How to clean a keyword list

1Paste your list, one keyword per line. The cleaned version shows up next to it as you change the options. Everything happens in your browser, so a 50,000-line export from a keyword tool is fine.
2Choose how to clean the text: lowercase, punctuation and accents. That always runs first.
3List any words to remove. They come out before duplicates do, which matters. Take "korvane" out of "korvane bypass pruners" and what's left is a copy of "bypass pruners", so it goes too.
4Choose which duplicates go. Exact duplicates are the same keyword once case and spacing are cleaned up. Plurals only count when both forms are in your list, so "pruners" folds into "pruner" but "shoes" never turns into "sho". Word-order duplicates have the same words in a different order, like "loppers heavy duty" and "heavy duty loppers", and the first one you pasted is the one kept.
5Decide whether to keep both forms. In an exact match campaign you might want each spelling as its own keyword. For a research list or your backend search terms, one is enough.
6Sort the list and add match types if you want them, then copy it or download the CSV.

FAQ

Remove brand names and stop words

List the words you want gone, one per line or separated by commas. You can delete just the word and keep the rest of the keyword, or drop every keyword that contains it. Matching is whole-word and ignores case, so removing "saw" leaves "sawdust" alone, and a two-word entry like "hand pruner" only matches as a phrase.

The word frequency tab is the quickest way to find what to remove. Brand names and off-topic words like "kitchen" show up near the top, and one click adds them to the remove list.

Combine lists into longer keywords

Put modifiers in one list and product words in another, and the combine tab writes out every pairing in order: list A times list B, times list C if you fill it in. It stops at 100,000 keywords. Send the result to the first tab to clean it like any other list.

Turn the list into campaigns

The match type option writes each keyword as broad, "phrase" in quotes or [exact] in brackets, and negatives with a minus. That notation is a reading aid other keyword tools use too. Amazon takes the match type as its own setting, which is why the CSV has the keyword and the match type in separate columns, with character and byte counts for each.

If you use Merch Jar, the campaign creation skill builds a structure from your ASINs and can use a list like this as the keywords for its research and exact campaigns. Everything it creates starts paused, and you decide what to turn on.

Already running ads? The search term report analyzer shows which searches spent money without an order, and the n-gram analyzer adds up spend by word, the same idea as word frequency with your own numbers attached.

Is my list uploaded anywhere?

No. The processor runs in your browser and makes no network requests with your keywords. Close the tab and it's gone.

Can I count characters and bytes?

Yes. The totals under the cleaned list show both, and the CSV has them for every keyword. To check a whole listing against Amazon's limits, use the listing character counter.

Why did a keyword disappear?

Open "See what was removed" under the cleaned list. Each removed keyword shows which one it duplicated and why.

Amazon keyword processor

Want a next step that fits your account?

Ask your own AI to do it.

Campaign creation