The Niche-Down Field Guide

How to take a proven, high-frequency app category and re-cut it for a specific person, life stage, or market — with the evidence for what works and what doesn't.

Research compiled 2026-08-11/12 · Sources: AppKittie (US App Store estimates), Apple's public storefront charts for 23 markets, and the Meta Ad Library. All revenue and download figures are estimated monthly and cover App Store IAP only — a limitation that matters enormously and is discussed at the end. §7 is a worked example applying the whole guide to FiftyStrong's own category.

---

The four rules

Everything below reduces to these. If you read nothing else, read this.

1. Build the decider, not the logger.

Same behaviour, same category, both excellent apps — roughly ten times the revenue per user for the one that decides rather than records. Logging is a commodity that trends to free. Judgement is what people pay for, and it is also what's hard, which is what makes it defensible.

2. Daily cadence is necessary but not sufficient — you also need anxiety or identity.

Baby feeding is the highest-cadence behaviour in this study (8–12 times a day) and the generic trackers earn $0. Islamic prayer is five times daily and earns $0.73/DL. Meanwhile baby sleep advice earns $8.75/DL. Cadence creates the opportunity for retention; fear or identity creates the willingness to pay. You need both.

3. Ask whether the niche changes the product, or only the copy.

Solid Starts changes the entire content — which food, what age, cut which way — and earns 12.50/DL * *.Vitachangestheinterface—largetype, notimers—andearnsupto * *6.65/DL for solitaire. The 807 menopause symptom trackers change only the label, and the best of them earns $3,783/month. If the underlying app would work identically for someone outside your niche, you have a positioning exercise, not a product.

4. A niche pays in proportion to where its population sits.

Christian devotional apps earn 4.50–10.02/DL. Islamic prayer apps earn $0.73/DL on the same behaviour at higher cadence — because the audience concentrates in Indonesia, Pakistan and Egypt rather than the US. The same logic governs geography: expat-majority markets never localize (Singapore and the UAE are ~14/15 Western apps), because the paying population is foreign.

---

1. What the move actually is

FiftyStrong is not "a workout app." It is:

Take a behaviour people already do on a repeating cadence, in a category that already has proven willingness to pay, and re-cut it for one specific person or one specific occasion.

Three parts, all required: a repeating behaviour (makes retention and subscription possible), a category that already monetises (you're re-aiming demand, not creating it), and a wedge nobody owns (an audience, a condition, an occasion, a constraint, or a market).

---

2. The pattern is live right now

Every one of these launched in 2025 or later and is a re-cut of a behaviour that already had a generic incumbent.

MeAgain is the cleanest lesson. A quarter of MenuFit's downloads, nearly the same revenue — because someone managing a $1,000/month prescription values a companion app differently from a general dieter. Narrower audience, higher price, less competition.

Sweatmates is the most instructive wedge, because it isn't demographic at all. "Do your workout with your partner" is an occasion. Occasion niches are underexplored relative to demographic ones.

How hard is it, honestly. Filter Health & Fitness to launched since Jan 2025 and above $30k/month and the entire field is 82 apps out of 92,070. The play works, and it is not a lottery — but roughly 1 in 1,000 new entrants clears a solo-founder-relevant line. Pick well.

---

3. The economics — revenue per download

This, not monthly revenue, is what decides whether a solo founder can win. Below $1.50/DL you need paid UA at scale. Above $4/DL you can be small and fine.

Note that revenue per download is not margin. MacroFactor's $30.01 reflects an expensive product with real content costs. High $/DL usually means high cost to deliver — which is precisely the FiftyStrong constraint, restated.

---

4. Where to hunt — the category scoreboard

"Top new entrant" = highest-revenue app launched since Jan 2025.

Two counter-intuitive ones. Medical is a trap — it feels serious and defensible, and it has a $31k/month ceiling for new entrants. Lifestyle's big numbers are the wrong kind of money — AI hairstyle try-on at $320k/month is a hard-paywall, single-session, paid-UA machine. That's a media-buying business, not a product that compounds.

