Rely on your experience - The main problem for startups is that they are looking for a problem. Find what you are passionate about, what you understand.
What is your feature - Have your point of view on your startup. Why is it that time now?
Never do without a draft - It is better to spend three, four pieces of paper to discuss the features than three weeks to write a project.
Study every week - Start by asking: “Do people want to upload their photos on the go? Can we make a good filter? ”- Spend a week on research, make a decision on Friday and move on.
Check on those around you - “We call it a bar exam. If you can’t tell your idea in a bar to a guy or girl, then it should be simplified. ” Speak not only with your irritable friends.
To know the moment when it is time to go on - If all the answers are received, it is time to continue.
Acceptance of the wizard of the OZ country for the prototype - no need to do everything at once. Part of the functionality can be simulated. So Michael and Sistrom (another Instagram creator) tested the idea of ​​sending a message that your friend had already connected to the service, manually simulating the behavior of the bot. It turned out that nobody needs it. Since we wrote 0 lines of Python code and these 0 lines of code we dropped.
Constantly create and show your audience - Do not waste a month on creating something, without checking whether someone needs it.
In principle, nothing new, but it is always useful to clarify the correctness of its course.