Manual eBay clothing keyword research without guessing.

Most eBay clothing keyword research gets messy because sellers change too many things at once.

They search one phrase, change the filters, try a broader phrase, switch to solds, change the category, add a fit term, remove a brand line, and then try to decide which keyword was better. By that point, the numbers are hard to trust because the search itself keeps changing.

The useful research is usually on eBay itself, because that is where the actual active and sold result pools are. But eBay only helps if the search is controlled enough to compare one idea against another.

Quotes are useful because they help tighten the test.

In practical terms, double quotes tell eBay that a word or phrase is intentional. A search like "Levi's" "501" asks eBay to pay attention to both quoted pieces. A search like "Levi's 501" asks for that phrase together in that order.

They are not magic. They do not make eBay search perfect. They do not remove the need to inspect the actual results. But when used carefully, quoted terms and quoted phrases can help you compare clothing keywords with less noise.

The point is not to make the search look fancy. The point is to change one wording choice at a time so the active and sold result pools are easier to judge.

Loose Searches Can Mix Too Many Buyer Pools

A loose clothing search can pull in a broad result pool.

That can be fine when you are browsing, but it is not always great when you are trying to test a specific keyword direction.

For example, a loose browsing search for BKE Tyler can behave differently than a cleaner research test like this:

"bke" "tyler"

The quoted version is a tighter test. It is trying to keep the important brand and line words treated as intentional parts of the search. That does not mean every visible result will be perfect, but it is still one of the cleanest ways to compare exact clothing keyword ideas on eBay itself.

That matters because clothing keywords often depend on specific details:

• brand
• model
• line
• fit
• cut
• wash
• fabric type
• style name
• buyer-language terms.

If the search gets too loose, you may be comparing items that do not really belong in the same buyer pool. If the search gets too tight, you may be looking at a result pool that is too small to trust. The job is to use quotes as a testing tool, then inspect the results and decide whether the pool makes sense.

What Quotes Are Useful For

In this workflow, quotes are useful when you want to test a specific word or phrase more deliberately.

The point is to isolate one wording choice. If you change the keyword, filters, category, and active/sold setup all at the same time, you cannot really tell which change helped or hurt the result pool.

For clothing research, that usually means quoting things like:

• brand names
• model names
• style names
• brand lines
• exact fit or cut phrases
• exact title phrases you are considering.

For example:

"bke" "tyler"

That search is not just asking, "What comes up for BKE Tyler in general?" It is a more deliberate test of whether those quoted terms create a useful active and sold result pool.

The important thing is to keep the language honest. Quotes tighten the test, but eBay is still a buyer-facing marketplace. If a search is too thin, eBay may broaden the result page, adjust filters, or show edge-case results so the buyer has enough items to look at.

Most of the time, that is something to notice, not something to panic about. The quoted search is still useful. You just inspect the result pool and decide whether eBay's adjustment matters for the item you are researching.

Separate Quoted Terms vs. One Quoted Phrase

There are many ways to combine quoted words and phrases. Two common patterns are especially useful to understand first.

First, you can quote separate terms:

"Levi's" "501"

That asks eBay to pay attention to both quoted terms. The words can still appear separately in the result pool. For example, a title might include Levi's near the front and 501 somewhere else.

Second, you can test the phrase as one quoted unit:

"Levi's 501"

That is a different test. Now you are checking whether that phrase together creates a different result pool than the separate quoted terms.

That difference matters because this is not just keyword research. It is also title-structure research. You are trying to learn whether buyers and successful listings are using the words separately, or whether the phrase itself is part of the stronger search pattern.

This is where quotes become useful for title research. You are not just asking whether Levi's 501 jeans sell. You are asking whether the wording structure changes the quality of the active and sold pools.

If the same filters are held steady, comparing these two searches can help you think through title direction:

"Levi's" "501"

versus:

"Levi's 501"

That does not mean the phrase version is automatically better. It means the phrase version is worth checking as its own result pool.