The daily-cadence behaviour map

Conspicuously absent from the winners: habit trackers, water reminders, sleep trackers, budgeting. Daily, but free-tool commodities. Rule 2 again.

---

5. Axis: persona and culture

Parenting is the strongest persona category found — better $/DL than fitness. Solid Starts $1.01M at $12.50/DL, Huckleberry $807k at $8.00, Napper $700k at $8.75, The Wonder Weeks $500k at $8.33.

I nearly missed this, and how is worth knowing. A title-scoped search for baby tracker feeding sleep newborn toddler returned 5,595 apps with a $20k/month ceiling and dozens of zero-revenue clones — it looked like a dead free-utility space. Searching the bare noun baby surfaced all four leaders. The winners name themselves after the answer ("Napper", "The Wonder Weeks"); the clones name themselves after the mechanism ("Baby Feeding Tracker"). A descriptive-phrase search returns precisely the losers.

Faith is the cleanest natural experiment for the culture axis. Prayer is daily and identity-driven in both traditions; Islamic practice has higher cadence.

Muslim Pro has 570,662 monthly downloads — one of the highest counts anywhere in this study — and earns $0.73 per download. Not devotion: geography and norms. Rule 4.

Two asides: faith apps live in Reference, not Health & Fitness — a much less crowded shelf. And Bible Note: Sermon Notes (Mar 2025, $208k/month) is a textbook recent niche-down: note-taking → sermon notes.

---

6. Axis: age and life stage

Seniors is the purest age-niche in the study. Vita Studio took the most commoditized content in software — casual puzzle games — and shipped seven re-cut for one age group: Word Search $154,685, Color Sort $129,434, Math Puzzle $68,777, Solitaire $45,007, Spider $43,796, Sudoku $33,215, FreeCell 29, 586. * * 504,500/month.** A rival sells solitaire at $6.65/DL. What changed: large type, high contrast, no timers, no punishing mechanics. The content is identical; the interface is the niche. And being unfashionable is the moat — nobody ambitious wants to build large-type solitaire.

Menopause matters directly to FiftyStrong, and the naive reading is wrong. 807 apps, top earner 3, 783/month, alongtailofsolo − devsymptomtrackersatliterallyzero.ButReverseHealthclaims * *1.1M + users * *andshowsonly 12k/month of App Store IAP, because its business is a web quiz funnel.

Menopause monetises through funnels, not app stores. The empty shelf is not absent demand — it's evidence the winners never listed there.

That is the strongest external validation FiftyStrong's model has, and it means those 807 zero-revenue apps aren't your competition and tell you nothing about your odds.

Anxiety, not age, is the underlying variable. Ranking life stages by revenue reproduces almost exactly the ranking by how frightened the buyer is: new parents (highest, sustained 24 months, captive cadence, two payers) → trying to conceive → seniors → menopause → general adult fitness (aspiration, not fear; priced accordingly). Age is a proxy for a problem. It works when the age reliably predicts an acute problem the person is already searching for.

---

7. Worked example: the menopause / perimenopause workout market

This is the whole playbook applied to FiftyStrong's own category, and it produces the sharpest illustration of Rule 4 in the document.

The App Store side is essentially empty

Only 159 apps have "menopause" in the title at all, and the highest-download one manages 11,229 a month. Almost every one is a symptom tracker, an HRT tracker, or a clinical service. Filtering to purpose-built menopause or 40+ workout apps:

Note the trap in searching this: sort a menopause+workout query by revenue and the top results are The Body Coach (100k), LESMILLS + (100k) and Glo ($100k) — general fitness apps that merely mention menopause in their description. They are not menopause products, and mistaking them for competitors would badly misread the market.

