Excel COMBIN and COMBINA both return a single integer — the number of ways to choose k items from n — not a list of the actual combinations. The difference between them is repetition: COMBIN counts combinations without repetition, so once an item is chosen it cannot appear again inside that same selection, while COMBINA counts combinations with repetition, so the same item may appear multiple times within a single selection. For n = 5 and k = 2, COMBIN(5, 2) returns 10 and COMBINA(5, 2) returns 15, because allowing an item to repeat increases the count.
The shared misconception that drives searches like "excel combin vs combina" is the word "combine" itself. Both functions count combinations; neither one generates them as text. If your real task is to take two lists of words and produce every possible pair (for ad groups, content briefs, taxonomy reviews, or internal keyword matrices), Excel's math functions will not get you there directly. That is a string-combination problem, and it deserves a string-combination tool such as the Keyword Combiner, which builds every pair as a deterministic, one-per-line output that you can paste straight into a spreadsheet, an ad-group draft, or a content brief without uploading your lists anywhere.
Excel COMBIN vs COMBINA: What Each Function Actually Counts
Microsoft's documentation treats COMBIN and COMBINA as siblings inside the Math and Trig category. Their arguments look identical — a "number" of items and a "number chosen" per subset — but the model behind each is different enough that the same inputs produce different integers, and Excel will return #NUM! or #VALUE! in defined ways when those inputs go out of bounds. Knowing which function answers which question is the first step before you ever open a spreadsheet.
| Feature | COMBIN | COMBINA |
|---|---|---|
| Purpose | Count combinations without repetition | Count combinations with repetition |
| Syntax | COMBIN(number, number_chosen) | COMBINA(number, number_chosen) |
| Repetition allowed inside one selection | No | Yes (sometimes called "with replacement") |
| Number argument | Must be ≥ 0 and ≥ number_chosen; non-integers are truncated | Same limits as COMBIN |
| Number_chosen argument | Must be ≥ 0; non-integers are truncated | Same limits as COMBIN |
| Negative or out-of-range input | Returns #NUM! | Returns #NUM! |
| Non-numeric input | Returns #VALUE! | Returns #VALUE! |
| Worked example (n=5, k=2) | 10 | 15 |
| Return type | Single integer in the active cell | Single integer in the active cell |
The "without repetition" rule means COMBIN treats each item as distinct and refuses to put it back into the selection after it has been used. The "with repetition" rule in COMBINA follows the formula C(n + k − 1, k). For a marketer the practical takeaway is that COMBINA always returns a number that is greater than or equal to the COMBIN value for the same inputs, and that gap widens as k grows. The functions never list which combinations exist; they only tell you how many exist.
The Worked Example: COMBIN(5, 2) and COMBINA(5, 2)
Both formulas take the same two numbers, but the underlying model is different enough to give a clean, verifiable contrast. Walking through the arithmetic makes the rule difference obvious.
For COMBIN(5, 2), the math is the standard combinations formula: n! divided by (k! times (n − k)!). Substituting 5 for n and 2 for k gives 5! / (2! × 3!). Computing each factorial: 5! = 120, 2! = 2, and 3! = 6. The denominator becomes 2 × 6 = 12. So COMBIN(5, 2) = 120 / 12 = 10.
For COMBINA(5, 2), the model allows the same item to appear more than once. The closed form is C(n + k − 1, k), which expands to (n + k − 1)! divided by (k! × (n − 1)!). Substituting gives (5 + 2 − 1)! / (2! × (5 − 1)!) = 6! / (2! × 4!) = 720 / (2 × 24) = 720 / 48 = 15.
Both numbers are what Excel returns when you type =COMBIN(5,2) and =COMBINA(5,2) into a cell. They describe how many distinct selections exist under each rule; they do not list which selections those are. That distinction is the entire reason the "excel combin vs combina" question feels unsatisfying once you start typing formulas — the cell holds one number, not a candidate list.
Why These Functions Don't Build Keyword Lists
Search-engine optimization work frequently involves pairing a base term with a modifier — `running shoes`, `running socks`, `running watch` — and the natural Excel instinct is to scan the function library for anything that sounds like "combine". COMBIN and COMBINA are the closest names, but both return a scalar integer rather than text, so they cannot build a campaign-ready keyword list.
