Before submitting Squat Alarm to the App Store, I spent a day on keyword research. I went in looking for a clever niche phrase that a squat-to-stop alarm could own. There isn't one. Almost all the search volume in this category sits in four plain phrases, none of which describe what my app actually does, and the small apps that rank well all handle their subtitle the same way.
The four phrases that matter
I used Appfigures on the free tier, looking at the US App Store for iPhone, on 4 September 2026. Its popularity score runs from 0 to 100, and 5 is the floor. A 5 means "measurable but negligible", not "5% of the top term".
- alarm: 65
- alarm clock: 54
- loud alarm clock: 51
- wake up: 39
Everything else drops off fast. "i can't wake up alarm clock" scores 21. "alarm app" scores 8. "wake up alarm" scores 7. If you are building any kind of alarm, those four phrases are the whole market.
The phrases that describe my app score nothing
Then I checked the phrases I had been quietly counting on. "squat alarm", "fitness alarm", "workout alarm", "heavy sleeper", "alarm clock for heavy sleepers", "no snooze", "stop snoozing", "anti snooze alarm", "wake up challenge". All of them sit at 5.
Competition on most of them is low, which sounds like good news until you ask why it is low. Almost nobody searches for them. Squat Alarm will probably rank near the top for "squat alarm" and "stop snoozing" on launch day, and I am not expecting much traffic to arrive with it.
One tool disagreed. Applyra gives "squat alarm" a medium traffic score and shows it at position three in autocomplete. Both tools agree the competition is low. I am reading the Appfigures 5 as "small compared with alarm clock" rather than zero, and I have a note to check Apple's own Search Popularity numbers in Apple Ads before I trust either one.
What the small apps do differently
Alarmy sits in the top five for every generic phrase. Appfigures estimates it at around a million downloads a month, so that is not a fight I can pick. The interesting apps are the small ones.
Wayk's subtitle is an exact phrase built from the top terms, and it ranks second for "alarm", second for "wake up", and sixth for "alarm clock". An app called "WAKE UP: stop snoozing" is number one for "wake up" on the strength of its title alone, then drops to 61st for "i can't wake up alarm clock" because "alarm clock" is not in its name. Rampit, which Appfigures puts at under five thousand downloads a month, ranks eighth for "squat alarm" and 34th for "i can't wake up alarm clock".
So a small app with an exact high-volume phrase in its title or subtitle lands in the top three for that phrase, while a small app with the same words scattered across the title, the subtitle and the keyword field does not. Apple seems to weight an exact phrase in a visible field far more heavily than one it has to reassemble from parts.
Where Squat Alarm stands
This is the metadata I had already applied before doing the research: the name "Squat Alarm: Stop Snoozing", the subtitle "Wake Up With a Fitness Task", and a keyword field with clock, heavy, sleepers, morning, routine, exercise, challenge, mission, loud, bed, oversleep, workout, and rise.
Scored against the list above, "alarm" is strong because it is in the name, and "wake up" is strong because it is an exact phrase in the subtitle. "alarm clock" only exists as "alarm" in the name plus "clock" in the keywords, which is the weak split version. "loud alarm clock" has the same problem. "snooze" is not indexed as a word at all; "snoozing" is in the name, and I do not know whether Apple's stemming bridges that gap. There is also no Spanish metadata, even though "alarma" scores 49 in the US store with lower competition than "alarm".
The decision I have not made yet
The obvious move is the Wayk pattern: change the subtitle to something like "Loud Alarm Clock to Wake Up", which puts "alarm clock", "loud alarm clock" and "wake up" into one field as exact phrases. That frees "clock" and "loud" from the keyword field, and the space pays for "snooze", "anti" and "cant".
I am hesitating for one reason. "Wake Up With a Fitness Task" describes the product. "Loud Alarm Clock to Wake Up" matches what people type into the search box, which is a different job. I wrote about that trade-off in The smallest useful promise wins, and this is the same tension with real numbers attached to it. A middle option is "Alarm Clock for Heavy Sleepers", which keeps "alarm clock" exact and matches the first line of the description.
The Spanish listing is an easier yes. The app already ships a Mexican Spanish interface, so a Spanish subtitle and keyword field is a metadata task rather than a product change. The one rule I have written down for it: write the Spanish terms for Spanish search, do not translate the English ones.
What will not help
Two things I would have happily spent time on a month ago. Apple does not index the description, so editing it only affects conversion after someone has already found the page. And piling on more niche keywords earns nothing, because every fitness-alarm phrase is already at the floor.
There is a limit to what any of this buys. Metadata only makes an app eligible for a phrase. Whether it actually ranks comes down to downloads, ratings and conversion after launch. A top-ten spot for "alarm clock" is not something a new app from one person can plan for on day one, and I would be fooling myself to build the launch around it.
Questions
Does a niche keyword ever work for a small app? It works for ranking. It does not work for volume. Squat Alarm should sit near the top for "squat alarm" from day one because nobody else wants the phrase, and almost nobody searches it either.
Why not just add the big phrases to the keyword field? Because the split version is weak. "alarm" in the name plus "clock" in the keywords is not the same as "alarm clock" in the subtitle, going by how the small competitors rank.
What would change your mind on the subtitle? Apple's own Search Popularity data for "squat alarm". If Apple shows real volume where Appfigures shows the floor, keeping the fitness-task subtitle becomes the safer choice.