The advertising side is enormous

Who is actually buying that inventory — none of them app companies:

• Quiz funnels — Emma Collins Active Wellness: "Get Your Personalized Menopause Training Plan". This is FiftyStrong's exact model, already running.
• 8- and 12-week programmes — EvoFit "A Plan Based On Your Biology · 12-Week Menopause Reset"; Women of Strength Gym "The 12-Week Strength System Built For Hormonal Changes"; Temple Fitness "Strength Sisters 8-Week Menopause Kickstart".
• Creator-coaches — Briana Prescott running five variants at once ("I Finally Fixed My Bat Wings", "I Had Muscle. Not Confidence."), Cristina Smith on brain fog ("Read this if you blank on words"), Lucy Chapman on GLP-1 muscle retention.
• 1:1 coaching positioned explicitly against apps — B4L Coaching: "Built for perimenopause, not against it · Real 1:1 coaching, not another app".
• Plus supplements, skincare, gut-health and hormone clinics, and V Shred's "Stalled metabolism? Take this 60 sec quiz".

The gap between $6,900/month of apps and 14,344 live ads is the entire finding. The App Store category is empty because it is the wrong container, not because there is no demand.

Menovation is your closest competitor, and the most useful data point available

Released September 2025, so roughly eleven months old:

• $3,783/month on 1,689 downloads, 235 reviews at 4.91★
• $28.99/month · $69.99/quarter · $199.99/year
• Growing hard: revenue +41% in 30 days, +56% in 90; MAU +273%, DAU +270% over 30 days
• organic_sources: [instagram] — growing organically, with no Meta ads detected
• Positioning: "strength, pelvic health, bone density improvement, balance"
• Content: 200+ workouts — 58 follow-along strength, 50 self-paced with exercise swaps, 36 Pilates, 33 Yoga, 18 Barre, 5 Yoga Nidra, 5 Core Basics — plus 160+ recipes, 30 nutrition lessons, 87 expert tips and 11 interviews with doctors and pelvic-health physiotherapists

Four implications, in order of how much they should change what you do:

1. 18/monthisunder − priced. * *Menovationholds4.91★at * *28.99/month and $199.99/year — 61% and 92% above FiftyStrong's current paywall. The CAC arithmetic in the FiftyStrong decision doc was built on $18; at $29 the break-even drops from ~1.9 months to ~1.2, and the kill criterion should move with it. This is the single most actionable finding here.

2. The positioning is confirmed, and it is not unique. Pelvic health + bone density + balance + strength is exactly the right frame — which you know, because the one app in the category that is growing chose the same words. Differentiation will have to come from delivery and voice, not from the promise.

3. The content bar is now quantified. 200+ workouts across five modalities, 160 recipes, 11 expert interviews. That is the non-code cost from §10, made concrete — and it is the real barrier to entry, exactly as this guide keeps arguing.

4. Paid acquisition is wide open. Menovation grows on organic Instagram; nobody in this app category runs Meta ads at meaningful scale. FiftyStrong's validated $0.36 cost-per-landing-page-view and its working creative system are precisely the asset none of these competitors have.

The strategic read: you are not competing with apps. You are competing with human coaches who have faces, Instagram followings, and 1:1 relationships — and who are already positioning against apps as a selling point. That is simultaneously the strongest validation of the web-funnel model (Rule 4 again: the money follows the audience's habits, not the store) and the clearest statement of what you would need to beat.

---

Why coaches dominate the ads — it is economics, not preference

The obvious reading of §7 is that this audience prefers human coaching to apps. 50 live reviews of Menovation say otherwise, and the correction matters because it changes what to build.

These women love apps. Menovation holds 4.91★ and the reviews are specific and enthusiastic. One is decisive on the point:

"It feels like I have a personal trainer with me when I do the workouts. I used to actually go to a personal trainer whom I loved, but life changes and I needed something I could do at home."

That is migration toward the app, from real 1:1 coaching. What they want is the feeling of being coached, not the delivery mechanism.

So why is the ad inventory all coaches? Because of LTV, not taste. A 12-week coaching programme at $600–2,000 supports a $200–400 CAC. An app at $29/month supports perhaps $60–70. Coaches simply outbid apps in the same auction. You would observe exactly the pattern in §7 — 111 coach ads, essentially no app ads — with zero difference in what anyone prefers. And B4L Coaching's "Real 1:1 coaching, not another app" is objection-handling aimed at the app-fatigued minority who respond to coaching ads; it is not a description of the market.

