i know you've seen everyone's agents build mobile apps on X.
literally every day, someone posts: “launched on iOS app store in 24h" then two weeks later: “i hit $4k MRR” and the replies are always the same:
“what did you build it with?” “how did you get users?” “drop the sauce”
after seeing enough of these, you eventually start wondering:
“should i be building mobile apps too?”
short answer: F*CK YEAH.
especially if you’ve spent the last few years building SaaS.
and especially if you’ve convinced yourself consumer apps aren’t for you because you’re a “b2b person.”
because there’s a decent chance you’ve already spent years selling to consumers. you just happened to put a SaaS dashboard in front of them.
here’s why i think that distinction matters, and exactly how i’d go from nothing to a mobile app worth publishing as quickly as possible.
you might already be doing b2c
there’s a pretty standard piece of startup advice that gets repeated everywhere:
- sell to businesses
- businesses have money
- b2b customers stick around longer
- consumers don’t want to pay
sounds reasonable.
until your “b2b SaaS” is a $19/month analytics tool sold to solo founders.
you didn’t escape b2c.
you just picked one of the hardest groups of consumers imaginable.
indie hackers and small founders are incredibly price-sensitive. they understand how software works, they compare everything, and they’ll happily spend three hours looking for an open-source alternative to avoid paying you $20/month.
and half of them are thinking:
“i could probably build this myself.”
putting a Stripe subscription on something doesn’t magically make it b2b.
the more useful distinction is why somebody buys.
a company usually buys software because there’s an economic reason to do it. it saves employees time, reduces costs, increases revenue, replaces another tool or makes some process easier.
consumers buy for a completely different set of reasons.
they want to sleep better.
look better.
save more money.
stop wasting time.
get stronger.
eat better.
feel more organized.
learn something.
quit something.
feel less anxious.
become more confident.
or simply feel like they’re making progress.
and these problems are enormous because basically everyone has them.
you also don’t need to convince a procurement department, integrate into somebody’s 14-tool stack or explain ROI on a sales call.
you need to make one person look at your product and think:
“wait, i want this.”
that’s a very different playground.
and right now, mobile apps are one of the easiest ways to play in it.
why mobile is suddenly interesting again
a few things are happening at exactly the same time.
1. ai has destroyed a huge part of the technical barrier
building a decent mobile app used to mean learning Swift or Kotlin, understanding a completely different ecosystem, fighting Xcode, figuring out app architecture and probably spending months before you had something worth showing people.
that’s changing extremely quickly.
a focused consumer app can now go from an idea in your head to something usable in a day.
the bottleneck is increasingly not:
“can i build this?”
it’s:
“should i build this?”
which is a much more interesting problem.
2. consumer distribution is everywhere
TikTok, Reels and Shorts can put a completely unknown product in front of millions of people without that company having an existing audience.
you don’t necessarily need SEO.
you don’t necessarily need paid ads.
you don’t necessarily need 50,000 Twitter followers.
one good piece of content can create the first few hundred or thousand users you need to figure out whether there’s something there.
3. mobile fits perfectly into that distribution
see video.
understand problem.
download app.
try product.
that entire journey can happen in a few minutes, on the same device.
there’s almost no context switching.
4. you can test ideas insanely fast
this might be the biggest change.
if building an MVP takes three months, choosing the idea feels incredibly important.
if building an MVP takes a day or two, the economics completely change.
you don’t need to discover the idea.
you need to find something interesting enough to test, build the smallest version that proves the core behavior, put it in front of people and see what happens.
if nobody cares, you learned something.
if people use it once but never come back, you learned something else.
if 100 people download it and 25 are still opening it a week later, now things get interesting.
so here’s exactly how i’d approach it.
FIRST STEP
1) find demand before finding an idea
don’t open a blank Notion page and brainstorm “startup ideas.”
you’ll probably come up with solutions to problems that exist mostly inside your own head.
instead, start by watching people.
for consumer products, one of the best places to do that is TikTok.
download it.
spend 5–10 mins/day intentionally looking for patterns.
not random viral videos. human behavior.
look for:
- things people constantly complain about
- habits they’re trying to quit
- things they’re insecure about
- things they flex about
- things they obsessively track
- new aesthetics and identities
- challenges everyone suddenly starts doing
- things people wish they were better at
- routines people keep sharing
- things people repeatedly ask for help with
- behaviors that already require some annoying workaround
you’re basically looking for human problems hiding underneath trends.
for example:
there’s a trend called “underconsumption core.”
the surface-level trend is people buying less stuff.
the obvious founder-brain reaction would be:
“let’s build an underconsumption app.”
don’t.
instead, ask why millions of people relate to it.
maybe the actual problems are:
“i impulse-buy when i’m stressed.”
“i keep buying stuff i don’t need.”
“saving money feels boring.”
“i have no idea where my money disappears every month.”
