If your listings sit on page 4 of eBay search, the problem usually isn't your photos or your price — it's your title. eBay's Cassini ranks listings on a mix of relevance, sell-through, and recency, and the relevance signal starts with the words a buyer actually types. An eBay keyword tool is how you find those words instead of guessing.
This guide walks through the workflow we use at Ecomli to research keywords for new listings: which tools matter, how to score the terms they return, and how to fit them into 80-character titles without keyword-stuffing your way out of the algorithm.
What an eBay keyword tool actually does
An eBay keyword tool pulls search data from eBay itself — usually by scraping autocomplete, parsing completed listings, or tapping a paid data feed — and turns it into a ranked list of buyer search terms. The good ones add three layers on top of raw search volume:
- Sell-through rate — how many listings with that keyword actually sold over the last 30–90 days. A 10,000-search keyword with 8% sell-through is a worse target than a 1,200-search keyword with 65% sell-through.
- Average sale price — tells you whether the term attracts buyers or window-shoppers.
- Listing competition — the number of active listings already targeting that term. Volume divided by competition is your real opportunity score.
A generic Google keyword tool won't give you these. Google sees "wireless mechanical keyboard" the same way for a Best Buy SEO writer and an eBay reseller, but eBay's buyer pool behaves differently. Buyers on eBay search with model numbers, brand-plus-condition phrases ("apple watch series 9 new sealed"), and category-specific qualifiers that almost never trend on Google. You need a tool that reads eBay's data, not Google's.
Which eBay keyword tools are worth using
The eBay keyword tool market splits into four buckets. Pick one from each bucket and you'll have everything you need.
1. eBay's own research data (free)
Inside Seller Hub there's a Research tab with terapeak-powered data on completed listings, sell-through, and average price by keyword. The interface is slow and the filters are limited, but the data is the source of truth — everyone else's tool ultimately compares back to this. Use it to validate any keyword another tool tells you to chase.
2. Autocomplete scrapers (free or cheap)
Tools like Keyword Tool Dominator and keywordtool.io hammer eBay's autocomplete endpoint with seed terms and return every suggestion eBay surfaces. This is how you find long-tail variants buyers type but you'd never invent. "Camping tent" returns a dozen seeds; "camping tent" piped through autocomplete returns 400+ specific phrases like "camping tent 4 person waterproof with screen room." That last phrase is the listing title.
3. Sell-through analyzers (paid)
Zik Analytics, Algopix, and Terapeak's premium tier add the sell-through layer. They scrape completed listings, calculate the percentage that sold, and let you filter by date range and category. This is where you separate "lots of searches" from "lots of buyers." Expect to pay $30–$60/month per seat. If you're listing 50+ products a week, the math works.
4. Title and listing optimisers (built-in or bolt-on)
Once you have keywords, you need to fit them into an 80-character title that reads like a human wrote it. Title-builder.com is the free option. Ecomli builds this into the listing flow: when you import a product, the keyword scoring runs automatically against eBay's autocomplete data, surfaces the highest sell-through phrases for that category, and assembles a draft title you can edit before pushing live. That's where the workflow stops being copy-paste-from-spreadsheet and starts being a one-pass operation.
The keyword research workflow we use at Ecomli
Tools are commodity; the workflow is what changes outcomes. Here's the order we run for every new product before it goes live.
Step 1: Pull seeds from completed listings
Open eBay's Research tab, search the product, filter by sold listings in the last 90 days, and skim the top 20 titles. Note every word that repeats. These are the buyer-validated tokens — brand names, model numbers, condition flags ("new sealed", "open box"), bundle qualifiers ("with charger"), and size/colour modifiers. Don't try to be clever; copy what's already working.
Step 2: Expand with autocomplete
Take your top 3–5 seed terms and run them through an autocomplete scraper. You'll get 200–500 long-tail variants. Most are noise. The 10–20 that match real buyer intent for your specific product become your candidate keyword pool.
Step 3: Score by sell-through, not search volume
Drop the candidate pool into a sell-through tool. Sort by sell-through rate descending, then filter to volume above some floor (we use 200 monthly searches for high-velocity categories, 50 for niches). Anything under 30% sell-through is dead weight regardless of search volume.
Step 4: Build the title
You have 80 characters and roughly 8–12 keyword tokens to spend them on. Order matters less than people think — Cassini reads the whole title as a bag of words for relevance — but readability matters for click-through rate, which feeds back into ranking. Lead with the brand and model, then condition, then 2–3 differentiators (size, colour, bundle), then the catch-all category term.
