How to Evaluate Your Startup Idea (or Investment)
Based on my experience as a startup co-founder, small business owner, and early employee at Uber. These are the things I check before jumping in.
I’ve run this search on Google, and the results are just too abstract for me. Instead, let’s get practical. I’m going to describe what it’s like to start a company, the hurdles you hit along the way, and then work backwards into how you should evaluate your startup ideas (or investments).
For context, I’m an… entrepreneur-in-progress. Not particularly good at it yet, but with enough experience to know some of the many things you shouldn’t do.
My “qualifications”:
I co-own a San Francisco bakery, called Butter& (140k followers on IG) that’s probably the #1 result when you search cake on Yelp here
I’m a co-founder at a startup in SF called Pastel (v1: suburban logistics for local businesses, v2: the first customer support tool designed for local businesses), for which I’ve raised $4M
I was an early-ish employee at Uber (there from 200 employees to 30,000), which was exciting and traumatic and has informed some of my opinions about startups
I’ve made 20+ investments in other startups (Snackpass, Citizen, Kyte, Magic Eden), which is enough to think about how to evaluate startups, but not enough to have a successful portfolio
When you start a company, here’s what the happy path looks like:
You make a product or service
You try to sell it to some people and they purchase it
You make enough money from those purchases to reinvest your gains into selling it to some more people
You keep selling to bigger and bigger groups of people in new ways
People copy you, and you feel sad but survive
You spend 5-10 years on it, if it’s going well
And hopefully things finally stabilize and you move onto something else
If you want to be as efficient as possible, you’ll want to work on something that will reasonably survive all those steps. I’m writing this because going through the company building process only to realize you were doomed from the start is one of the most painful feelings. Starting and running a company is so, so, so much work that to do it and not have a solid chance of succeeding is tragic.
So here’s the list I’ve learned to check against any time I think about starting a new company (or investing in someone else’s):
Urgent problem. You’re solving an urgent, specific problem (or a strong desire, though this is much harder) for someone. Test: Are these folks already trying to solve this problem with money today? They might pay you then. Test: When you talk about this problem with them, do their eyes light up? After you tell them you’re working on this problem, do they follow up with you proactively? Sales will be easy then.
Top solution. You can be the #1 or #2 solution to this problem. Ideally, there’s no large, well-funded competitor at all — that’s rare. Test: Can you be 10x better than the closest competitor? If so, your ads will convert really well and your customers will promote you for free. Test: Are there a lot of companies already competing for these keywords in Google Search or this audience on Instagram/Facebook? Those are the two easiest way to grow a business, so if others are spending heavily in those channels already, you’re going to need a very creative alternative to reaching those customers.
Good business model. You can make good money once users adopt that solution. Test: What are the gross margins here? Most tech companies try to be over 80%. Most local businesses try to be over 50%. Test: If you sell this to a customer once, will they keep purchasing from you in the future? Recurring revenue (even if it’s not a subscription) is the easiest way to have high margins and sticky growth. Test: What’s the average purchase value going to be? Can you pay to acquire those users and still make good money?
Large market. Your market is large, so those good margins get multiplied over a big group of people. Ideally, it’s a small market right now that’s growing really fast — that’s usually a blind spot for your competition. Test: What’s the incremental cost of serving another city? Another vertical? If the answer is near 0, your product will scale gracefully. Otherwise, its expansion will hinge on future investment, which adds more risk to your path to success.
Defensible. Your success is hard to imitate. Test: If you did this for a year and someone copied you, how much business would you lose? How much harder would growth become? This is a test of the moats around your business. All great business ideas (and many bad ones) get copied, sadly.
Right team. You and your team can actually pull this off. Test: Imagine you were competing against a dream team. How much better are they? What advantages would they have over you? What are the chances a team like that exists and will copy you?
Good vibes. You don’t hate your life. 5+ years is a long time to work on something you hate (with people you hate) — or are just ambivalent about. Conversely, the more you can work on things that other people find incredibly boring, the less likely you are to face competition.
Most ideas don’t pass tests 1 and 2. The most common errors are:
Solving a problem that’s just kind of annoying. “This is something that would make people’s lives better, if they just gave it a chance!”, you argue. Selling products and services that just solve annoying (but not urgent) problems is really slow work. A master marketer might be able to pull it off, or someone with an enormous existing audience. This does not describe most of us. (If it does describe you, please reach out – there are so many on-the-fence business ideas we could let soar together!)
Solving an urgent problem that is already solved — and not being 10x better. Competitors have already gobbled up all the easy Google Search keywords, and they’re bidding the heck out of the Instagram/Facebook audiences. The only way you’ll win in this scenario is by having a solution that is so much better (10x cheaper, 10x more effective, whatever) that customers will convert on your ads at a much higher rate and your customers will talk about you for free. When was the last time you promoted a product to your friends? This is a very high bar.
Let’s use an example. Here’s what a bad idea looks like. This is something I actually built with my co-founders, tried to launch, and failed at.
We were making a tool that small businesses in the food industry could use to plan out their weekly production. (Are you sensing a theme here? I really like thinking about small businesses. For better or for worse — that’s a whole other post.) They usually have some number of preorders that they know they need to make, then they are making a prediction about orders that might happen, and they need to plan out when they and their team are going to make which items, and therefore which ingredients to buy and what labor to plan out. So let’s take it through the framework:
✅ Urgent problem. Planning production is pretty urgent each week. It’s often the first thing a small business owner or manager will do in the morning when the workweek starts. Labor and ingredients are the two biggest costs in a food business, so it all needs to be planned out carefully to keep the business’s margins on track.
❌ Top solution. The current solution is using Excel and Google Sheets. Most businesses have developed some sort of custom template, they drop order data into it after exporting from their ordering system, and a little massaging gets them to the plan they need. Our solution was automatic, instant, and easy to update on the fly if orders changed. But it wasn’t 10x better than Excel and Google Sheets.
❌ Good business model. It would be a software business, but when we started asking about price points, customers told us it might be worth $10–20/month for them. That might work for Quickbooks (which is quite cheap) but that’s because they built their business at a time when there was almost no competition for search keywords. Acquiring users and still making money on those accounts was going to be very challenging. And we’d need a lot of those users to build a big business. Even 100k users at that price point would be a smaller business than we wanted.
✅ Large market. There are 33M small businesses in the US. Maybe 1M of those are restaurants and bakeries. In theory, production planning is something all of those businesses do, so no problem here. And we were sure there were other business types who had production processes that also required planning outside of the obvious retail food businesses.
❌ Defensible. The ordering systems themselves, say Toast or Shopify or Square, could decide to copy us and it’d be game over, since they own the underlying order data. Not a good place to be — and perhaps a hint that this was more of a feature than a standalone product.
✅ Right team. We’re 3 co-founders and are ourselves small business owners (that bakery experience!), plus our technical co-founder, Anh, is the most talented frontend developer I’ve ever met. We have an online audience of about 150k followers in total, many of them small businesses. It’s hard to imagine a more competitive team for this idea.
✅ Good vibes. We really enjoy spending time together, and we enjoy helping small businesses. So no worries here.
What do you think about this framework? Put it to the test against your own experiences and let me know how it does. I’d love to keep improving my heuristics so I can get smarter about how I spend my time and resources, so let me know if you have any feedback or ideas in the comments.