“i want to feel rewarded for not buying something.”
those are much more interesting.
now you can start imagining actual product loops.
maybe every time you resist buying something, you add it to the app and your “money saved” counter increases.
maybe you photograph something you’re about to buy and the app makes you wait 24 hours.
maybe friends compete over who avoided the most unnecessary purchases this month.
the trend gave you the signal.
the underlying behavior gives you the product.
the 3 consumer app formats i keep coming back to
you don’t need to invent some completely new category either.
most interesting consumer apps fit into a few basic structures.
tracker
turn invisible behavior into numbers.
spending. screen time. sleep. habits. mood. food. focus. workouts. sobriety. reading. studying.
people love seeing themselves quantified because something abstract suddenly becomes progress they can see.
“i’ve been focusing more lately” is vague.
“my average focus time went from 41 minutes to 76 minutes” feels real.
coach
help somebody become a slightly different version of themselves.
daily missions. challenges. plans. reminders. personalized recommendations. feedback.
people often don’t need another complicated tool with 40 buttons.
they need something that understands their goal and tells them:
“do this next.”
the product becomes valuable because it removes decisions.
simple utility
take one annoying thing and make it pleasant.
timers. lists. journals. notes. widgets. planners. calculators. scanners.
the functionality can be ridiculously simple if the experience feels good enough.
a product doesn’t need 25 features to deserve a place on somebody’s home screen.
sometimes one feature used every day is much stronger.
steal ideas from the comments
one more trick:
when you find a trend, search the comments for words like “app.”
people literally write:
“someone needs to make an app for this”
or:
“is there an app that does this?”
or:
“i wish something tracked this automatically.”
that’s basically free product research.
and it’s much more useful than asking people:
“would you use an app that does X?”
because they’re already expressing the problem without you putting the idea in their head.
2) define the loop before you build anything
before touching code, answer one simple question:
what does somebody repeatedly do inside this app?
not what features does it have.
what’s the loop?
for a spending app it might be:
almost buy something → log it → resist purchase → see money saved → feel progress → repeat
for a fitness app:
open app → receive today’s workout → complete it → see progress → come back tomorrow
for a focus app:
choose task → start timer → finish session → build streak → repeat
if you can’t explain the core loop in one sentence, the app is probably still too complicated.
then ask yourself four more questions:
what makes someone download it?
there should be a very obvious promise.
what makes someone understand it in 10 seconds?
the value shouldn’t require a tutorial.
what gives them their first win?
get them there as quickly as possible.
what makes them open it tomorrow?
this is the one founders often forget.
downloads are nice.
retention is the product.
you don’t need perfect answers yet. you just need enough clarity that you’re not asking an AI to invent your entire business while it writes the code.
3) borrow the patterns, not the pixels
once you have an idea and a basic loop, don’t design everything from scratch.
you’re probably not a product designer.
neither am i.
instead, find 5–10 successful apps around the same problem or user behavior.
they don’t even need to be direct competitors.
if you’re building a savings app, maybe one app has incredible onboarding, another has a great streak system, another has a satisfying progress screen and another has a paywall you like.
download them.
actually use them.
then screenshot everything:
- first launch
- signup
- onboarding
- home screen
- navigation
- core action
- empty states
- progress screens
- streaks
- notifications
- upgrade prompts
- paywall
- settings
the useful part isn’t the colors or rounded corners.
it’s the decisions underneath them.
where do they ask questions?
how many onboarding screens are there?
when do they show the actual product?
when do they ask for notification permissions?
when do they ask for money?
how quickly do you get your first win?
what information is permanently visible?
what gets hidden?
what makes you come back tomorrow?
these companies have already tested thousands of tiny decisions you’d otherwise be guessing about.
so don’t invent every interaction from zero.
study what works, understand why it works, combine the best patterns and add your own touch.
for younger consumer audiences, i generally like:
- one obvious action per screen
- huge typography
- very little text
- visible progress
- streaks
- milestones
- satisfying numbers
- personalization early
- very obvious feedback when something is completed
basically:
make progress impossible to miss.
if somebody completes something, celebrate it.
if they’ve used the app for seven days, show them.
if they improved 18%, show them.
if they saved $143, make that number impossible to ignore.
the user should constantly understand:
“this is working.”
4) turn the references into an actual app
this is where building gets stupidly easy.
i’ve tried most of the tools people use for building software with ai.
but if i want to go from an idea to an actual mobile app quickly, i use Shipper.
at this point you should already have:
- your app idea
- your core user
- the outcome you’re promising
- your main product loop
- screenshots of apps that solve similar problems well
take all of that and give it to ChatGPT, Claude or Grok first.
don’t say:
“build me a budgeting app.”
you’re giving the model almost nothing to work with.
instead, give it a proper job:
“i’m building a mobile app that helps {user} achieve {outcome}. study the attached references and break down the ux patterns, visual hierarchy, onboarding, navigation and interactions they use. redesign those patterns around my product. define every mvp screen, what happens on each screen, the full onboarding journey, main navigation, core user loop and one mechanism that gives users a reason to return regularly. remove anything that isn't necessary for the first version. finally, turn everything into a detailed build prompt.”
now you have something much closer to a product spec than a random prompt.
read it.
remove the stupid stuff.
add anything it missed.
then take that output, attach your screenshots and put everything into Shipper.
tell it exactly what you want.
the screens.
the interactions.
the flow.
the logic.
the tiny details.
then keep talking to it like you would a developer sitting next to you:
“make this screen simpler.”