Example for a refurbished iPhone case: "OtterBox Defender iPhone 15 Pro Case Black Heavy Duty Shockproof Belt Clip" — brand, model, condition implied by the absence of "used", three differentiators, and a final commerce term ("belt clip") that catches buyers searching by feature rather than brand.
Step 5: Track and iterate
After 14 days, pull the impression and click data from your seller dashboard. If a listing has impressions but no clicks, the title isn't matching buyer expectations — the keywords are too broad. If it has no impressions, the keywords aren't matching searches at all — rebuild the title with the next ranked candidate. Most listings need one revision to stabilise. Some need three.
Common mistakes that kill keyword research
The fastest way to waste a keyword tool subscription is one of these patterns. We see them constantly in seller audits.
Stuffing synonyms instead of variants. "Phone Case Cover Cover Cover for iPhone" gets flagged by Cassini and demoted. eBay's algorithm penalises repeated tokens. One instance of each keyword is the cap.
Optimising for volume, ignoring intent. "iPhone case" gets 90,000 monthly eBay searches. It also has 1.4 million active listings. You won't rank. "iPhone 15 Pro Max case magsafe clear" is 2,400 searches with 18,000 listings. Different math.
Copying competitor titles verbatim. Cassini detects near-duplicate titles and de-ranks them. Use the same keywords, but in a different order with different connector words.
Ignoring item specifics. Item specifics (the structured fields below the title) are themselves a ranking signal. A keyword tool tells you what to put in the title; item specifics tell Cassini which categories and filters to surface your listing in. Fill every relevant one. Ecomli's listing builder pre-fills item specifics from supplier data when it imports, so you're not typing brand and model 200 times.
Not refreshing keywords seasonally. "Halloween costume women" peaks in October. "Air conditioner portable" peaks in July. Keyword tools that show 12-month volume history will catch this; tools that show only "monthly average" will hide it. Always check the trend line before committing a title for a seasonal product.
Where Ecomli fits in
If you're managing 50+ listings, manual keyword research per listing stops scaling at about 20 hours a week. Ecomli's listing engine runs the autocomplete + sell-through scoring automatically when you import a product from your supplier feed, surfaces the top-ranked keyword candidates for the category, and drafts a title you can edit before publishing. The same engine flags listings whose impressions drop more than 30% week-on-week, so you know which titles to rewrite without reading every dashboard.
For deeper category research, see our eBay product research method and our writeup on how to add keywords to eBay listings — together they cover the upstream (which products to list) and downstream (how to wire keywords into item specifics) sides of this workflow.
Frequently asked questions
Is there a free eBay keyword tool that's actually worth using?
eBay's own Research tab inside Seller Hub is free and pulls from terapeak data. For autocomplete expansion, keywordtool.io's free tier returns the top 30–40 suggestions per seed, which is enough for low-volume sellers. The paid layer is worth it once you're listing more than 10 products a week.
How many keywords should I put in an eBay title?
Use every one of the 80 characters, but don't stuff. Eight to twelve meaningful tokens is typical: brand, model, condition, 2–3 differentiators, and a category catch-all. Repeated words are penalised, so each keyword appears once.
Does eBay use the same keywords as Google?
Some overlap, but the buyer behaviour is different. eBay searches skew toward brand-plus-model and condition phrases ("ps5 console disc edition new sealed") that almost never trend on Google. Always research keywords with an eBay-specific tool, not a generic SEO tool, even if the SEO tool is more polished.
How often should I update keywords on existing listings?
Re-run the research every 60–90 days for evergreen products and every 30 days for seasonal ones. Pull each listing's 14-day impression and click data, and rebuild any title where impressions dropped more than 30% week-on-week. Most stable listings need one revision per quarter.
Can keyword tools predict which products will sell?
They can rank candidates by demand and competition, but they can't predict winners. A keyword with high sell-through tells you buyers want this category of product — it doesn't tell you whether your specific listing, at your specific price, with your specific photos, will be the one they click. Treat keyword data as the floor, not the ceiling.
Ready to automate your eBay business?
Ecomli handles product sourcing, listing, repricing, and fulfilment — so you can focus on growing.
Start for $1 →14-day trial · Cancel any time · No questions asked