The main lesson is the double-quote method: quoted terms and quoted phrases let you test wording more deliberately.

Quotes Work Better When The Filters Stay Steady

Quotes are only one part of the research setup.

If you change the keyword and change the filters at the same time, the comparison gets weak.

For clothing, the filters that matter might include:

• category
• condition
• color
• fabric type
• fit
• style
• Buy It Now.

The exact filter stack depends on the item. A common jeans research setup might hold steady around something like:

Pre-owned
Men's Jeans
Blue
Denim
Straight
Buy It Now

Those example filters are not random. For a real item, the filters you choose for research should be the same honest item details you would be comfortable carrying into the listing. The exact stack changes by item, but once it matches the item, keep it as steady as possible while you test the wording.

If a detail is a judgment call, like a fit that could reasonably be labeled two different ways, treat that as its own later comparison instead of mixing it into the first keyword test.

If you are testing:

"bke" "tyler"

against:

"bke" "straight"

the comparison is more useful if the condition, category, color, fabric type, and other important filters are not changing in the background.

The cleaner the setup, the easier it is to tell whether the keyword changed the result pool or whether the filters changed the result pool.

When The Search Gets Too Tight

Quotes are useful, but tighter is not always better.

For this workflow, if a word belongs in the keyword test, I usually want it in quotes. The bigger risk is narrowing the search too far with too many search terms, too many filters, or both at the same time.

A small result pool is not automatically useless. Clothing research often works inside small corners of the market. But a tiny active or sold pool should be read with more caution because one or two listings can swing the number.

That is especially important with clothing because small wording differences can split the market into little pockets.

Sometimes that pocket is real. Sometimes it is noise.

If a quoted term suddenly shows a much stronger sell-through signal, that is not an automatic answer. It is a clue.

Open the solds. Look for the common factor.

Sometimes the term is pointing to a specific brand, sub-brand, branded style, collector phrase, item code, or niche that does not apply to your item. Sometimes it really is the buyer-language term you were looking for.

That difference matters. If the common factor applies to your item, the keyword may be valuable. If the common factor belongs to something you do not have, the keyword can make the number look better without making your listing better.

That is not wasted research. Sometimes you discover a corner of the market that does not help the item in front of you, but it does teach you what to recognize later. Make a mental or written note. The next time you see that brand style, item code, or buyer phrase while sourcing, you will know why it matters.

That is why quotes should lead to better inspection, not less inspection.

Always Inspect The Results

The number is never the whole research.

Even with quotes, eBay search can behave imperfectly. Filters can change. Result counts can shift. eBay can interpret searches in ways that are not obvious from the search bar. Sponsored or promoted results can also make the page feel less clean than the search looks.

That does not make the method useless. It just means the seller has to stay in control of the research.

The most common reason a search gets loosened is simple: the result pool is thin, and eBay is trying to show a buyer enough items to shop. That is not always a dealbreaker. If eBay drops a minor filter and the result pool still describes the same buyer market, the search may still be useful.

Before trusting the result, check the parts that matter:

• whether the important filters still describe the same item
• whether eBay broadened the search because the result pool was thin
• whether a critical filter like price, condition, size, or style changed enough to affect the conclusion
• whether the quoted term has a specific meaning you need to understand.

The goal is not to reject every search that is slightly imperfect. The goal is to understand what you are looking at before you trust the number.

A Simple Way To Use This In Real Research

Here is the basic workflow.

Start with a clean filtered search that matches the item.

For jeans, that might mean choosing the right category, condition, color, fabric type, fit, style, and Buy It Now filter.

Then test one wording choice at a time.

Example:

"bke" "tyler"

Check the active result count.

Check the sold result count.

Inspect whether the results actually match.

Then test a nearby wording change while keeping the filter setup steady.

Example:

"bke" "straight"

or:

"bke" "tyler" "straight"

The exact searches depend on the item. The method is what matters:

Same filters. One wording change. Inspect the result pool.

Over time, this helps you see which words are creating cleaner title direction and which words are pulling the search into a noisier pool.