What the reviews say the job actually is, in rough order of frequency:

1. Don't make me plan. "I love following her program and not having to plan what I need to do" · "quickly start and not need to think about what I am going to do."
2. A named human to trust. "Ashley" appears by name in a large share of reviews. Not a brand — a person.
3. Adapt to my body today. "easy to tailor to my abilities" · "offers encouragement when you need rest days."
4. Explain why, in small doses. "digest it in bite sized doses."
5. Don't sell me weight loss. "Great workouts if you are tired of the 'lift to lose weight' BS and want to lift to get strong." — FiftyStrong's exact positioning, in a customer's own words.

The structural correction: Menovation is a cohort migration, not an acquisition business

The reviews reveal how it actually grew. Its creator, Ashley, ran SLAM (Strong Like A Mother) and Get Mom Strong for years; as that audience aged into perimenopause she launched Menovation and moved them across:

"I've been a subscriber to SLAM for 3 years. I'm going to switch…" · "switched to Menovation from Slam in October of 2025" · "I did Ashley's SLAM for years… It's like SLAM grew up."

Three consequences:

• Menovation's $3,783/month says nothing about cold-acquisition economics in this niche. It is a pre-sold audience being re-housed, at a price a three-year relationship supports.
• The paid-acquisition lane is therefore even emptier than §7 suggested — nobody has proven it works here, in either direction.
• But FiftyStrong has no Ashley. The non-code cost is now specified precisely: it is not only video production, it is a person with a face and a relationship. That is the real barrier, and it is a hiring or partnership problem, not an engineering one.

Two warnings from the critical reviews

• 1★: "there is a lot of good content, the price is way too high! I can't really dig into anything because I can't afford it." Price resistance at $28.99 is real — test the higher price, don't assume it.
• 3★: "it requires more planning than I'd like to do… Other apps are better at being able to just press start and having a plan in place. I also don't find the videos to be as phone friendly." The failure mode is friction, not content quality. Press-start-and-go beats a library. That is a direct product spec.

Evidence base: 50 most recent US reviews, one app, rating-skewed 4.91★ and self-selected. Directional, not representative.

---

8. Axis: geography and localization

Tooling first — AppKittie can't do this, and Apple can

get_supported_countries lists JP, KR, CN, TW, HK, SG, but the data behind them is largely absent. And search_apps's countries parameter filters on an app's primary language, not the market it succeeds in. My first pass used that filter, sampled China and SE Asia, skipped Japan/Korea/ Taiwan, and concluded "geography loses." That was wrong — see the corrections log.

Use Apple's free feed instead:

https://itunes.apple.com/{cc}/rss/topgrossingapplications/limit=25/genre={id}/json

cc = any storefront · genre = 6013 Health & Fitness, 6017 Education, 6023 Food & Drink, 6015 Finance. Also topfreeapplications / toppaidapplications. No key, no credits, every market. It gives rank order only — pair it with AppKittie for sizing.

Localization resistance varies enormously

Share of Top Grossing Health & Fitness top 15 held by local or regional apps:

Korea and Japan are closed to exactly the degree that makes them worth entering. MyFitnessPal is #1 grossing in both the US and UK and charts in neither.

Europe is pan-European, not per-country. US apps get displaced, but by European companies operating continent-wide: Yazio (DE) is #1 in Germany and Turkey plus top-4 in FR/IT/ES/NL; komoot charts in six markets; WeWard (FR) is #1 in France and Italy. The barrier is competitive, not linguistic — lower opportunity than Asia. Poland is the exception, with three genuinely Polish apps in the top 10 and a hard language barrier.

LatAm has national champions: Fitia (Peru) MX #3 · silBe (Colombia) MX #9 · Befit BR #3 · and Monju: Emagrecer com GLP-1 at BR #6 — MeAgain's play, localized into Portuguese.

