Startup Idea Generators

December 18, 2018

Lately, I’ve been been thinking more systematically about startup ideas. Sometimes you’re just not finding organic startup ideas, or you want to think of tons of ideas for fun.


I’ve been using a few symbolic patterns lately that seem good for coming up with relevant startup ideas. That’s just a fancy way of saying I take a pattern or concept that seems interesting right now, and then I come up with ideas that fit into that mold. It’s pretty easy to come up with several ideas all based on the same insight.


I’ve got about a dozen patterns I’ve been thinking of lately, but here I’ll just include three of them, with concrete examples of how I go from a pattern to a rough product concept.


As with any creative exercise, it helps to get specific. With any of these ideas, it’s not enough to say something like “Uber for X.” You really want to be naming specific features and loosely defining an MVP so that you can feel out your idea and force your imagination to do some work.

1: Vertical-specific chat

Pattern: Group chats are used as a “good-enough” solution for several problems.

Process: Find any widespread, specific use-case or for a group chat and build a bespoke app around that.


Examples: People use group chats for all kinds of things. Amongst them are:

  • Conversation with friends
  • Scheduling a hangout
  • Planning an event
  • Talking about crypto prices
  • Live discussion during a broadcast or event
  • Sending memes / content

Because group chats are such a flexible tool for communication, they’re good enough to work for all these use-cases. But a tailored solution around any of these use-cases could steal part of that market.


Example usage:

For example, I’m in a few group chats that are just for general conversation, but whenever a University of Virginia basketball game comes on, that becomes sole focus of the chat. It’s actually confusing because I’m usually not watching the game. I think this is a pretty common, widespread behavior for sports fans.


So, let’s imagine a group chat app specifically for people who are watching sports. Even more specific, an app for group chats during college basketball games. You could form a group around any specific game, and the chat UI could be highly intelligent around that game.


Let’s think of some specific features we could build around that idea:

  • Form your group before gametime, everyone gets a push when the game is about to start
  • People in the group chat can “side” with either team, so you can see who’s rooting for each side. Chat bubbles for people rooting for each team appear on opposite sides of the screen (e.g. Duke fans on the left, UNC fans on the right)
  • Game clock pinned to the top of the chat
  • Key highlights and game events appear in real-time as bubbles in the chat, which people can react to
  • Ability to turn live audio commentary on/off while in the chat
  • Names of players currently on the court for each team are always in a sidebar, with their stats
  • Betting against other members of the chat using Apple Pay (I know this would be illegal, but just getting ideas out there)
  • Dedicated “highlight clips” keyboard for sending replays back and forth in the chat
  • HQ-style polls from ESPN that gauge your group’s reaction after key plays. Ability to see sentiment of everyone else in the world who is watching that game

Again, I’m in quite a few group chats that get completely taken over by basketball when it’s on, then go back to normal conversation afterwards. A thoughtful “group chat for basketball” app could take that time from iMessage, and eventually become “group chat for sports” and then “group chat for events.” Ultimately you could probably just build the world’s best messenger by starting on something so specific.


If you think that’s a terrible way to start, this same pattern can be used for more ideas. Group chat for planning weddings. Group chat for therapy. I know this isn’t a unique insight anymore, but I think there’s room for tons of different group chat apps around different verticals.

2: “Found time”

Pattern: Apps with business hours. Filling gaps in the day with something fun.

Process: Think of common gaps of time in the the day that many people experience. Figure out an app that could make that experience more interesting or useful.


Sachin Monga wrote this piece recently that I really enjoyed, and he mentions the concept of apps with "business hours." For example, HQ owns a specific time of the day, and doesn't hog your attention with an infinite feed of diminishingly meaningful content. What are the other gaps in the day that an app could own?


Examples

  • Waking up and checking phone for a few minutes before getting out of bed
  • Waiting for the train
  • Riding in an Uber or driving to work
  • Waiting on hold for customer service
  • Brushing teeth before going to sleep

Example usage:

An app to use while you wait for the train. I wait for the subway all the time, and it's really interesting to look at all the unique people standing around on the platform. Maybe there’s an app that would make this more fun.


What if there was an app called “Train Waiter?” You open the app and press “I’m waiting.” The app uses your GPS to figure out your current station, and everyone waiting in that station enters a competitive game. Maybe the game is even based on guessing when the train will arrive.


Or maybe you get paired in an anonymous chat with a stranger with the prompt, “Where is this train taking you?”


Or maybe you do nothing in the app, but after awhile, the app tells you which people you’ve been standing next to every day for the past month and tells you to say hi.


I don’t know, some of this could be a little creepy, but it would definitely turn a frustrating daily experience into something exciting and new. That's your distribution hack. People are sitting around bored for certain times of the day, and might be more open to using a product that can change that.

3: Downmarket creativity

Pattern: Finding the lowest possible bar for creative work

Process: Think of any creative or artistic work that people struggle with due to complexity. Figure out what it would take to make the creation easy enough that anyone could do it.


It’s helpful to walk through the steps: Hard -> Medium -> Easy


Describing the hard and medium steps forces you to keep pushing the idea further and get specific with what the easiest version would look like.


Examples

Daniel Gross recently tweeted this with the title “Downmarket Creativity.”


Concretely:

Photography: Film photography -> DSLR -> Instagram


My follow-up to that would be:

Filmmaking: Filming a movie -> Youtube vlogging -> Instagram Stories


Examples:

Music production: Pro music production software -> GarageBand -> ???


Writing: Writing a book -> Medium Blogs -> ??? (maybe this is Twitter)


Painting: Oil on Canvas -> ProCreate for iPad -> ???


Example Usage:

Taking the pattern for music, we could take the challenge of, “How could I design an iOS app that enabled anyone to make good music?”


Some key things to keep in mind would be:

  • Composition: It’s really hard to write a whole song. Maybe it’s just a looper instead. All clips are less than 15 seconds, so that the only thing the user has to do is record one bar of music that loops automatically.
  • The concept of being “in-tune”: Maybe all possible buttons are in the same key.
  • Rhythm: Maybe the app can adjust all of the user’s notes so their timing is perfect.

Bandimal actually made an app with all these considerations in mind, and they won a design award from Apple for it. By redesigning this app for adults and adding the ability to share musical loops to a feed like Instagram, maybe you could make the Instagram of music.


I think this pattern could be really powerful for any difficult creative art form.

Conclusion

These are just three patterns I think are helpful for coming up with new ideas. In the future I'll write about more.


That’s it for now. If you’ve got any patterns of your own, or any cool ideas inspired by the post, hit me up on Twitter @john_c_palmer.