Where CompDock Fits

You can do this manually.

The hard part is that eBay can make the workflow repetitive. If changing the search term forces you to rebuild filters over and over, the research gets slow fast.

That is where mistakes creep in. One test keeps the filter stack. The next test quietly loses a filter. Then the numbers look like they are comparing keyword ideas, but they are really comparing two different searches.

CompDock was built for that repetitive part.

CompDock is independent and is not affiliated with, endorsed by, sponsored by, or approved by eBay.

It is a local Mac helper for reusing a filtered eBay search URL that you provide, changing the keyword text, opening or copying the new search when you choose, and saving local research history. It does not decide what keyword is right. It does not replace checking the active and sold results. It just makes the repeated manual testing less annoying.

The research still belongs to the seller.

The point is not to make the search fancy. The point is to change one wording choice at a time so the active and sold result pools are easier to judge.

Ask to test CompDock

Active and sold results can be useful, but only if the searches are measuring roughly the same thing.

That is where a lot of eBay clothing research goes wrong.

The useful numbers usually come from eBay itself because that is where the actual active and sold result pools are. But those numbers only help if you control the search tightly enough to compare one pool against another.

A seller looks at one active result count, then checks a sold result count from a slightly different search, then compares the numbers like they belong together. The numbers may look clean, but the comparison is already damaged if the filters, category, condition, or keyword pool changed in the middle.

The goal is not to make the number look good. The goal is to make sure the number is comparing the same kind of search.

What Active Results Tell You

Active results are the current competition pool.

They show what is listed now under that search setup. A huge active count is not automatically bad. Sometimes a large pool gives you room to cut through weak listings if the sold side is strong enough and your listing is built cleanly. A small active count is not automatically good either. It may mean a tighter niche, or it may mean there is not much buyer activity there.

The active number matters most when it is compared against the sold number. That comparison is what starts to show whether the pool is crowded, healthy, or too thin to trust by itself.

The active count is only useful when the search matches the actual item.

For clothing, that means the active pool should be filtered around details that actually matter, such as:

• category
• condition
• color
• fabric type
• fit
• style
• Buy It Now when you are trying to avoid auction noise.

Buy It Now deserves its own callout. If you are researching fixed-price clothing listings, mixing auctions into the count can distort both the active pool and the sold pool. Auctions can sell differently, price differently, and pull in a different kind of buyer behavior. That does not mean auctions never matter, but they should not be mixed into the comparison by accident.

If those details are loose, the active count can be misleading. It may include items that would never compete with yours in a real buyer search.

What Sold Results Tell You

Sold results are a past demand signal.

They show what sold under that search setup. That can be valuable because it gives you a sense of buyer activity, but sold results do not prove your exact item will sell.

Sold results become more useful when you compare them back to the active pool. The simple manual sell-through formula used in this guide is:

sold results / active results x 100 = STR

For example, 30 sold results and 60 active results would be:

30 / 60 x 100 = 50%

You may see some tools or sellers use a different version, such as solds divided by active plus sold. That version can be useful too. The important part is to use one formula consistently when comparing keyword tests. This guide uses solds divided by actives because it is fast to calculate manually and easy to compare from one search to the next.

Solds can get noisy for the same reason actives can get noisy. If the sold pool includes the wrong condition, wrong category, wrong buyer intent, wrong brand line, or a bunch of auction results, the number starts to mean less.

That is why sold counts need context.

Do the sold listings actually match the item type?

Are they the same kind of jeans, shirt, jacket, or shoe?

Are they the same condition range?

Are the prices in the same world as the item you are listing?

Did the search pull in a broader pool because the wording was too loose?

The sold count is not the answer by itself. It is a signal that needs context.

The Filter Stack Has To Stay Steady

The cleanest comparison starts with one filtered search.

For example, if you are researching a pair of men's jeans, your setup might hold steady around:

Pre-owned
Men's Jeans
Blue
Denim
Straight
Buy It Now

That is not a universal rule. The right filters depend on the item. The point is to build a search that matches the real item, then keep that setup steady while you compare active and sold results.

