The key considerations for how much to raise in a seed round are:

• What are the milestones? (e.g. MVP product, iterate with early customers, early team count and makeup)
• Working backwards from there, what will you need to do it? How much runway?
• What do you need from the funds / angels in the round? (e.g. intros, access to portfolio companies, expertise)
• What similar deals are being made and what are those terms?
• What are the additional expenses or timelines that are needed for the product? (e.g. manufacturing, inventory, professional service fees)

Seed investors are primarily evaluating the team and the space in order to make an investment decision. Having an unknown co-founder they would need to work with in the future is a red flag. If you still intend to do that, be up-front and work closely with them in finding the right person. This is important for dilution in future rounds. You want your investors to share in the dilution if you bring on a co-founder later with material amounts of equity instead of diluting just the current founder(s) share.

See also:

• How a seed round works
• Key considerations for how much to raise in a seed round

The following are remote native companies with all employees working remotely.

• GitLab
• InVision
• Lambda School
• Basecamp
• Zapier
• Automattic

An outsourcing firm that typically offers payroll services, tax withholding, HR, and benefits. Employees contract directly with the company (e.g. the offer letter origin is the company not the PEO) and then the PEO enrolls them in payroll and other services. PEOs charge two to seven percent of payroll.

See also:

• Employer of record
• Back office

A payroll partners that is responsible for employment compliance and liability of employees. Companies contract with them to outsource the work of setting up operations in all jurisdictions in which they have employees.

See also:

• Back office

An EOR is much more expensive because they take on all of the liability and compliance on behalf of the company. Company’s use an EOR to hire employees globally where they do not have a local legal entity.

A PEO is a co-employment model and helps the company set up the compliance, tax, payroll, benefits, and HR processes, but tend to operate more as a consultancy. Company’s use a PEO to outsource HR functions where they have their own local legal entity. Most company’s find it too expensive to have a PEO after 100 employees.

See also:

• Back office

When using AWS S3 as an origin to serve a static website distributed by AWS Cloudfront you need to choose how to enable brotli compression. These choices are mutually exclusive—you can either a) have Cloudfront compress to brotli at the edge with a 20% ratio with fallbacks for gzip or no compression or b) compress to brotli by pre-compressing assets where you can use a much higher ratio (80%), but with no fallbacks.

As an alternative you can use AWS Lambda Edge to rewrite requests based on the Accept-Encoding header, but then you need to pay for that in addition to S3 and Cloudfront costs.d

See also:

• The complexity of AWS Lambda stems from the need to internalize the runtime model (you also need to internalize the architectural model in this case)
• Integrating between AWS services adds incidental complexity

Part of Google’s algorithm for indexing content is determining whether a page is useful. One heuristic for not indexing content or un-indexing it if it has already been indexed is whether or not the page is ‘thin content’. Pages that are thin have few words on the page or might be considered duplicate.

See also:

• Understanding SEO
• Effects of note blogging on SEO

Most business registration requires a company to have a real physical mailing address. This is problematic for remote native companies as they don’t have one. Further, most mail forwarding companies use P.O. boxes which are note allowed. To work around this, remote-first startups use the address of the founders and employees in each state or convince someone to let them use their mailing address.

See also:

• Back office

When selling a compliance product, you often need to sell non-compliance; what happens to the customer if they are not compliant. The challenge with that is there are alternatives, for example selectively ignoring the obligation or doing the bare minimum to avoid the worst of it. This can also present challenges with who the buyer is. If a legal and compliance person is the champion, they can get stuck convincing the decision maker (e.g. CEO) and it can take a long time adding uncertainty to the sales process.

See also:

• RegTech

This note does not have a description yet.

When trying to hire someone in a new country, it can take on average two months to put in place all off the infrastructure for employment. This includes research (local laws, regulations, taxes) setting up local payroll (often through a consulting firm familiar with the location) and even setting up a subsidiary company which requires it’s own research into taxes and maintenance.

The most difficult part is determining if it is worth it to set up a legal entity (subsidiary). Key factors include, payroll, sales tax, legal cost of managing entity, and employment risk due to employment law. To decide, it typically involves researching, speaking with an international law firm, getting a second opinion from an employment law firm, talk to a tax consulting firm, and speaking with accounting.

However, once it is set up properly once you can mostly hire as many people you need.

See also:

• Back office
• Remote native companies, especially in the EMEA face this problem very quickly when hiring
• RegTech

Due to COVID-19, employees are spread throughout the country and moving around. This is a challenge for employers who are obligated to pay taxes in the jurisdictions their employees are working. Remote native companies and companies that support a growing remote workforce will continue to face this challenge.

See also:

• Back office
• RegTech

A chicken sexer is a job where, shortly after hatching, someone determines the sex of the chicken. There is no explicit knowledge on how to do that, accuracy is built up through calibrating an intuition in an apprenticeship model (you can only learn to be a chicken sexer by working repeatedly with a chicken sexer).

