How to Evaluate Startup Opportunities (feat. Math)
On my first day at business school I tucked my polo shirt into khaki shorts.
I was 24 and had decided to leave a finance job in NYC to go to UNC Chapel Hill, and despite never having tucked a polo shirt into khaki shorts before, a few happy hours leading up to the first day made it crystal clear that that’s what people did here. Social proof is a hell of a drug. We’ll get back to that.
My first class was Accounting, and the first words out of my professor’s mouth boomed as anxious students shuffled in.
“Good morning! How many people in this class share the same birthday?”
Confused silence.
“Ok, let me ask it another way. There are 40 students in this class. Raise your hand if you think any two people share the same birthday.”
No hands. Sloppy mental math in my head said 40 people and 365 days equaled probably no shared birthdays.
“How about this - raise your hand if you don’t think anyone in this classroom was born on the same day.”
A smattering of hands. One kid raised his a bit too quick and untucked his polo. He wouldn’t make that mistake again.
“Ok, great. Let’s go around the room. You-”
*Professor points at someone in the front row*
“What’s your birthday?”
“January 18th.”
“Anyone else in the room born on January 18th?”
A hand in the back goes up.
“That’s… my birthday, too.”
“Ok, then. You all fail accounting.”
Humans don’t get math and it’s a big reason your startup won’t work
The really important math - the stuff that’ll drive your startup - isn’t intuitive.
For example, humans don’t innately understand probabilities or exponential growth. We can learn each, but very few people can internalize them to the point that they identify and use them in everyday life.
Which is a big problem for entrepreneurs, because that’s the math that’ll help you evaluate the impact and likelihood of strategies with long-odds. And your startup will depend entirely on a few of your long-odds strategies working out. The thing that our success relies on is the same thing we’re cognitively unable to understand. Not great, Bob.
So we all end up just kind of back-of-napkining the NPV of different potential strategies, and we’re usually way off. Then we fail, take a job at Deloitte and blame the whole thing on not being able to raise money.
In the birthday example, basically every student heard the professor and thought:
“There are 39 people other people in the room and there are 365 days in a year, so there’s roughly a 1 in 9 shot someone shares my birthday. So, I guess there’s a 1 in 9 shot anyone in the room shares anyone else’s birthday.”
Most math mistakes we make are because we’re the stars of our own little movies and overestimate our impact. We assume we’re all Harry Potter, but in reality we’re probably just a random kid in Hufflepuff.
So we don’t realize that 40 students actually means 780 potential pairs - 780 chances for people to share a birthday, most of which have nothing to do with us. And, not surprisingly, someone did.
Our teacher showed us the best way to think about the problem, which is to calculate the chances that all 40 birthdays are unique. Some relatively simple math followed:
(364 / 365) ^ 780 = ~12%
There was a 12% chance all birthdays were unique, which meant there was an 88% chance of a birthday match. That number goes up to 99% with just 56 students. But if you ask 10 people if they think a room full of 56 students is likely to have a birthday match, I guarantee you they all say no. We’re intuitively bad at math. And our startups suffer.
Which brings us to private chefs and why you shouldn’t start a newsletter.
“Uber for Private Chefs”
We’ve had at least 10 entrepreneurs apply to Tacklebox with a similar idea over the past five years: Uber for Private Chefs.
It’s a gig economy play. The thinking is that there are probably private chefs with downtime, and definitely people who love good food. These startups connect the two. It’s a classic Whisper Idea - your friends will love it and fall all over themselves to tell you how often they would use it, which will make you rush to GoDaddy to buy the URL for a random clever food name like “Jicama.”
These founders approached the idea in two ways:
Ghost Kitchens - create kitchens that chefs can rent out like a WeWork. Private chefs pop in and cook in large quantities, they leverage delivery services like Postmates to get that stuff to people, and magic ensues.
Traveling Chefs - chefs go to people’s houses and prepare meals. These range from the traditional “cater a 12-person dinner” model, to the “cook for 5 hours on Tuesday afternoon and leave 5 days of leftovers in the fridge of a busy family” model. And, as always, magic ensues.