If you change the filter stack while comparing active and sold counts, you may not be measuring the same buyer pool anymore.

A one-filter difference can be enough to damage the comparison. That is what makes this workflow annoying manually: even one or two filters can take time to rebuild, and one missing filter can quietly change what the numbers mean.

A color filter can change the pool.

A condition filter can change the pool.

A fit or style filter can change the pool.

Auction results can change the pool.

Category changes can change the pool.

Even if the keyword looks the same, the comparison can be bad if the filters are not the same.

Do Not Compare A Broad Search To A Narrow Search

This is one of the easiest ways to fool yourself.

Say the active search is broad:

"bke"

Then the sold search is tighter:

"bke" "tyler"

Those two counts are not measuring the same thing. The broad active search may include a much larger competition pool, while the tighter sold search may be measuring a more specific buyer pool.

That can make the comparison look worse or better than it really is.

The reverse can also happen.

If the active search is narrow and the sold search is broad, the active/sold comparison can look stronger than it deserves because you are comparing a smaller current pool against a wider sold history.

The safer rule is:

Same keyword structure. Same filters. Then compare active and sold.

Once that comparison is clean, you can test a new keyword direction. But when you do, change one thing at a time.

That is the same idea behind the broader keyword workflow: isolate the wording change. If the keyword changes and the filters change too, you are no longer learning cleanly from the active/sold comparison.

Spot Check Result Quality Before Trusting The Count

Even when the active and sold searches look consistent, it helps to spot check what eBay actually returned.

That does not mean you need to audit every result every time. Most normal search variation is just part of working inside eBay. The point is to catch the moments where the number looks cleaner than the actual pool.

For clothing, look for obvious issues such as:

• brand line
• item type
• condition
• fit
• wash
• fabric type
• price range
• whether the buyer intent seems similar.

This matters because a result count can look useful while the listings inside the count are not really the same kind of market.

The best time to slow down is when something feels off: a keyword suddenly spikes, a result count changes harder than expected, the price filter refuses to stay, or the page starts showing a visibly different kind of item.

If the result pool clearly does not look like your actual item, the number should not carry much weight. If the pool is mostly right with a few weak or messy listings mixed in, that is normal. Those weak listings may be part of the reason a better-built listing can still compete.

eBay Can Still Change The Loaded Search

There is another wrinkle: eBay controls the loaded result page.

You control the search inputs. You can choose the keyword, quotes, filters, category, active/sold view, and Buy It Now.

But eBay still decides how the page loads. It can interpret the search, change what filters are available, remove or ignore a filter, broaden the result pool, or treat one category/search combination differently than another.

That does not make the research useless. Most of the time, eBay is trying to load a result pool that has enough items for a buyer to look at. If eBay broadens a search slightly for you, it may be doing something similar for the buyer who searches that term.

That can happen in ways that feel inconsistent. A price filter might refuse to carry cleanly on one item, then behave normally on the next item. A filter that looks basic can still disappear if eBay decides the result pool should load differently.

The practical rule is to pay attention when an important filter changes the meaning of the research.

If a price filter matters and eBay keeps dropping it, that is worth fixing or noting.

If a minor filter changes but the visible result pool still matches the buyer pool you are trying to understand, that may not be worth stopping over.

This is not about blaming eBay or blaming the tool. It is about staying aware of what changed. The seller controls the search inputs, but eBay still has final say over what appears on the page.

A Simple Active/Sold Workflow

Here is the practical version.

Start with one clean filtered search that matches the item.

Check the active results.

Switch to sold results using the same search setup. If you are doing this manually on eBay, expect to rebuild or re-check the important filters because many of them may not carry over cleanly. If you are using a saved filtered URL or CompDock, the setup is easier to repeat, but eBay can still adjust the loaded page.

Spot check the pools when the number or the visible listings feel off.

