How To Evaluate Startup Ideas

Part of my job at?Minko involves thinking of new ideas.

Ideas that will appeal to our users and possibly give them 10x more value.

It’s almost like I am trying to build mini startups at a startup.

I got my second shot yesterday and woke up feeling a bit woozy/feverish. My calendar was also empty (yay monday) so I decided to do some thinking instead of regular work.

One thought led to another and I was looking for a decision making framework which can help us evaluate ideas in a systemic way using data and market feedback ( rather than gut feel )

On Youtube ,I first came across this?video?by Kevin Hale, Partner at YC where he talks about how YC evaluates startup ideas.

It’s interesting and here are some notes from the video.

• When a founder is pitching ideas to Kevin for the first time – He never says it’s a bad idea.
• He is not being polite. Kevin is following Paul Graham’s approach – unlike other investors who are looking to poke holes into your ideas, he is trying to look for reasons which will make the idea win.
• The bets which went big were very non obvious at the idea stage. e.g. Airbnb.
• It’s not going to sound very obvious when founders first tell you their idea.
• You have to rewire your brain to work on your optimism, creativity and think how to take the story the founder is telling you and think how it can become a billon dollar idea.
• A great investor will then pitch that story back to the founder

A startup idea is a hypothesis – and the hypothesis is how can it grow quickly – Basically think of the plan that will make things grow, and faster.

The first part is the problem?– what is the setting of the problem

• what problems do a large no of people have(1 mil+), and is it growing
• problems that are urgent
• really expensive to solve
• mandatory problems – where the laws have changed
• frequently occurring problems
• it’s ideal if you have multiple – frequency is super important though as it gives people a lot of opportunities to convert – it can help in changing people’s behaviour.
• Motivation – Ability – Trigger : Fundamental to changing user behaviour

The second is the solution?– what experiments are you running

• Don’t start here aka don’t fit a solution to a problem e.g. blockchain or the new new thing
• Don’t SISP – Solution in search of a problem

The third is the Solutions insights?– why is your experiment going to work

• what is the reason it will work
• what is the unfair advantage you have – why will you win instead of somebody else.
• How will you grow faster

Unfair Advantages – are the things that help you grow faster

1. Founders – Are you 1 out of 10. Are you a super expert. ( A product manager at Google isn’t an expert. Somebody with a Phd or a patent is)
2. Market – is your market growing 20% yoy – you have picked the right space but if you are the only company then you are most likely the weakest.
3. Product – Is your product 10x better than the market
4. Acquisition – do you have a sustainable acquisition model? If paid advertising is the only one, then it gets discounted – you need to find acquisition paths that cost no money i.e. word of mouth at $0
5. Monopoly – do you have network effects or a marketplace

Beliefs

• what is the default belief i.e can the founder build it.
• miracle belief – if you are bunch of engineers building an enterprise startup, then the default belief is that they can build it but can they sell it? Do they have sales expertise? If they cracked sales, they would be home.

Example of?Y Combinator as a startup idea pitch

Problem: Hard for founders to raise money without knowing someone in venture capital

Solution: Investing in companies through open applications – you don’t need to know anyone

Unfair Advantages for YC

• Incredible founders?( e.g. Paul Graham had immense credibility – built and sold the first SaaS company)
• Market?– future $B companies will be technology companies powered by software
• Product?– 3 month living expense + access to VCs after MVP
• Acquisition?– PG already had a large audience especially of hackers
• Monopoly?– As the YC alumni grew, it become more powerful and influential

Results since 2005

• Funded 2000 companies
• YC is a community of 4000 founders – CEOs of companies including Dropbox, Airbnb, Reddit, Twitch, Gitlab, Doordash, Stripe etc
• 15 $B+ companies
• 93+ $100M+ companies
• Total market cap of companies is over $100B

Your startup idea is a hypothesis – How can you prove that you can grow quickly.

?

Phew – If you like reading instead of videos, the above is pretty much a synopsis of the 30 min video.

And if you are still reading this, you are most likely interested in startups or exploring an idea or a startup founder. Read on as the its going to get even better.

While the above framework is interesting, I wanted to see if there was something deeper. That’s when I discovered an interesting presentation by?Clint Korver?, Managing Director at ULU Ventures.

He is an expert in the field of Decision Analysis Framework and has helped large companies develop decision making frameworks.

Clint outlines how his?firm?uses a decision model to decide which companies to invest in.

The video is 1 hour long and has two parts.

• The first 30 mins is relevant for startup founders to understand how VCs might evaluate startups
• The second 30 mins is more for VCs on portfolio construction – it debunks some established norms like ‘doubling down’ especially for micro VCs

Clint outlines a model which is quantitative in nature and uses math to help you make a structured decision matrix.

One thing that stood out was failure rates. He took two sets of startups his company had invested in. In the set in which they used the Decision Analysis model, the failure rate of startups was 5%. In the second set where the didn't use it, the failure rate was 50%

I am building a model on similar principles which might hopefully allow us at Minko to make better decisions on new features for products (remember mini startups in a startup)

DM me if you would like to try it out.