Spreadsheet workarounds do exist, and they all share the same fragile pattern: paste one list into a column, paste a second list into a row, use INDEX/MATCH or TEXTJOIN with a clever row/column index trick, then drag the formula down and out. The technique works for a 5 × 5 grid but quickly becomes unreadable when one list grows past a few dozen terms, and it produces silent duplicates when the lists contain overlapping phrases or terms with irregular spacing. For a marketer who just wants every ordered pair, the formula approach is the wrong shape of solution.
The mathematical concept behind the right tool is the Cartesian product. If you have three colors and two product names, you get 3 × 2 = 6 input pairs — every color paired with every product. The same pairing logic, applied to two keyword lists, gives you the full set of candidate phrases to review. The deeper background on this pairing concept lives in the Cartesian product primer for keyword combinations, and the same idea translates directly into practical two-word keyword building described in building keywords from two lists. Understanding that the keyword combination task is the same combinatorics problem in a different costume helps you choose a tool that handles the strings, not just the count.
What Marketers Actually Need: A Real Keyword Combination Tool
The task marketers are usually trying to do in Excel is one of three things: build a starter list of long-tail variants for an ad group, draft a content brief from a topic plus modifiers, or sanity-check the internal taxonomy of a category page. None of those tasks are solved by counting combinations; all of them are solved by listing combinations.
A dedicated keyword combination tool takes one list of base terms and one list of modifiers, lets you pick the order and the separator, and produces every pair as a clean, one-per-line output. Because the operation is mechanical, the output is also deterministic: identical inputs and identical settings always produce identical text. That property is what makes the output suitable for version comparison, copy-paste into a spreadsheet column, or piping into another tool for analysis later.
The Keyword Combiner in particular runs entirely in the current browser tab, which means your lists are not uploaded to a server, do not reach a search-engine API, and are not stored after you close the page. Every generated phrase is a string candidate only — the tool does not estimate search volume, competition, intent, language fit, or grammatical quality, so relevance still depends on a human reviewing the list before any of it lands in a campaign or a published page.
How to Combine Two Keyword Lists With the Keyword Combiner
- Paste list A and list B with one term per line. The base terms go in list A, the modifiers or second terms go in list B. The tool trims outer whitespace on each line, ignores empty lines, and removes exact case-sensitive duplicates inside each list, so `Running shoes` and `running shoes` stay separate on purpose.
- Pick A-then-B or B-then-A order, and type the separator. A space, a hyphen, or nothing (an empty string) all behave exactly as typed — there is no hidden spacing added. Switching the order changes both the loop order and the visible phrase order, which keeps the diff between runs predictable.
- Run the combination and read the audit summary. The summary shows normalized list sizes, the theoretical pair count (A × B), how many duplicates were removed from each input list, and how many output collisions were collapsed. Use it to confirm the math matches your expectation before copying.
- Review the candidate list and remove anything irrelevant. Some pairs will sound unnatural, repeat a competitor's brand name, or duplicate a phrase that already exists on a landing page. Treat the output as a draft, not a final list.
- Copy the result as a single block. The one-per-line output is meant for direct paste into a spreadsheet column, an ad-group draft, a content brief, or a later keyword-data check. Reordering happens outside the tool if you need it.
Limits, Gotchas, and Audit Checks Before You Publish
The Keyword Combiner enforces hard limits because silent truncation is the worst possible failure mode for a keyword list. Each term may contain at most 100 Unicode code points. Each editor accepts up to 100,000 UTF-16 code units of pasted text and 500 unique terms. The pair budget is computed before generation, and any request above 10,000 input pairs fails the entire request rather than returning a partial sample. Knowing these numbers up front lets you plan your input lists around them.
Output collisions are different from input duplicates. With a space as the separator, three colors times two products produce six phrases and the count lines up with the math. With an empty separator, pairs such as `ab` + `c` and `a` + `bc` can both produce `abc`, and the tool keeps the first occurrence and reports the collision separately. If your summary shows an unexpectedly large output collision count, change the separator or shorten one of the lists before copying.
Before publishing or pushing the list into a campaign, remove irrelevant phrases, confirm landing-page fit, and check policy and trademark concerns. The tool does not strip trademarks, normalize capitalization, add Google Ads match-type syntax, quote phrases, or expand synonyms — those edits belong in your destination workflow. Demand validation still needs an appropriate keyword-data source; the combination step is upstream of that, and the boundary should stay clean between "list of candidates" and "list worth bidding on."