Web vs. mobile vs. neither

You picked your idea yesterday. Today you decide what shape it should be in.

There are basically three answers to "what shape should my thing be," and most beginners go straight to the wrong one because the wrong one feels the most legitimate.

Door 1, web app

A web app lives at a URL, runs in a browser, works on every device with a browser (which is essentially every device), and gets updated by you pushing a new version. The user does nothing, they just see the new thing on next visit.

This is the right answer for the vast majority of small builds. Web apps are cheap to host (Day 12), easy to distribute (just send a link), and faster to ship than mobile apps by a wide margin. The phone web browsers in 2026 are good enough that a well-built web app feels nearly as native as a real app, with none of the App Store overhead.

Pick this if your build:

• Will be used at a desk or while sitting down.
• Is going to need updates often.
• Does not need the camera, GPS, push notifications, or other phone-specific features in a serious way.
• Will be used by a small audience initially.

Door 2, mobile app

A mobile app lives in the App Store or the Play Store, gets installed on a user's phone, and is updated by the user choosing to update (which most users do not, ever).

Mobile apps are the right answer when:

• The thing fundamentally needs the camera, the GPS, push notifications, or background processing in a way the browser cannot really do well.
• The user will use it often, on the move, and you need it to feel like a first-class part of their phone.
• You are willing to maintain two separate versions (iOS and Android) and pay the App Store taxes (review process, 15 to 30 percent revenue cut, version-update pain).

That is a high bar. If your idea is "an app for X," before you commit, ask yourself whether it could be a really good web app instead. About 70% of the time the answer is yes, and you just saved yourself six months.

Door 3, neither

This is the door beginners forget exists, and it is often the right answer.

Lots of useful things are not apps at all. They are:

• A script. A small piece of code that runs once a day, or runs when you ask it to, and does one specific thing (sends you an email, updates a spreadsheet, posts to Slack). No UI, no users, just an internal tool that does a job.
• A browser extension. Lives inside your browser. Useful when the thing is "I want to change how every webpage I visit looks or behaves."
• A Google Sheet with formulas. Sounds like a joke, often is not. Many viable businesses have run on a Google Sheet for years.
• A spreadsheet plus Zapier or Make. No-code automation can do astonishingly much before you need to write any code.
• A simple form (Tally, Typeform, Google Forms) plus an email sent to you. A real MVP for many ideas is a form that collects requests, which you then fulfill manually.

This door exists because the question is not "what is the most impressive shape my idea could take," it is "what is the smallest shape my idea could take that would still tell me whether I am onto something." Many of the best builds we have seen in this course started in Door 3 and graduated to Door 1 later.

A two-minute exercise

For your idea from yesterday, write down which door you would pick if you had to ship in one week. Then write down which door would tell you fastest whether the idea is actually useful.

If the two answers differ, pick the second one. Day 21 will come back to this when we sequence the build, because the first version of your thing should always be the version that teaches you the most about whether the idea is real.

Why Door 1 is the right default

For most readers of this course, the default is Door 1, a web app. The reasons:

• It works on every device, including your mom's phone and your roommate's Linux laptop.
• It updates the moment you push.
• It costs almost nothing to host until you have real users.
• The AI tools you are using are best at building web apps in 2026.
• You can show it to anyone with a link, with no install friction.

Pick a mobile app only if your idea fails for a clear, specific phone-feature reason. Pick Door 3 if your idea is small enough that even a web app is overkill.

Forward references

Day 17 asks where AI fits inside your shape choice, which is downstream of today's call. Day 18 sketches the stack for your specific choice. Day 21 sequences the build, including which shape you ship first and which shapes (if any) you graduate to later.

Day 16 wrap

The thing you can now say plainly. There are three shapes (web, mobile, neither) and the answer is "web" most of the time, "neither" more often than beginners think, and "mobile" only when a specific phone feature requires it.

The thing you can now do. Pick your shape for your build, on purpose, with a one-sentence reason for the choice.

The guardrail to remember. "Should this be an app" is the wrong question. The right one is "what is the smallest shape that would teach me whether my idea is real."

See you on Day 17, where we ask where AI actually fits.