Either approach leads to a marketplace with chefs on one side and people eating delicious food on the other. Both sides make sense in a vacuum, but connecting them… does not.
It’s a textbook underpants problem.
Ghost Kitchens and Traveling Chefs both lead entrepreneurs to two realizations:
Chefs are straightforward - if people want their food and you’ll pay them well, they’ll cook.
Eaters are downright finicky - people will say they’ll use the service, but when you try to nail them down for a meal or set up a chef to cater a dinner, they’re flakier than the halibut the chef was going to serve.
These realizations also led every founder to the same conclusion - they’d need to build up a “waitlist” of eaters before they could launch. They’d need thousands of “eaters” to get any sort of reliable demand for chefs.
But how do you get those eaters?
And that’s the core question that spurred this post:
How do you choose a strategy to acquire customers before you launch?
Paid acquisition is a disaster. Instagram, Facebook, Adwords are all flooded with VC-backed companies inflating prices. These channels will help you test intent and messaging, but acquiring 2,500 good users would easily cost +10 grand, and who knows how good they’d actually be. Plus, you’re competing with professional growth marketers. Plucking people willing to try a new food service from social media is just jumping into the deep end of a pool full of very smart, very well-funded professionals.
So, every founder landed on a “free” acquisition funnel - content. Here’s their strategy - you might recognize it, as it seems to be everyone’s strategy in 2021:
Create content to build an audience / waitlist of a few thousand people
Show the chefs the waitlist to get them excited and on the platform
Launch with enough demand to combat the naturally halibut-flaky customer
And now we need to talk about the Survivorship Bias Staircase, and why content is probably an awful use of your time (and I do recognize that I’m telling you this via content <spiderman gif>).
The Survivorship Bias Staircase and the Consistency Lie
How would you build an audience of eaters?
Content doesn’t seem crazy, right?
Maybe you had a great content idea. A weekly newsletter where you interview a professional chef each week, highlighting 3 recipes for healthy lunches under 20 minutes. Covid has pushed tons of people remote, so home-cooked lunches are definitely a thing. This also gives you an excuse to create relationships with private chefs - featuring them gives them value immediately before you ask them to join as a partner.
And, you’d be consistent - consistency is the key. Even if you didn’t get traction after the first month or two, in 6 or 7 months if you stuck with it you’d have a few thousand subscribers and have featured a hundred or so chefs.
That’s how it works, right?
Successful entrepreneurs talk about this path constantly. They built an audience early and it drove their later success. They say things like “the best time to start building an audience was 2013 and the second best time is today,” and you nod excitedly and launch your Substack.
Investors aren’t helpful, because they’ll tell you the same thing. The way to de-risk this opportunity for them is for you to show up with an audience that already trusts you. So, go do that, and then they’ll chat.
Here’s how you should look at people giving you advice. At Tacklebox we call it the Survivorship Bias Staircase.
It’s natural for them to suggest the things they did that made them successful. But it assumes three dangerous things:
They actually know what made them successful
You can do what made them successful
That same strategy would make something successful today
The most overlooked of these is number 3. All non-Dogecoin successful people in 2021 have been at it for a while - let’s say 10 years. Any tactic advice they give you is 10 years old.
Treat all tactic advice like a gallon of milk. Would you drink a gallon of milk from 2013? Tactics spoil fast and they’ve all got an expiration date.
Let’s take an example that comes up constantly: James Clear. He’s a wonderful habits writer who started a simple newsletter in 2012 and translated it into all sorts of successes, most recently a NYT Best Seller (and terrific book) called Atomic Habits. He’s a deceptively good writer - he writes simply, so most people (mistakenly) assume they can create similar content.
Here’s his journey:
In 2012, his newsletter was novel. No one was writing consistent newsletters, and certainly no one was writing about habits.
Now, here’s you trying to build a newsletter for 20-min lunch recipes in the gauntlet of content that is 2021.
Founders who look to create content that will build an audience just so that they can launch their startup in 2021 are basically saying “hey, I know just building a startup is as hard as solving a rubik’s cube in twenty minutes, but you know what? Let’s up the ante. I’m going to do it blindfolded and hammered drunk.”
