How to come up with strong startup ideas

Team SaaSBox


This essay by Paul Graham suggests tinkering with ideas that attract users instead of sitting and thinking of plausible business ideas. This is certainly a good way to come up with a solution that people want to use and pay for - You just gain some skill in some domain, then come up with a solution that makes something efficient, fun, or useful in that domain (like a developer tool, or a game).

However, there are two forces that work against you.

The first one is that technology is complex. If your product is addressing one layer in a technical stack, it is hard to find a fit. If we take AI as an example, there are now many vendors working on different layers of the AI problem, some of which are tools that let you build AI models easier. If the technology is complex, how do you find your buyers and audience? You have to pinpoint the group of people that will use your tool or tech, within that stack. Is it possible? Yes, but there is ambiguity. When starting out you don’t know who will use your product, in what ways? It is a discovery in itself - you have some complex technology or tool, and now you need to figure out in what ways your users will use it. It is a real problem.

The second one is that markets are crowded. They were always crowded, but tech used to be a smaller world back in 2010s. Today there are so many products, at least some fields such as web, cloud and mobile apps, are saturated. AI is already commoditized before it became a wave. In this case you are lost in the noise, you either need strong distribution, marketing, network, or you won’t stand out. This is another important problem, your solution may make a difference, but you won’t be discovered, or considered because there are many others that look like you. For the user it is hard to know if you are really better.

The first one is nasty - how do you find your buyers and audience and figure out how they will use your tech product? The solution is to focus on existing user behaviors. Just build something that solves a problem where there is an existing user behavior. Most of the FAANG companies are doing this. People like watching movies, it is such an obvious behavior and need. Netflix makes this more efficient and convenient. People always bought books, and Amazon made the experience so much better. You still just need and buy stuff. Amazon is targeting this behavior. For Amazon to start, there was absolutely no ambiguity on how the users would use the product or whether they would need it. Maybe the question was, will they shop online? They made something people wanted by improving an existing user behavior. Contrast this to a developer tool. Do developer tools sell? Most certainly. But it is riskier starting out in that it is ambiguous who will use it and how it will be used.

It is no doubt that such a generic user behavior is shared by the entire world (watching movies, searching for things, buying stuff) that these companies have the biggest market share. Facebook was more tricky, it was a tech spin on an existing behavior but tricky to come up with. When Google launched, search was already an existing behavior, and when iphone came out, phones were already very popular.

The takeaway is that focusing on an existing, popular user behavior greatly simplifies your buyer audience problem, which is one of the biggest problems for any new business. If you are struggling to fit your technical solution to a user profile or use case, maybe there is no need to struggle. Just change your solution to target an existing behavior.

For example, with SaaSBox - the user behavior needs more clarity. The standard behavior for launching a SaaS is to pick a developer framework of your choice. Most recent popular way is to pick a code template or framework, and starting from scratch. The default behavior for non-technical folks is hiring engineers and paying them. The behavior is “find and hire devs”. Nobody currently is looking for a “platform” where they can launch their SaaS, since this would be a new behavior. Also launching a SaaS is a niche problem with a smaller audience.

The second problem is also tricky. How do you stand out? One of the solutions, is to come up with a new spin, new approach, or deep tech, or amazing UX on a given problem, and make it 10X better. The effect of this is that people notice it due to its novelty, and your solution stands out. Word of mouth kicks in. When you reach out, people respond and you can build your network. Everyone wants you on their side with your 10X product that your networking attempts are fruitful.

Here is the Jeff Bezos interview that touches on both points - he built a solution that targeted existing behavior (simple to know users want it) and had a spin on it that made it 10x better (novelty made it easier to discover) . Also, he succeeded by thinking of a business idea and make it work, (of course not everyone has the same thought process as JB) instead of tinkering with a hobby etc. where targeting a known behavior or standing out would be an afterthought to figure out.

For SaaS companies, targeting existing behavior is especially important - as Tawheed Kader puts it in this tweet. A lot of SaaS startups die due to obscurity. People don’t know how to use them or understand what they are for. Founders are also blind, trying to understand why users behave the way they do.

The conclusion is to focus on an existing behavior if possible, and make it 10x better.

Join The Discussion