Then calculate the simple STR:

sold results / active results x 100 = STR

Then write down:

Keyword tested
Filters used
Buy It Now or auction status
Active count
Sold count
STR
Notes about result quality

Only after that should you judge whether the comparison is useful.

Then, if you want to test another keyword or title direction, change one piece of wording and repeat the same process.

This keeps the research from turning into a guessing game.

The point is not just to collect counts. The point is to understand whether the same kind of buyer pool is showing up on the active side and the sold side.

Where CompDock Fits

You can do this manually, but it gets repetitive.

The slow part is rebuilding the same filters over and over while testing different keyword wording. That is where mistakes happen: one search keeps the filter, another search loses it, and suddenly the active/sold comparison is not as clean as it looked.

CompDock is independent and is not affiliated with, endorsed by, sponsored by, or approved by eBay.

CompDock helps with the repetitive part by reusing a filtered eBay search URL that you provide, changing the keyword text, opening or copying the new search when you choose, and saving local research history.

It does not replace checking the active and sold result pools. It does not know whether the comps are good. It does not make eBay search perfect.

It just helps keep the manual workflow more repeatable.

The goal is not to make the number look good. The goal is to make sure the number is comparing the same kind of search.

Ask to test CompDock

Sell-through rate can be useful, but it is easy to give the number too much power.

For eBay clothing research, STR is basically a comparison signal. It can help you see whether one keyword direction looks stronger or weaker than another, but it does not tell the whole story by itself.

The number can point you toward a better question.

It should not make the decision for you.

What STR Is Measuring

The simple version is:

sold results / active results x 100 = STR

If a filtered search has 50 active results and 25 sold results, the STR would be:

25 / 50 x 100 = 50%

That number can help you compare one search setup against another search setup. But it only means anything if the active and sold searches are measuring roughly the same kind of result pool.

That is why the filter stack matters first.

If the active side is filtered to pre-owned men's jeans and the sold side is accidentally pulling in new items, auctions, a different category, or a broader keyword pool, the STR number is already damaged.

If the active count is zero, the formula is not useful in the normal way. That does not automatically mean the keyword is amazing. It usually means the result pool is tiny enough that you need to inspect it manually before treating the number as meaningful.

STR is not magic math. It is only as clean as the search behind it.

It also inherits every eBay search problem underneath it. If eBay drops a filter, changes the category, broadens the result pool, or treats the sold page differently from the active page, the STR number can look precise while the comparison underneath it is not.

Start With A Baseline

The cleanest use of STR is not to look at one number by itself.

The better use is to build a baseline, then compare new keyword ideas against that baseline.

Start with one clean filtered search that matches the actual item.

For clothing, that might mean holding steady details like:

• category
• condition
• color
• fabric type
• fit
• style
• Buy It Now or auction status.

Then calculate the STR for that starting point.

That gives you a baseline.

After that, change one wording choice at a time. Change a brand-line phrase, a fit term, a wash term, a style term, or a title structure. Then compare the new STR against the baseline.

The baseline is what keeps the research grounded.

Without it, a single STR number can look impressive without telling you whether the keyword actually improved the search.

That is the heart of the method: the baseline tells you where you started, and each keyword test tells you whether the result pool moved in a better or worse direction.

What A Higher STR Can Suggest

If a keyword variation has a higher STR than your baseline, that may be worth a closer look.

It may mean the keyword is narrowing the search into a stronger buyer pool.

It may mean the wording matches the way buyers describe the item.

It may mean the item line, fit, style, or title phrase is more useful than the broader wording you started with.

But "may" is the important word.

A higher STR does not automatically mean the keyword belongs in your title. It means the keyword is worth checking.

Open the active results.

Open the sold results.

Look at what is actually there.

If the listings match your item and the sample is strong enough, the higher STR may be pointing you toward a better title direction. If the listings do not match your item, the number may be attractive but useless.

What A Lower STR Can Suggest

A lower STR can also be useful.

It may suggest the keyword is weakening the search.

It may suggest the word is too broad.

It may suggest the word is pulling in items that are technically related but not actually competing with yours.

