A listing title does two jobs at once, matching searches and winning the click against nineteen neighbors, and every marketplace weighs the jobs differently. The same product needs different titles per channel, not because gurus say so, but because the platforms’ search engines and buyers reward different shapes, one truthful core, adapted per dialect.
The dialects
Amazon: spec-dense, structured. Brand + product line + key attribute + size/count + differentiator, in roughly that order, because Amazon’s search leans on structured data and its buyers scan for specs. Category style guides exist and matter, and category choice sets them. Filler adjectives cost characters that specifications should hold.
eBay: search-term coverage. eBay buyers type wildly varied queries, and titles are the primary match surface: the 80 characters want brand, product type, model, size, color, condition, and the honest synonyms buyers actually use (sold-listings research shows the winning vocabulary per category). This is coverage, not stuffing, every term must truthfully describe the item.
Etsy: human phrasing with search anchors. Etsy queries read like gift briefs (“personalized leather journal for him”), and titles that mirror that phrasing match both the engine and the browsing mood, anchored by the concrete nouns, with the personality Etsy buyers came for.
Your store: brand voice with clarity first. No marketplace engine to appease, your data quality standards and your customers’ vocabulary set the bar.
The universal rules under the dialects
- Front-load ruthlessly. Every surface truncates; the first 40-60 characters must carry the product. The test: does the truncated title still identify it?
- Truth is the conversion strategy. Titles write checks the product cashes, or returns and disputes cash them instead. “Fits most” and inflated superlatives are future reason codes.
- Write for the scanner. Twenty results, two seconds each: capitalization discipline, no ALL-CAPS shouting, no emoji clutter, the concrete noun early.
- Variants get specific titles: the child’s color and size in ITS title, per the variation discipline, so search lands on the right child.
- Machine drafts, human finals: generation tools draft dialects fast; the ranking weight titles carry deserves a human pass.
The testing loop that beats folklore
Titles are one of the few listing elements with a clean feedback loop:
- Baseline: note a listing’s impressions/views and conversion for two weeks (marketplace seller dashboards expose enough of this).
- One change at a time: reorder the front-load, swap a synonym, add the missing spec, single variables, or the result teaches nothing.
- Read velocity and click-through over the next two weeks against baseline, the same measure-then-decide habit as everything else.
- Roll winners across the family: a title pattern that lifts one child usually lifts its siblings, and sometimes the category.
The meta-rule: your category’s buyers vote with clicks, and their vote outranks any template, including this article’s.
Common questions
Should titles include the brand name if nobody searches it?
On Amazon, usually required by style. Elsewhere: early brand mention costs prime characters; young brands often lead with the product and earn the brand-first title later, the brand-building arc.
Do keyword tools help with titles?
For vocabulary discovery, yes, sold-listing language and marketplace autocomplete are the ground truth; volume tools rank the candidates.
How often should I revisit titles?
When the testing loop says so, and at seasonal turns where gift-phrasing enters queries. Constant fiddling without measurement is motion, not optimization.
Does changing a title hurt existing rank?
Marketplaces re-evaluate on edits; a better-matching title recovers and exceeds, a worse one does not. Test on secondary listings before touching the flagship, A-item discipline again.
One product, every dialect spoken
Truthful cores, per-channel titles, and the sync that keeps every listing consistent, from $49/month with unlimited orders. See pricing.