The science behind the magic of good business ideas

The science behind the magic of good business ideas

How do you find a great business idea?

Ideas are worth dime a dozen, and execution is everything? A good hard working team will win, no matter what, right? I call bullshit on this. I would rather be an average team working on a great idea (in a great market) than a great team with a shitty idea. No matter how much of a genius you are, the market you serve will determine your success. 

With that rant on why over, let’s move on to the how. I want to specifically talk about finding an idea for a B2B SaaS startup, because the B2C side can sometimes be more of an art than a science. 

What will a good idea look like

Before we start though, lemme give you a disclaimer - your job is not to find an idea. It is to find a problem which is painful enough that people are either already paying for it, or they spend a huge amount of time putting up band-aid solutions. 

When you do find the right problem - you will see that there are X people in the market with that problem, Y number of solutions already exist, and those X are already spending Z efforts in solving that problem. Ideally, X should be growing at A% every year, rapidly. Now you have to simply find a smaller niche within X that is still unhappy with Y, and build something 1% better for them. It really is as simple as that. You can look at every B2B company out there, and this is what they did. 

But Aditya, that’s boring. I want to be an innovator

Fuck yeah it is boring, but so can be the daily grind of building a business (sometimes). You can try and invent something completely new, like the cure for cancer or put rockets on Mars, but then you are playing a very different game. And I promised you science, not magic. That game is beyond the realm of mere mortals like me, at least. 

Make no mistake, 99% of innovative companies also started out the same. Facebook was a slightly better social network on launch day, just for Stanford students. Google was a slightly better search engine because they focused on linking different web pages together, not just their content. 

And what do you think, building something 1% better than others, and keeping up that game is easy? No it’s not. Go ask the successful ones.  And playing a niche helps you build something that is 100x loved by your audience which grows like wildfire in that community. And then (if you want) you can expand from there. The more crowded a market, the better it is for you, because you can then easily find a niche within those bloody red ocean waters and win it. 

You need to find out if others have paid for solving that problem in the past. They should have paid money to expert consultants in that category, hired employees to fix the problem manually, or bought another software. 

What will a bad idea look like 

All really bad SaaS ideas have one thing in common - they try to change the user’s behaviour. I mean you can still try to do this, but not unless you have something 1000x better than before, or you have network effects, or boat loads of VC money. 

Bad ideas also try to solve a problem which no one seems to care about enough. 

I think the market's readiness determines if your journey is going to be an up-hill or down-hill one from an "effort to outcome" perspective. And nothing like if you are able to catch a tailwind (which is unlikely if you are building for a market that's not ready yet) - Mathew John, CEO at Typito

Step #1 - Build your framework

The reason why people feel overwhelmed while choosing a startup idea is because the universe of possible ideas to choose is too vast. There are problems to fix in every category; from logistics, food, and team building to security and productivity. Even within a category, you can choose different personas e.g. hiring issues of a newly funded tech startup are very different from those being faced by a bank on wall street.

My framework had three very simple criteria, but that helped me define the ideal business in a very good way.

I wanted to prove business without depending on external funding 

  • Keep it simple. Can’t do AI, or other deep tech businesses that require me to hire people
  • Must be a proven crowded market. I won’t have money to build a category from ground up
  • Faster validation would needed so that I can double down only if it’s gonna work
  • I should have experience in the industry, so that I know how to quickly win

I want to be happy solving the problem

  • Will I feel joy building it? Will it make the world a better place? That removed ad-tech. 
  • Will I use it myself and dog-food it?

I want to play the long term game

  • Will the problem exist after 100 years too? Or is it a fad?
  • Is the market growing?

This is how Kaapi was born  - employee engagement & feedback for remote teams. It is a simple software within a crowded market. I have 7 years of experience in the HR industry. I would be its own first customer to dog-food it. I find it joyful and satisfying to solve this problem. Remote work is a growing market. Building on top of Slack allowed us to validate it fast. 

Step #2 - Input problems you encounter into the framework 

This is where you list down everything that comes to your mind over the next 4-8 weeks. There are no rules here, just keep dumping your brain into a notebook. To help you organise a bit, I would recommend a simple project with 3 columns:

  1. New ideas - dump everything here. Doesn’t matter how crazy it is. Personally, I am a fan of building for problems that you face personally. Because then you already have the first customer for your business, yourself. 
  2. Promising ideas - Every few days, keep pruning through the 1st column, and if you find yourself excited by an idea multiple times,  then move that idea to this column. You should still not worry about the validity of the idea. But simply trust your gut a bit to point at one. 
  3. Actively exploring - Now comes the exciting part. Every few days, pick the most exciting one from the 2nd column and start validating it. More details below.

Here is the link to my project from 2019

Step #3 - Validate the idea

If you are a developer, you would most probably want to start coding right now. Stop! Your brain is excited about the idea, but your brain already has a lot of inbuilt biases to fool you. Here are some things I recommend:

Talk to customers

This is a pretty obvious one.  And better people than me have talked about it, so I will keep it short. 

  • Go read The Mom Test. It changed my life, and I am sure it will change yours too.
  • The biggest thing you need to keep in mind while talking to customers is that past behaviour is more important than future aspirations. Asking “Will you pay for this?” is a bad idea. Ask “What did you do when last time you had this problem”
  • It is very important to identify the right persona - who has the problem? Where do they work? How old are they? E.g. with Kaapi, we got different answers when we talked to HRs vs CXOs. 

Look at market trends

I recommend breaking this down into 4 categories - 

  • What are people typing into Google in the past - This is very easy to do and will tell you the actual intent of potential users. You can go to the Keywords Tool in Google Adwords to do this research very quickly. Simply type a few keywords, and then the tool will show you gold. 
  • What is the trend of that keyword -  Can you catch a wave in the market vs trying to roll a boulder uphill for the next 3 years? I was also exploring online scheduling softwares at that time, and it was very exciting to see it trending really high up, but it did not meet my criteria of “I want to be happy solving that problem”
  • What is bad about current solutions - While talking to current customers will help, you can also scale this by looking at online reviews of your competition. For Kaapi, I googled “Best employee engagement softwares” → landed on → filtered for small business solutions → looked at negative reviews of all such softwares 

As always, I want to be 100% open and transparent, so here is the actual excel sheet I still use every month to map the market.

Ensure that buyer and user is the same 

A buyer is the decision maker in the organization and a user is the well, the person who uses it. If you are a payroll software, the buyer might be the CEO, but the user would be the HR team. This means that you have to go through a large enterprise sales process of pitching to multiple people, finding champions, and even building multiple features to keep everyone happy. In the early days of validation and MVP, you need to embrace speed, which means that this route is out. You need faster feedback loops!  

If all the signals from the steps above are positive, ok, go and build! 


Not everything will happen in such a smooth flow. It didn’t happen to me at all, and even right now, I am not 100% certain that Kaapi will surely be a big success. During the early phase, you will go through weeks of anguish and frustrating research. You will doubt yourself multiple times..

Edit - If you REALLY REALLY love an idea, and it excites you a lot, then you can probably fuck all this science and just go and build it. I am no expert! But just remember to build a quick MVP vs wasting 2 years of your life on it. 

I hope this framework helps you as much as it helped me! 

The 1% better newsletter

Actionable guides on remote work, leadership & lifestyle design.
No spam, ever.

Read more

Creativity is overrated

The science of being creative

Read Story

Change your heroes

Want to change your life? Start here.

Read Story