Where do startups ideas come from?

A framework for ideas from someone that has seen thousands of startups

They overvalue ideas. They think creating a startup is just a matter of implementing some fabulous initial idea. And since a successful startup is worth millions of dollars, a good idea is therefore a million-dollar idea.

Paul Graham, Ideas for Startups

It’s a question I have been asking myself for the past dozen years. After recovering from the ups and downs of my previous startup, I thought I would jump back in immediately with another idea. Every idea I came up with though eventually progressed along the same path. First it was exciting, then came the doubts, and finally the resignation that this idea was not worth pursuing.

I also get asked the “where do I find a good idea for a startup” question from people exploring the leap into entrepreneurship. Often this happens at conferences and meetups where I am speaking, and folks will come up afterwards eagerly pitching me their idea to get some advice or validation that their idea is worthy of being a startup.

The most I can muster at such times is a faint smile and a “sounds good” or “go for it”, though I am pretty sure I do not sound all that convincing. The truth is that I have no idea what could be a good idea from a bad one. If you think back over the past decade, the idea of driving random people around in your own car or renting out a room to strangers in your home seemed ridiculous. But these became two of the biggest startups and changed industries.

Can you just think up new startup ideas?

Of course, most of you reading probably already have a startup. Even if this question is not one you would ask now, it will eventually come up. When it does, it is because you are contemplating a pivot, expanding upon the existing business through new products, monetization channels, etc., or you are at the end of the journey with your current startup and pondering the future.

Nowadays there are plenty of outlets online to get advice on finding startup ideas. While many of them are valid, I often go back to this one guy that has been writing all about startups since 2000. Today that would not be all that strange, but back then it was still a novelty. I called this guy “Mr. Startups” as he had been a startup founder, sold the startup, and spent his time inviting other founders to come out to Silicon Valley for a few months to build out their ideas.

If you haven’t figured it out already, I am talking about Paul Graham and the program he co-founded Y Combinator. I liked his writing style and way of thinking. Nothing seemed revolutionary, but everything he wrote just made things so much clearer when I was launching my own startup.

On the topic of ideas, he was quite prolific in sharing his thoughts over seven posts that specifically mentioned “ideas”. As someone that has engaged with thousands of startup founders through YC, his thoughts on generating startup ideas seem particularly valid. In analyzing his posts, I summarized his thinking into five key insights:

  • Obvious ideas are not ones you "think up" but "notice"

  • Live in the future and build what seems interesting

  • Ideas are a starting point, not a blueprint

  • Focus more on the idea and less on the startup

  • Startups are rarely ever killed by competitors

Obvious ideas are an often an outcropping of working on the right type of problems. This often leads to eureka moments that point you to something broken or missing that you could uniquely solve because of your firsthand knowledge of the domain. What seems like a flash of inspiration however is often the culmination of observing the problem over a long period of time rather than just conjuring up ideas at random.

When thinking about ideas, it is also helpful to place yourself in the future. This is not to suggest going into science fiction mode. Rather you want to understand what things in the present age would seem antiquated to your future self. This works in any sphere such as life sciences, marketplaces, automation, and any other situation where one could imagine a large shift in user behaviors like what we have seen with mobile phones and now with AI.

The thing about ideas is that they never come fully formed at the start. You have the shape of the idea, but it still seems fuzzy. Paul says the best approach is to think about ideas as questions. As he puts it, “If an idea is a blueprint, it has to be right. But if it's a question, it can be wrong, so long as it's wrong in a way that leads to more ideas.” With more iterations and knowledge, the idea becomes more concrete.

The biggest barrier for me launching a new startup has been asking my questions in the frame of whether it is a good startup. While you eventually want to build a startup that is capable of making money, monetization is often takes a longer time to figure out. I was at Stack Overflow to help launch the SaaS business, a company that had for the previous eight years struggled to find the right business model. If you have a good idea, the answer to how it makes money will eventually come.

Lastly, many entrepreneurs get hung up on whether their idea is in “too crowded” of a space. Having competitors is actually a good sign as it shows there is a market with buyers. Companies like Facebook and Dropbox were launched when the prevailing wisdom was those markets were already saturated with entrenched incumbents. Much like the asking whether a startup idea is a good business, it takes time to find your niche and unique positioning.

In a sense, what Paul Graham shares in his writing influenced my own journey into my own startup. I did not search for a good idea for a startup when I started, it was just obvious to me. I was at Oracle at the time and every one of my customers had and despised their HR system, PeopleSoft. Besides being both expensive and complicated, it did nothing to help address the enormous misalignment of talent in large organizations where people were not recognized or valued. The missing ingredient was data on skills, and that was the spark that nudged me to launch a startup.

Keep these five thoughts in mind is a help framework for at getting started with startup ideas. This can also help when considering a new product direction or full pivot. Many well-known startups began as something completely different. Slack, Twitter, Groupon, and Shopify are just a few startups of many that started as one thing, noticed something interesting, and focused on solving that newer problem.

What about your startup or other startups that you have seen? Where did the idea start and how did the idea evolve over time?

One of the big controversies that came up last week was about serverless. It has been a huge boost for productivity, especially for startups, but a post from the Amazon Prime Video team brought up many questions. Understandably folks were confused by the viewpoints, but here's our take for startups:

  1. Always read beyond the headlines and understand the context, many of these situations were based on very specific use cases.

  2. Speed is everything early on, so the best strategy is to build using what you know and can allow you to iterate quickly.

  3. Learn from experts like Werner Vogels, Adrian Cockcroft & Adrian Hornsby, they have seen it all when it comes to architecture.

We have shared the links below to three posts that are well worth reading to clarify the key point here, which is that startups should use the appropriate tool for the appropriate use case!

"Monoliths are not dinosaurs" by Werner Vogels

It has been a whirlwind tour for the Startup Advocate team over the past couple of weeks! Check out some of our recent posts from Berlin and Kenya, where we got to meet so many awesome startups and founders to learn about the nuances of the local startup ecosystem and how AWS plays a part in supporting and helping startups to build and scale.

Over this week and next, Mark is on a Southeast / Midwest USA swing with community events Nashville and Indianapolis. If you are in either place, come an join Mark at our Founder Night Event or the Powerkeg Conference. Will have more updates on the world tour next week 👋