And this is where we get back to math. Because you should never employ a strategy, especially a long-term strategy like content with long feedback loops, without understanding your odds.
So let’s look at them.
There’s obviously no way to tell if YOU will be successful, but for nearly every strategy you can find data on base rates. Basically, what is the most likely scenario for someone doing X. This forces you to remember that you’re just a bit character in Hufflepuff, not Harry Potter (and I know even as I write this people reading it are like - “yeah - most people reading this are in Hufflepuff. Obviously I’m Harry Potter, someone has to be, but they should all take this advice.” I know this because I’m… thinking it, too. We’re all in this together!).
Back in 2012, there weren’t many newsletters. There certainly weren’t professional newsletters - it was a nascent space. You were far more likely to be successful with a weekly newsletter, and the problem really was consistency. You just needed time for your base to spread. Compound interest takes a while to really start kicking in, so exponential growth really was a function of not quitting too early.
The Base Rate of someone starting a newsletter - any newsletter - was much higher. Anecdotally, I know this. I started two and had no problem growing to a thousand subs with inconsistent content no where near as good as my startup stuff now.
Base Rates for newsletters in 2021 are drastically different. Content is mature - there are full industries around newsletters. Substack has created tons of professional newsletter writers (there are over 500,000 Substacks). Millions of people pay for newsletters. Morning Brew - a daily news email - sold for $75 million dollars, and it’s looking like that was a steal.
The base rate of a newsletter in 2021 - the most likely scenario for all newly created newsletters in 2021 - is less than 40 subscribers. An adjacent content space with a slightly higher barrier to entry - podcasts - has a base rate of less than 26 downloads per episode. More than 50% of podcasts get less than that. Basically zero independent podcasts break the 5,000 sub barrier.
The question you need to ask is - for this strategy to be successful, what percentile do I need to be in? And is that realistic?
If you’re looking to have an audience of 2,500 engaged subscribers to your 20 minute healthy lunches newsletter, you’d be in the top 1% of newsletters. Here’s the percentage visually:
Maybe you’ll be able to create the Harry Potter of newsletters and grow. But to do that, you’d have to beat well-financed professionals leveraging serious growth tactics. You’d have to write top 1% content. You’d have to be imminently shareable in a world that doesn’t really share content anymore.
It’s not that you can’t. It’s just… is that what you want to bet your startup on? Is that what you are truly exceptional at? Because in a game as competitive as startups and content, you’ll only win at what you’re uniquely great at.
And one more crazy important point on any strategy: people will see through ‘strategies.’
The “why” for any great product is simple - “because the customer wants it.”
The “why” behind most newsletters and podcasts is - “because the entrepreneur wants to leverage it to get customers to sell them something else.”
Customers will sniff that out like pigs to truffles: Customers want what they want - not what you want them to want.
So, what do you do?
Find where you are Harry Potter… and do some math
Most content doesn’t work because there isn’t anything unique or differentiated about it.
The vast majority of content ideas I hear are kind of half-heartedly proposed by the creator. They don’t have a strong point of view. They serve ulterior motives. When I ask why the content needs to exist they defensively say things like “no content needs to exist - James Clear’s content didn’t need to exist,” but for James, it did. And when you read it, that part is clear. Combine that with the fact that he was uniquely gifted as a writer, his product and value actually was content, there are always 10+ habit books on the Amazon top 25 bestseller page, and the base rate for newsletters at the time was high, and you get a very nice little Venn Diagram for a strategy:
If you take away two things from this post, make it these:
We stink at math - dig in and find the base rate of the strategy you’re pursuing.
Remember, you’re the random kid in Hufflepuff. You aren’t Harry Potter. You won’t be able to make any strategy work - don’t bet this huge opportunity on a strategy that isn’t stacked in your favor. You will be able to execute on a strategy that hits the Venn Diagram above. That’s where you live. That’s your home.
Choose where you compete wisely.
Solving a Rubiks Cube is hard enough. No reason to do it in the dark while you’re hammered.