The Gulf is the sharpest experiment here. Saudi: local Kalee at #1, MyFitnessPal pushed to #2, four local apps in the top 10. The UAE, adjacent, same language: entirely Western. The difference is population composition — check who actually lives there before researching anything else.

The finding that survived all 23 markets

Calorie counting is localized everywhere. Sixteen markets, sixteen different answers:

US MyFitnessPal · GB Calorie Counter + ($1M/mo, $25/DL) · DE Yazio · FR Foodvisor · SE Lifesum · PL Fitatu · ES & MX Fitia · BR Befit/Monju · SA Kalee, Mezan, Miran · TR EatBetter · TH Kalguroo, NubCal · VN Nutree, Dietfit, CalSnap · JP あすけん, カロミル · KR 필라이즈, 칼로그 · TW WiseMeal · HK CaLoMei

A food database is the most locale-bound data structure in consumer software — the single most robust finding of this whole programme.

The operators already running this play

Learna AI is the literal answer to "US winner → localize" — top-5 in US Education and charting in HK, JP, TW. Gymwork ships EN/JA/KO from one binary, ~$100k/month, with a price ladder from $138.99/yr down to a $22.99/yr "eligible only" tier — per-market price discrimination is part of the play.

CaLoMei — the template and the reality check

A Hong Kong solo developer, released ~3 March 2026, #1 top grossing Health & Fitness in Hong Kong five months later. The entire moat, from its own listing: 香港地道食物資料庫 (a HK food database — cha chaan teng, cart noodles, roast meat rice) and 每次掃描完都會用廣東話寸你一句 (after every scan the AI roasts you in Cantonese). Everything else is Cal AI.

Then the number: ~1,000 downloads and **~6, 000lastmonth * *, MAU1, 408, DAU317.Pricesraisedtwice(4.99→$9.99 monthly).

Being the #1 grossing health app in Hong Kong is roughly a $6k/month business.

Real solo-founder income in five months from a food database and a joke — and nowhere near a mid-chart US app. Rank is not revenue. Prize ordering: Japan (~125M, top-3 global market) > Korea (~52M) > Taiwan (~23M) > Hong Kong (~7.5M).

Solo devs holding chart positions: CaLoMei (HK #1) · 寶寶輔食日記 baby weaning (TW F&D #7) · 菜單貓 (TW #10) · Sipor (TW #8) · 냠냠일기 baby weaning (KR #11) · CalSnap (VN #9) · Bitalyze (PH #9).

---

9. The candidate shortlist

Hypotheses, not validated niches — §9 is how you'd falsify them.

Tier 1

1. A local-food-database calorie tracker. The only niche proven to work in every market anyone has tested. Either cuisine-for-diaspora sold in the West (high iOS share, Western ARPU), or a market-local product. Reference implementation: CaLoMei. Reality check: CaLoMei is ~$6k/month.

2. Baby 0–2, at the deciding layer. Best $/DL of any life stage. The logging layer is worthless; the "what should I do about it" layer is worth $500k–1M/month. Caveat: the cultural version is already running in Taiwan and Korea via solo developers.

3. Spoken English × a professional occasion. 8–9 of the top 12 grossing Education apps in every CJK market are English-learning — the most repeatedly-confirmed wedge found. Duolingo and Speak own the general slot; medical English for nurses, accent work for Cantonese/Mandarin speakers, and aviation English are still won by small teams.

Tier 2

4. Seniors, via interface rather than content. ~$504,500/month across seven puzzle games says the accessibility gap is real and repeatedly monetisable. Wider than games: any daily behaviour with a hostile default interface.

5. Strength programming × a persona. Ladder does $6.5M/month at $20.06/DL and charts nowhere in Asia; Korea's chart is all loggers. A Ladder-grade programming product in Traditional Chinese has no visible incumbent in TW or HK.

6. Meal planning × a household constraint. mise proved the base at $482k/month in two months. Halal, coeliac, renal/diabetic, "one dinner my toddler and I can both eat."

7. GLP-1 adjacency. MeAgain $736k at $8.76/DL, with a localization proof in Monju at BR #6. Highest ceiling, most contested, expect funded competition.