“move the paywall after the user gets their first result.”
“add a 7-day streak here.”
“this onboarding feels too long. cut it in half.”
“save this state when the user closes the app.”
“make this interaction feel more native to iOS.”
“there are too many choices on this screen. make one action dominant.”
this is where people get AI builders wrong.
you don’t need to know Swift.
you don’t need to manually create every component.
you don’t need to set up a giant dev environment just to find out whether anybody wants your idea.
but you still need taste.
you still need to make decisions.
you’re basically directing the product while Shipper builds it.
and the quality of the result depends heavily on the quality of those decisions.
the biggest difference is how you talk to the ai.
don’t just tell it to “make an app.”
keep forcing it to think about the user:
- what do they see first?
- what do they need to understand here?
- what is the single most important action?
- how quickly do they experience the main benefit?
- where could they get confused?
- what information can we remove?
- what makes this satisfying?
- what gives them a reason to open it tomorrow?
that usually produces something much better than endlessly prompting for new features.
5) make the first version good enough to charge for
once the main experience works, stop adding random functionality.
focus on three things.
onboarding
treat onboarding like its own product.
because for most users, it basically is.
they haven’t experienced your product yet. they have no loyalty to you. they can close the app in two seconds and never think about it again.
find good onboarding flows, screenshot them and repeat the same reference process.
your goal is simple:
the user should understand why they downloaded the app and experience value within 30 seconds.
every onboarding screen should earn its existence.
if you ask a question, use the answer.
if you ask for a permission, explain why.
if something can wait until later, move it later.
and if you have eight onboarding screens because every other consumer app has eight onboarding screens, you’re doing it wrong.
then give those references to Shipper and iterate until the whole thing feels obvious.
monetization
once the product works, add your subscription and paywall.
don’t spend three days debating whether your yearly plan should cost $27.99 or $31.99.
you have no data yet.
a perfectly normal place to start might be:
- $4.99/week
- $29.99/year
you can test pricing later.
what matters more initially is when you ask for money.
if possible, let the user understand the value first.
let them create something.
see a result.
finish their first session.
get their first personalized plan.
then put the paywall in the path of continuing that value.
you want the user thinking:
“i want more of this.”
not:
“what the hell am i paying for?”
polish
then use the app.
a lot.
don’t stare at the home screen and decide it looks finished.
actually behave like a user.
start from a fresh account.
tap things in weird orders.
deny permissions.
close the app halfway through onboarding.
open it again.
leave fields blank.
use absurd inputs.
come back the next morning.
give it to friends without explaining anything and watch where they get stuck.
you’ll discover an unbelievable number of tiny things that made perfect sense to you because you built it.
every confusing interaction you remove makes the product feel more trustworthy.
and whenever something feels wrong, go back to Shipper and describe exactly what you want changed.
you’re not trying to make the first version perfect.
you’re trying to make sure the core loop feels finished.
6) ship before you feel ready
once the main loop works:
SHIP IT.
don’t spend another month adding social features, achievements, AI assistants, custom themes and 14 settings because you’re scared of publishing.
the first version isn’t supposed to prove you’re a genius.
it’s supposed to answer one question:
does anyone actually want this?
publish it.
make content about the problem.
send people to it.
watch what they do.
read the reviews.
look at where onboarding drops.
look at how many people actually reach the core action.
look at how many come back the next day.
look at how many come back a week later.
look at who pays.
then fix what’s broken and ship again.
personally, i’d start with iOS and only worry about Android once the idea has earned the extra work.
and i’d keep the first version almost painfully focused.
one audience.
one problem.
one promise.
one core loop.
you can always make the app bigger after people care.
making something smaller after you’ve built 30 features is much harder.
the weird thing about building consumer apps right now is that the technical barrier that stopped most of us from doing it has basically disappeared.
you used to spend most of your energy figuring out how to build the thing.
now you can spend much more of it on the parts that actually make the thing work:
finding a real behavior.
turning it into a product people understand immediately.
giving them a reason to come back.
figuring out distribution.
you can figure out what people want on TikTok, study the apps already winning their attention, turn those patterns into a proper product spec and have Shipper build the thing.
then put it in front of real people.
you don’t need to know whether it’s a $10k/mo app before you start.
you just need to get version one into somebody’s hands.
preferably before the 18-hour timer runs out.
