A keyword modifier in a Keyword Combiner is the second term or short phrase you place in list B, and choosing it well means matching a specific intent, audience, or attribute against every base keyword in list A so the tool can produce a useful Cartesian product of candidate phrases. This article explains what belongs in the modifier list, when modifiers are better placed in list A instead, the categories of modifiers that tend to produce usable output, and how to evaluate the first run before exporting. The Keyword Combiner treats both lists mechanically — it does not judge whether a modifier makes business sense, so the choice of lines for list B is yours. Each non-empty line in list B becomes one input across up to 10,000 input pairs, and the order you select (A then B or B then A) together with the separator you enter determines the literal text of every output phrase. Picking modifiers therefore has three consequences: which phrases exist at all, what the surface text looks like, and which pairs fail the 10,000-pair cap before any output is generated.
What Counts as a Keyword Modifier in a Combiner
In a Keyword Combiner, a modifier is any line you place in list B — or, if you reverse the order, list A. The tool does not assign one list to "the modifier" role; it treats both lists as equal inputs, and the distinction is about which list contains the variable part of the phrase and which contains the fixed head. If your campaigns are built around product names, the product names usually belong in one list and the descriptive attributes belong in the other.
A useful modifier is a single term or short phrase that, when joined to every base keyword, produces phrases a real user might type. Modifiers commonly take the form of intent words ("buy", "review", "vs"), audience words ("men", "kids", "small business"), attribute words ("cheap", "premium", "organic"), or format words ("guide", "template", "pdf"). Lines that are not single modifiers — full sentences, multi-clause descriptions, or marketing slogans — will combine into phrases that read awkwardly and are most likely removed during your manual review step.
Lines that look like modifiers but actually repeat the same idea with slight spelling variants waste pair budget. The combiner deduplicates exact case-sensitive matches inside each list after trimming outer whitespace, so "Shoe" and "shoe" remain distinct on purpose, but " running shoes" and "running shoes" collapse to a single entry because the trim step runs first.
Common Modifier Categories Worth Trying
Choosing modifiers becomes easier once you know the categories that consistently produce useful phrases. The table below summarizes the most common categories, the kinds of list B lines each one holds, the kind of list A head terms they pair with, and the failure mode to watch for when the run completes. The exact counts vary by project; the relationship between modifier size and output size is what the tool reports.
| Modifier category | Example list B lines | Best paired with list A containing | Watch for |
|---|---|---|---|
| Intent | buy, best, cheap, review, vs, alternative | Product or topic head terms | Too narrow, may ignore informational searches |
| Audience | men, women, kids, small business, enterprise | Product or service categories | Phrases that assume a gender or market segment |
| Attribute | organic, premium, handmade, waterproof | Product nouns | Attribute-product mismatches that read as nonsense |
| Format | guide, template, checklist, pdf, video | How-to or topic heads | Format-only modifiers that fail to signal intent |
| Location | near me, UK, California, Austin | Service or retail categories | Location modifiers with no matching landing page |
A common starting size is 10 to 25 modifiers across two or three of these categories, which keeps the pair count manageable before you scale up. Categories outside this list — seasonality words, brand-specific qualifiers, pricing tiers — also work, as long as each line is short enough to read as a modifier when joined to a base keyword.
Where Modifiers Belong: List A or List B
The combiner gives you two lists and lets you pick the order. The choice is not arbitrary. If your modifiers describe how a searcher wants to engage with the topic, list A often holds the topic ("keyword research") and list B holds the modifier ("tool", "guide", "vs ahrefs"). The resulting phrase reads as "keyword research tool", "keyword research guide", and so on. If your modifiers are the more variable part — color variants, for example — list B usually holds the colors and list A holds the product, so the head noun stays in a predictable position for grammar and ad-group grouping.
A practical rule: put the smaller, more variable list in the position whose tokens you want to control most tightly. If you are generating phrases like "red running shoes" and "blue running shoes", the colors belong in the list that produces the trailing token, because search engines and ad systems usually prefer a stable head term followed by a variable attribute. The same rule applies when modifiers are intent words rather than attributes — the head term should sit in the position it already occupies in your existing keyword taxonomy.