See also:

• Thought Leaders and Chicken Sexers makes the analogy of thought leaders that can not be proved right or wrong because knowledge is never made explicit
• This is an example of tacit knowledge like riding a bike

A single company operates as the ledger for all stock transactions in the United States. Cede & Co. essentially own all publicly issued stock and stock is traded through contractual rights.

This is a extreme example of 1 of 1 trust model, but surprising in how much power is centralized in a single entity (that must do the right thing).

See also:

• Trust is an important factor to economic prosperity

When building a company and product, having a singular obsession with working on it and solving the problem is an advantage over competitors that do not. The obsession leads to exploring the area in depth, more than any rational person would do. This leads to all sorts of discoveries overlooked by others.

See also:

• Having genuine curiosity and obsession might be a characteristic for dealing with skepticism
• Time horizons as a competitive advantage has a similar in beating short-term speculators
• Naval Ravikant talks about this at length in ‘how to get rich’

Functions that must be called or handled in a particular way can be thought of as having different colors. The programmer needs to understand the color of functions and this adds complexity (and bugs).

This is particularly bad with callback based control flow (either functions return values or nothing). Promises improves things and async / await is significantly better, but you still can’t call an asynchronous function in a synchronous one, forcing you to structure your program around it.

Read the essay

A staggering statistic about the number of prescription painkillers that have been legally flooding the United States. Highly addictive opiates like Oxycontin have created a drug problem that spans all walks of life. To put it in perspective, there are 319 million people that live in the US so nearly 1 opiate-painkiller has been prescribed per person.

Read the essay ‘Down here they sometimes call it boy.’

Many companies have people the work from home occasionally or employees that work fully remote. However, a growing number of startups are starting entirely remote and growing the company with the expectation they will stay a distributed organization.

Examples:

• GitLab (currently valued at $6B)
• Basecamp

Considering there actually are conventional-minded people and independent-minded people in the world, no one considers themselves as conventional-minded. It’s more favorable to think of oneself as independent-minded, perhaps due to the way we value individualism. In Paul Graham’s essay How to think for yourself, he provides an example of this with conspiracy theorists. On their surface they reject some commonly held belief, but the strength of their conviction reveals their conventional-mindedness just within subset of society (not to mention how conspiracy theories tend to be planted by others for political gain).

See also:

• Contrarian dynamic is a predictable pattern of comments reacting to content

Those who care a great deal are the most effective people in any pursuit. It’s difficult to imagine the opposite being true, someone who doesn’t really care about what they are doing being the most effective at their job. Caring is a low-level characteristic that is difficult (or impossible) to fake and does not have any preconditions (caring seems to be a behavioral default not everyone has). It has a large impact on the quality of work and depth of contribution (effectiveness).

See also:

• Complacency comes from mistakenly believing success is assured, but those who care a lot will be the first ones to point out that past success does not guarantee future success.

Companies on a clear upward trajectory can still fail not from external threats, but from complacency of the people running and operating the company. This happens when they mistakenly believe success is assured, but empirically this is not true (plenty of examples of industry darlings getting big and going under).

Complacency is made worse when new people join merely to have a seat on a rocket ship. They exhibit less of the qualities of what built the rocket ship in the first place and causes those that did to also be lulled into complacency (argumentum ad populum).

See also:

• COVID-19 has shown how complacency can have a direct impact on the spread of a deadly virus
• Growing startups are a microcosm of an economy and we can extrapolate the complacency from once fast growing companies to trends in the economy of stagnation and decline

There are four kinds of luck.

The simplest form is extrinsic, something fortunate happens to you—we tend to call that blind luck.

Then there is luck that happens by increasing the number of chances to be lucky—we often see that in people who output an enormous amount and are in constant motion.

The next is luck from being skilled in a particular area and spotting lucky breaks that others don’t notice—this is prepared luck.

Lastly, there is luck that is drawn to you from being in a position where other’s luck becomes your luck—this is due to your unique characteristics, second-order luck. This is the hardest to understand, but consider the example of a deep sea diver who is the best in the world. When a sunken ship filled with treasure is discovered deep underwater, this opportunity will go to them because they are known to be the best.

See also:

• Naval Ravikant who originated this distinction in types of luck
• Lecture - You and Your Research provides a good example of prepared luck.
• Charlie Munger and Warren Buffet are both examples of luck due to unique characteristics (deals come to them, often favorably)

When using multiple AWS services together new problems emerge—IAM permissions and roles, configuration, and load-bearing quirks.

For example, using Cognito and Lambda (to workaround issues with usernames) signing up could fail because the Cognito role doesn’t have the right permissions to invoke the Lambda handler.

Even between multiple instances of the same service, calling one Lambda from another requires configuring the name of the Lambda to call (particularly if you have a separate dev and production version).

Then there are the quirks to discover, for example uploading a file to S3 in a Lambda that sets an acl on the object requires a hard to discover permission and returns a confusing ‘Access Denied’ otherwise.

See also:

• This is a good example of fighting the framework or maybe fighting the architecture