Demoted

8. Strength training × men 50+. Previously Tier 1. The life-stage sweep found no proven entrant anywhere — ambiguous, not encouraging. Menopause has an acute physiological trigger that sends women searching; "being 50 and male" does not. Still cheap to test since it reuses your FiftyStrong machinery, so painted-door it before building.

Do not build: symptom trackers of any kind · anything in Medical or Productivity · general cycle tracking (Flo at $9.1M) · expat-majority markets (Singapore, UAE) · anything globally standardised like wine · the finger-on-the-camera vitals class (Heart Beat, Heartica, Finger Blood Pressure), which earns real money on measurement claims its sensors cannot support.

If you're choosing a market, not just a niche

Your native advantage sits in the smallest market with the weakest economics. Three resolutions, ranked:

A. Taiwan primary, Hong Kong as test bed. Traditional Chinese covers both. TW is 3× HK's population and shows the most solo-dev penetration of any market tested. Validate in HK where your instincts are native and ad costs trivial, then ship the same build into TW. This is the one I'd take. B. Build once in English, localize into 3–4 CJK markets. What EpopSoft, Gymwork and Munice do. Higher ceiling; you compete with them directly. C. Build for the West from Hong Kong. Singapore- and Türkiye-registered studios earn far more selling into the US than any Asian market pays. Highest ceiling, no language moat.

Not exclusive — A is the cheapest way to learn whether you can build a genuinely local product.

---

10. The screen — testing a candidate in ten minutes

1. Is the base behaviour monetised? search_apps on the generic behaviour, sortBy: revenue. You want an incumbent above $50k/month. If nothing monetises, the wedge can't either.

2. Is the niche unoccupied? Title-scoped search on the wedge term — and the bare category noun, per §5. AppKittie slugs freeze the original name, so a slug disagreeing with the current title means a rename into the niche (e.g. app-fortify-strength-training-plan now titled "Fortify Strength for Women 40+"). A generic app repositioning into your niche and then growing is the strongest evidence the niche is real and still open.

3. Are new entrants breaking in? sortBy: growth, growthPeriod: 90d, minRevenue: 30000, releasedAfter: 2025-01-01. If recent launches are climbing, incumbency isn't decisive.

4. Check revenue per download, not revenue. Under $1.50/DL you need paid UA at scale. Over $4/DL you can be small and fine.

5. If the niche is a market, use Apple's feed (§7), not AppKittie — and check population composition before anything else.

6. The judgement no query answers: can you deliver the content? FiftyStrong's binding constraint isn't code, it's producing a credible video library with a credible coach. Every candidate has an equivalent — a cuisine app needs a real food database, a nursing-English app needs real clinical vocabulary. Ask what the non-code cost is before committing, because that's the cost that will actually stop you.

---

11. Caveats

• All figures are third-party estimates covering App Store IAP only. Web-funnel businesses are invisible — which is why Reverse Health looks like a $12k/month company and why the menopause category looks dead. This is the single most important limitation here.
• **/DLisaratiooftwoestimates. * *Sortintobands(<2 / $2–5 / $5–12 / $12+); don't compare $8.00 against $8.75.
• "No proven entrant" is ambiguous. An empty niche can mean untried or unwanted. Men 50+ and the cuisine tracker are both in this category.
• Search phrasing shapes results — see §5. Search the bare noun as well as the phrase.
• Apple's legacy RSS feed gives rank order only, with no revenue and no documented refresh cadence. Relative position is solid; size comes from weaker sources.
• AppKittie systematically under-indexes Asian apps. Treat any figure for an Asia-focused app as a floor. This blindness is also a mild advantage: if standard Western ASO tools can't see these markets, Western competitors aren't systematically hunting there.
• **CaLoMei's ~6, 000couldbeUSDorHKD * *—ifHKD, nearerUS770, and the calibration gets sharper, not softer.
• Growth percentages on very young apps are unreliable, and category counts are search-scoped rather than official Apple totals.

Corrections log

Kept because the failure modes are reusable.