Taro Logo

Beginning Side projects - How to plan and start

Profile picture
Entry-Level Software Engineer at Taro Communitya month ago

I want to get started badly with side projects. I feel like I never know how to truly get started with them, and most importantly, planning with scope. I don't know if something is ever too basic or too complicated with my knowledge. I wanted to ask, how do you all really get started with a project most importantly finish?



  • 18
    Profile picture
    Founding ML Engineer @ Lancey (YC S22)
    a month ago

    My criteria for picking side projects:

    • Should be easily demo-able/show proof
      • Web apps are great because you can drop a link and interviewers can check it out. I've done in this interviews a couple times where I drop the link and they check it out while I'm explaining it
      • Android/iOS apps are great because people can check the playstore link and see it has X downloads
      • Something that is hard to demo/not publicly accessible like a CLI tool or software that runs private is generally less preferrable

    When it comes to side projects dont over think it. Software can always be made more and more complex with additional features. Pick something you can hack together in 2-3 hours and keep iterating over months. I dont think any software is ever "done".

    Even something as simple as a login page can be made enormously complex (string validation, SSO, login with google/apple/github, reset password, ensure new password is not the same as old, ensure it shows properly on windows/android/desktop, dark mode/night mode, passwords should match, username should be unique)

    I wouldn't worry about picking something too easy or too hard. Just figure it out as you go! use chatGPT or google to learn what you need as you need.

    If you're feeling stuck a simple idea is pick an app you use and clone it and give it out for free-- gym tracker, nutrition tracker, productivity apps like notes, chrome extension to track submitted job apps -- and then make it free


  • 9
    Profile picture
    Tech Lead/Manager at Meta, Pinterest, Kosei
    a month ago

    RE: how to finish: make the completion criteria for publishing the first version of the project easy. Like, laughably easy.

    No one will care or use your initial version of the product, it's just for you to build momentum and start doing more iteration.

    How Long to Crank Out a Decent Side Project?

  • 8
    Profile picture
    Tech Lead @ Robinhood, Meta, Course Hero
    a month ago

    Sai pretty much nailed it. The important thing here given that you're junior and just starting is to choose an idea that's stupidly simple. Something so simple that you're afraid that other engineers will make fun of you for it (don't worry, they won't). Follow the advice here: "Finding a mobile app idea - How to do it?"

    After you have the idea, break it down and ship it within a week: "How do I make software less overwhelming?"

  • 1
    Profile picture
    Entry-Level Software Engineer at Unemployed
    17 days ago

    Ideally the best project idea is one that you've been wanting to make for a while and is not too complex. This takes time to come up with and exposure to different problems/fields. I have a Obsidian file where I list every new idea that I think could be viable.

    Another way is to get ideas from others. For this, you can simply ask people or find collections. Here is a tweet with many ideas: https://x.com/\_buildspace/status/1793805155080777955.

    I'm currently making a carpooling website at antride.ca. I remember talking about the idea 2.5 years ago in a internship interview (which I did not end up getting), and I'm finally getting around to making it. The idea is from personal experience of a uni club I was in that would be have off-campus events intermittedly. Carpooling wasn't attempted by the organizers and the current websites are unintuitive. No carpooling website currently exists that auto-matches people into groups based on location to make the optimal carpooling routes. So I decided to make one.

    Once you start your idea, make a project overview document. List the problem, the solution, and the minimum viable product you would have to make to bring value. Keep scope as small as possible. Resist the urge for perfection.

    If it has the ablity to get users, do a bit of upfront research by making a survey or asking potential users what they think of it.

    And remember to enjoy the process by working on an idea that excites you.