Heuristics

One thing I have noticed about interesting people is that they have short, pithy rules for making decisions. The rules are short because they need to be memorable. And yet they often yield the right decision.

I collect such rules here. This is mostly selfish: these rules seem to work, so it pays to remember and apply them. Thus, if I think a heuristic is wrong, it isn’t included here.

It’s also important to note that none of these rules are perfect. For example, you don’t always need to work on stuff that would make your friends say ‘wow’. It’s just a good way of focussing on impact.

I’d love to collect more of these. If you know any good ones, please leave a comment.

Food

The ‘native’ rule: for ethnic food, pick places where you see a lot of people from that ethnic group eating. (Source: observation)

The ‘lesser known’ rule: pick lesser known ethnic food; e.g. prefer Vietnamese food to Thai food, Pakistani to Indian. Mass popularity generally leads to ‘watered down’ cuisine to appeal to the widest possible audience. (Source)

The ‘long black’ rule: to find good coffee shops, find places that serve a ‘long black’. It’s a fancy name for an Americano, but this rule works pretty well. (‘Flat white’ would’ve worked a few years ago, but it’s since gone mainstream.) [Source: I saw this somewhere on Twitter.]

Hiring (for startups)

The Sunday test: If this person were alone in the office on a Sunday, would that make you more likely to come in and want to work with them? (Source)

The ‘animal’ test: Could you describe the person as an animal? If not, don’t hire. Being an ‘animal’ means having a combination of tenacity and obsessiveness with your work. (Source)

The ‘explanation’ test: Can they relay complex ideas in simple terms? If so, they are probably intelligent. (Source. Also a common Google founder interview tactic).

People

The ‘third world prison’ test: Could this person get you out of a third world prison? Although designed to find a person to start a relationship with, this test works for finding a resourceful person. (Source). [A nice variant from Hilary Mantel’s ‘Wolf Hall’: “Lock Cromwell in a deep dungeon in the morning, and when you come back that night he’ll be sitting on a plush cushion eating larks’ tongues, and all the gaolers will owe him money.”]

The ‘waiter’ test: To figure out how nice someone is, watch how they treat people in service professions (waiters, checkout clerks etc.). (Source: proverbial).

What to work on

The ‘wow’ test: For a given project: is this something that would make your friends say wow? If not, it’s probably not worth doing. This only works if you have credible friends, of course. (Source)

The ‘dispersion’ test: It’s better to make a few people really happy than to make a lot of people semi-happy. (Source). (Nice variant from Scott Adams: “I was chatting with the television executive in charge of the project and asking what the cutoff was for an acceptable test-audience response. The executive explained that for television shows, the best predictor is not the average response. Averages don’t mean much for entertainment products. What you’re looking for is an unusually strong reaction from a subset of the public, even if the majority hates it.” See also ‘1000 True Fans’)

The ‘regret minimization test’: When you’re 80, are going to regret having tried this? If not, then do it. Famously used by Jeff Bezos when deciding whether to leave his hedge fund job to start Amazon. (Source)

The ‘founder’ rule: “Pick an industry where the founders of the industry — the founders of the important companies in the industry — are still alive and actively involved.” It’s probably still a young, vital and in-flux industry, meaning lots of opportunity. (Source)

The ‘people’ test: Look for smart people and hard problems. If a field has both of these, it’s probably a good one to work on. (Source) Bonus: add ‘small groups’ to this list of criteria.

The ‘user’ rule: Focus on users, not competitors. (Source)

The ’10 years’ test: Don’t start a startup you’re not willing to work on for ten years. (Source)

Miscellaneous

The ‘quality, not quantity’ rule: When you want to get better at doing something, do lots of it, and don’t worry too much about quality at the start. When you are looking to improve, learning to do something fast beats accuracy. (Source; also here applied to software engineering.)

The ‘island’ test: Suppose you’re going to an island with no access to any goods; aside from the obvious stuff, what do you make a point of packing? That’s what you’re addicted to. (Source)

5 thoughts on “Heuristics

  1. Nabeel, this is a really great list of heuristics. One to potentially add to your list comes from Andy Johns (Growth roles at Facebook, Twitter, Quora, and Weatlthfront). Its about hiring certain types of individuals for solving different kinds of business problems:

    Someone very smart once told me his thoughts on hiring while I was at Facebook. He said “When you’re trying to solve a problem that’s never been solved before, you don’t start by hiring people that have a predisposition towards solving a known problem in a particular way. Instead, you hire smart, hungry people without predispositions and put them in a resource-constrained environment. Then, in that environment, you push them to find a solution to the problem that’s yet to be solved. Later, as your business scales and you begin to encounter problems with a known solution, you hire people that have previously solved that problem and have them solve it again.”

  2. Really like the idea of keeping a short list of heuristics. You’ve inspired me to start collecting some of my own (will share on my site soon).

    I saw Sam Altman tweet this yesterday and I thought you might like it. I’m calling it the ‘gut’ rule –> great advice from @justinkan: “My advice is flip a coin. If you regret the way it landed, then go with your gut.”https://twitter.com/sama/status/583853101838376960

Leave a Reply

Your email address will not be published. Required fields are marked *