the trust layer once; it pays off across every channel.

4 How it's built — and the order

Brain skills

  • sellable-landing-page-audit — audit current collection/product pages
  • frontend-design · sellable-landing-page-generator — rebuild them
  • nhb-copy + market-research — advertorial copy in real buyer language
  • sellable-voice-of-customer — mine reviews/returns for the exact objections to kill
  • quiz-funnel / settledloop — hero-product funnels · generate-image/remotion — the media

Sequence

  • 1. Install measurement (GA4 + Clarity + Stape) — get a real number
  • 2. Trade Me Tier-1 levers — immediate revenue on the live channel
  • 3. Product-page trust rebuild (grade, media, warranty, cross-sell)
  • 4. Collection-page filtering + trust bar
  • 5. Advertorial LPs for hero SKUs, fed by paid traffic
Trade Me mechanics from Trade Me Help, RNZ & integration-partner guidance (2025–26); Best Match weightings beyond title are inferred, not published. Website CRO benchmarks from Baymard/industry. Fee figures and effective dates must be confirmed on Trade Me's live pages before relying on them. Not financial advice.
← Back to the document hub