The order selector changes both the loop order and the phrase order. Switching from A-then-B to B-then-A rewrites the output sequence in a predictable way, which makes diffs between runs easy to read. For more on how order, separator, and pair counts interact, see the guide on combining keyword lists with order, separator, and pair counts.
How to Build and Combine Modifier Lists Step by Step
- Write each base keyword or phrase on its own line in list A — one per line, no commas or semicolons separating entries.
- Write each candidate modifier on its own line in list B. Aim for 10 to 30 lines drawn from the categories above; more than 50 raises the pair count quickly.
- Choose A-then-B or B-then-A depending on which token should lead the phrase. Pick the separator that produces natural text — usually a single space, sometimes a hyphen for compound modifiers.
- Run the combine. The tool computes the theoretical A × B pair count, applies trimming and within-list deduplication, and rejects the request entirely if the input pair count exceeds 10,000.
- Read the summary block: unique output count, normalized list sizes, theoretical pair count, removed input duplicates, and removed output collisions. If the unique output count is much smaller than the theoretical pair count, expect colliding concatenations and revisit the separator or the lists.
- Manually review the output for relevance, landing-page fit, and grammatical quality. The combiner does not perform these checks.
- Copy the deterministic one-per-line result with the copy button. Re-running with identical input yields identical text, which makes the output suitable for version comparison.
Reading the Summary After the First Run
The summary panel is the audit trail for your modifier choices. It reports five numbers that together tell you whether list B worked as intended. The unique output count should be close to the theoretical pair count unless your separator is empty and adjacent terms happen to form identical strings — for example, "ab" plus "c" and "a" plus "bc" both yield "abc". When that happens, the later phrase is dropped and reported as an output collision. A high number of removed input duplicates means you repeated modifiers inside list B, often because of trailing spaces or capitalization drift between drafts. A high number of removed output collisions usually means your separator is empty or your terms are too short; switching to a single-space separator almost always reduces collisions without changing the meaning of legitimate phrases.
If the summary shows the request failed before any output, the cause is the pair cap. Reduce one list, narrow the categories, or split the work into several runs with different modifier slices. The tool never returns a partial sample or silently keeps the first 10,000 pairs, so a clean failure is the only signal that the modifier list needs to be cut down.
Limits That Shape Modifier Choices
The combiner's published limits interact, and modifier choices affect all of them. Each list accepts up to 500 unique case-sensitive terms, each term is capped at 100 Unicode code points, and each editor accepts up to 100,000 UTF-16 code units of pasted text. The pair budget is calculated before output deduplication and rejects the whole request above 10,000 input pairs. A 200-line list A and a 200-line list B produces 40,000 theoretical pairs, which the tool will refuse. Choosing a narrower modifier list, or splitting the work into several runs, keeps the request within budget. Modifiers that exceed 100 Unicode code points are also rejected because the trim-and-cap step runs before deduplication. Internal spaces, punctuation, accents, emoji, and non-Latin scripts remain unchanged inside each term, so a modifier written with an em dash or a non-Latin script is preserved as written.
After the Combiner: Demand Still Needs Validation
A successful combine proves only that every pair exists on paper. It does not prove that any phrase has measurable search demand, that a phrase matches a landing page, or that a phrase respects trademark or platform policy. The tool does not estimate search volume, competition, cost per click, intent, language, relevance, or grammatical quality, and it does not query an advertising account, search engine, keyword provider, or AI service. Before pasting the output into a campaign draft, remove modifiers that produced unnatural or off-brand phrases, confirm each remaining phrase has a destination page, and validate demand through a first-party source such as your own analytics or a paid keyword-data provider. The Keyword Combiner is a string-combination utility, not keyword research, and treating the output as a candidate list rather than a final target keeps expectations honest.
If you're weighing options, Keyword Density Checker Example: Formula and Numbers covers this in detail.