It may also suggest that the buyer pool does not care much about that wording.

That does not mean the keyword is always wrong. Sometimes a lower STR term may still be needed because it accurately describes the item. A title is not only a numbers game. It still has to be honest and readable.

Sometimes a lower overall STR can also hide a stronger sub-niche. If the sold results share a pattern, such as a specific brand line, fit, fabric type, item code, era, or style name, try narrowing around that pattern before dismissing the keyword. The broad keyword may be weak overall but useful inside the smaller buyer pool where it actually belongs.

But if a keyword consistently drags the comparison below the baseline, that is a reason to question it.

In plain terms, it may be making the sell-through comparison drop because it does not fit the buyer pool you are trying to reach.

The useful question is not:

Is this keyword good or bad?

The better question is:

What kind of result pool does this keyword create?

Tiny Samples Can Lie

Small samples are one of the easiest ways STR can fool you.

If a search has 2 sold results and 1 active result, the STR looks huge.

That does not automatically mean you found a great keyword. It might just mean there is not enough data.

The same problem can happen in reverse. A tiny active pool with almost no solds can make a keyword look weak, even though there may not be enough volume to say much either way.

Small samples should not be ignored, but they should be treated carefully.

When the result pool is tiny, use STR as a reason to inspect the listings, not as a final answer.

Check whether the solds are similar enough to teach you anything useful.

Check whether the active items are actually similar.

Check whether the keyword is describing a real niche or just creating a weird little search pocket.

Keyword Spikes Need Investigation

Sometimes a keyword creates a dramatic STR change.

That is where the research gets interesting, but it is also where you can fool yourself fast.

A spike might mean you found a strong buyer term.

It might mean you found a specific brand line or item name.

It might mean the keyword has a meaning you did not realize.

It might also mean eBay pulled in a strange result pool.

For example, if you are researching women's denim capris and a phrase like pedal pusher or Pedal Capri suddenly looks much stronger than the broader terms, do not just drop the phrase into your title and move on.

The exact terms are not the main lesson. The lesson is the pattern: when one phrase suddenly changes the numbers, slow down and figure out why.

Open the results and figure out what the phrase is doing.

Is it a generic style term?

Is it tied to a specific brand, line, or item?

Are buyers using it broadly, or are they searching for one very specific kind of item?

Does your item actually belong in that result pool?

That kind of spike can be valuable. It can reveal a real corner of the market. But it can also trick you into using a keyword that does not fit your item.

That is why this kind of research is not only about math. It can teach you how buyers talk, what terms belong to specific item lines, and where a small corner of the market may be hiding.

The number starts the investigation.

The listings finish it.

STR Does Not Replace Listing Judgment

Even a useful STR signal does not replace the rest of the listing decision.

You still have to look at:

• condition
• size
• price range
• shipping
• photos
• brand line
• seasonality
• defects
• title accuracy
• buyer intent.

A keyword with a strong STR may still be wrong for your item.

A keyword with a weaker STR may still be necessary because it describes the item accurately.

A keyword can also look weak in a broad search and then become valuable once you find the smaller pattern where buyers actually use it.

A sold result may matter less if the condition, size, brand line, or price tier is different.

An active result may matter less if the listing quality, shipping, or item details are not similar.

STR helps you ask better questions, but you still have to inspect the market.

Where CompDock Fits

You can calculate STR manually.

The repetitive part is running the same filtered search over and over, changing one keyword or phrase, checking active and sold counts, and remembering what you already tested.

CompDock is independent and is not affiliated with, endorsed by, sponsored by, or approved by eBay.

CompDock helps with that repetitive part. It lets you reuse a filtered eBay search URL that you provide, swap keyword text, enter active and sold counts you choose to record, calculate STR, and save local research history.

It does not decide what to buy.

It does not decide what to list.

It does not know whether a sold comp actually matches your item.

That judgment still belongs to the seller.

STR is strongest when it sends you back into the actual results with a better question.

Ask to test CompDock