Long-Tail App Store Keywords That Fit the 100-Character Field

A long-tail App Store keyword is a specific phrase a person types, like “sourdough baking timer,” instead of a single crowded word like “timer.” Specific is good. Pasting the entire phrase into the 100-character keyword field is often a waste, because Apple combines words across the name, the subtitle, and the keyword field. If “sourdough” and “timer” are already in the name, the field only needs the word that is missing.

This is the mistake that makes a careful app look like it has no keywords left. The developer finds ten perfect phrases, drops each one in whole, and also splits them into single words “just in case.” The field fills up with repeats. Apple’s rules on Creating your product page already tell you not to repeat words and not to add plurals of words you included. The combination rule is how you obey that without losing the phrase. Formatting details are in the keyword field guide.

What “long tail” means inside 100 characters

Long tail, on the App Store, is not a 12-word sentence. It is a query more specific than the category, short enough that the missing pieces fit the budget. “Timer” is a generic. “Sourdough baking timer” is a job. You want to be eligible for the job. You do not want to store the job as a single comma-separated blob if two of the three words are already indexed in the name.

Every character counts, including spaces inside a phrase and the commas between terms. There is no space after a comma. “stretch and fold” is 16 characters because of the two spaces. “stretch,and,fold” would be a different, worse string: it spends commas to separate words you might have combined, and it still repeats nothing useful if those words were not the goal. Prefer one phrase with internal spaces only when the words must travel together and are not already elsewhere in the 160.

Apple combines words you already paid for

The indexed budget is the name (30), the subtitle (30), and the keyword field (100). Words in any of those three can combine. That is why the practical rule is: list the unique tokens, not every phrase those tokens can form. If the name contains “sourdough” and “timer,” and the keyword field contains “baking,” you have covered a search for a sourdough baking timer without storing the six-word echo “sourdough baking timer, baking timer, sourdough, timer.”

Apple does not publish a weight for exact phrases versus combined words. Do not invent one, and do not “buy safety” by including both the phrase and the pieces. The safety is fake. The duplicate words are real, and Apple’s own skip list tells you to avoid duplicates. Combination is the reason the skip list is affordable. You are not losing the query. You are refusing to pay for it three times.

The before-and-after for a sourdough timer

The sample app times starter feedings and bulk ferment for people who bake sourdough at home. Name: “Proof: Sourdough Timer” (22 characters). Subtitle: “Feeding and bulk ferment” (24). Tokens already indexed: proof, sourdough, timer, feeding, and, bulk, ferment. The phrase people type, “sourdough baking timer,” is missing only one word: baking.

The before field spends 78 characters to add the word baking and then repeats the name and subtitle. The after field spends all 100 characters on terms a baker actually types.

Put “baking” in the after field if you still need that exact token and it is not in the name. In this version the name already has “sourdough” and “timer,” and the subtitle has the method. “baking” is one extra word you can swap in by cutting a weaker neighbor, not a reason to restore the whole before string. The before string also repeats “proof,” “sourdough,” “timer,” “bulk,” “ferment,” and “feeding,” which Apple tells you not to duplicate.

What to do with the characters you free

The after field is the point of the exercise. Starter, levain, autolyse, stretch and fold, aliquot, crumb, hydration, banneton, cold proof, rye, jar, lid, pan. Those are the words in a baker’s notes that the 30-character name could not hold. A few of them are long-tail phrases with internal spaces: “stretch and fold” and “cold proof.” They earn the space because the pair is the term, and the pieces are not already indexed. How to find those neighbors from autocomplete and reviews, without a paid tool, is in App Store keyword research.

Count the string in the ASO Character Counter after you delete spaces that snuck in after commas. The after example is 100 characters on purpose. If you are at 70, you did not finish. Add another specific term a reviewer used. Do not add “best,” “free,” or “app.” If you are naming the app from scratch so the strongest phrase starts in the 30-character name, do that first. The name is a better home for the head term than the hidden field. See how to name an app.

The screenshot still has to match the phrase

Combination gets you eligible for the query. It does not get you the install. Someone who searches “sourdough baking timer” and lands on a generic clock will leave. Screenshot one should show a feeding timer or a bulk-ferment countdown with the starter named on screen. Screenshot two can show stretch-and-fold intervals if that phrase is in the field. The caption should use the same job the metadata assembled, in plain language.

A new app with no download history is especially exposed here. You do not have behavioral data to lean on, so a mismatched screenshot wastes the only query you bothered to assemble. That constraint is the subject of ASO when you have zero downloads. Brand versus these generic baking terms is a separate budget decision, covered in branded vs generic keywords.

A quick audit you can run on a live field

Export the name, subtitle, and keyword field into three lines. Split the keyword field on commas. For each token, ask whether every word inside it already appears in the name or subtitle. If yes, delete the token. If some words are new, keep only the new words, unless the full phrase is a set term like “stretch and fold.” Then look for a singular and its plural side by side. Apple says not to include both. Keep the form people type.

• Circle any keyword that is a complete copy of the name or subtitle. Delete it.
• Circle any single word that already appears above. Delete it.
• If a phrase is half new, keep the new half and let combination do the rest.
• Refill the opened characters with a neighbor from reviews or autocomplete, then recount.