How to Get Startup Ideas

Live in the future, then build what's missing

The way to get startup ideas is not to try to think of startup ideas — it's to notice problems, preferably ones you have yourself, and preferably problems that exist because you're standing slightly ahead of everyone else. The best ideas usually share three traits at once: the founders want the thing themselves, the founders can actually build it, and almost no one else realizes yet that it's worth doing. Drop any one of the three and you get a feature, a hobby, or a business plan competition entry — not a startup.

Where organic ideas actually come from

1. Ideas that grow out of your own work, not ones invented in a notebook to sound impressive at dinner
2. Being an early, serious user of some new technology puts you ahead of problems other people haven't hit yet — the tooling gaps around a new API, model, or platform reveal themselves to users, not analysts
3. Being a genuine expert user beats being a clever outside observer of a market you've read about but never lived in
4. The problems worth solving are the ones that annoy you personally, on a recurring basis — not the ones that sound biggest in a pitch
5. If it looks 'too small to be a company,' that's often exactly what makes it invisible to everyone except you

Schlep blindness

There are great startup ideas lying around unexploited because they require some unpleasant, tedious, unscalable task that nobody wants to do — paperwork, compliance, integrations, cold outreach, manual reconciliation. Your own unconscious mind will hide these ideas from you, filing them under 'chore' instead of 'company.' The most dangerous version isn't administrative schlep, it's technical glue work: connecting two systems that don't want to talk to each other, which reads as beneath a real engineer right up until someone builds a business on doing it well.

Signals you might be schlep-blind right now

1. You've said 'I should just write a script for this' more than three times about the same task
2. The task involves a human manually bridging two systems that should already talk to each other
3. You've mentally filed it as 'ops work' or 'busywork,' never as 'thing I could sell'
4. When you describe it to someone in the same field, they immediately say 'oh god, yes, that too'

A five-minute idea audit

1. Write the idea as the problem, not as the product — 'reconciling vendor spend by hand' beats 'a FinOps platform'
2. Name one real person, not a market segment, who has this problem today — ideally yourself
3. List every schlep required to actually deliver the solution, including the boring integration and support work
4. Ask why a bigger company hasn't already done this — a good answer is 'it looks too small or too schleppy for them to bother,' a bad answer is 'no one thought of it'
5. Check whether you are the target user, or at most one honest degree removed from one