'Should I learn to code?' is the wrong question

New Indie Hackers often ask this question:

Should I to learn to code or can I get by with no-code tools?

I suggest that this is the wrong question. A much better question is:

Am I committed to learning whatever it takes to build a money-making robot?

Let's break that question down.

Firstly, let's understand the goal - most Indie Hackers want to build a money-making robot. They want a machine which will deliver value to customers and collect their money on a 24/7/365 basis. Usually this isn't a physical machine - it's software which directs a CPU to do something with information - gather it, process it, modify it and spit it out to customers in a way that they value. And then, of course, receive money from those customers.

Secondly, you must understand that the path of being an Indie Hacker is learning. Whatever you already know, you don't know anywhere near enough. Not even close.

If you think that you already know everything except a few coding skills, you are horribly wrong. On this journey you'll constantly encounter situations in which you don't have a clue how to accomplish your next task, and you'll have to learn your way through it.

Thirdly, your learning is not upfront - it's continuous. Do not attempt to learn everything you need before you start - you'll waste your time by learning the wrong things.

What you need to learn right now is whatever is needed to perform your next task - the most important task which, therefore, should be done right now. As you perform your next task, you'll learn which task to do after that and how to do it. And so on.

Building a money-making robot means taking on many new and unpredictable tasks and, therefore, it requires learning each task just in time. But don't take my word for it - here's a complete misquote of Jesus which says the same thing:

"Do not worry about learning for your next task, for your next task will worry about itself. Each task has enough learning of its own." (Not) Matthew 6:34

Fourthly, the hardest part of building a money-making robot isn't the building the robot part - it's the money-making part. It's the human-focused work of finding out which problems people actually have (and are so painful that they'll give you money to solve them). This is very different to the problem that you think they have - read 'The Mom Test' to understand the difference.

This means learning things that may not seem interesting, e.g. how to:

• choose a market,
• interview customers,
• design solutions,
• pitch solutions to customers,
• close sales,
• market to prospects,
• onboard new customers,
• satisfy customers and much, much more.

The good news, however, is that these things become more interesting when you actually need them (and can make money from them).

And now, finally, we're ready to look at the question of building your robot:

Should I to learn to code or can I get by with no-code tools?

By now you can probably guess my answer - you will need to learn which tools are available and which tools are best for your next task.

Code and no-code are essentially the same thing - they're tools for telling a CPU what to do. They do, however, offer different trade-offs with respect to:

1. how much you need to learn upfront (limiting your ability to start building a money-making robot), versus
2. your flexibility in telling CPUs what to do (limiting the possibility space in which your money-making robot can exist).

A coding language has a completely open structure and provides a wide range of discrete commands for telling a CPU precisely what to do. This gives you immense flexibility at the cost of upfront learning (how to set up your own coding environment, the basics of coding, how to execute and deploy your code to a CPU, etc.). Then, once you've done that, there are a complex set of commands to learn, coding patterns to master, security practices to understand, deployment options to consider and so on. The learning is literally endless.

A no-code tool is a tool which someone else has coded using one or more coding languages. Their tool aggregates many discrete coding commands into a smaller set of more powerful / leveraged commands which are easy to learn. On top of that, they do the hard work of deploying, executing your code and keeping it secure. They do all of this because it's their money-making robot. The easier they can make their tool to use (i.e. the less learning you need to use it), the more people they can attract to their tool.

That's not to say that no-code tools can't be powerful or flexible - they can. But a no-code tool will always limit your flexibility because:

1. The only way that a no-code tool can become as flexible as a coding language is to become as complex (and difficult to learn) as a coding language. Flexibility comes from having precision in the set of instructions that you can give to a CPU. And the cost of allowing precision is to present the user with a complex system which they must learn. This is why code (and coders) will always be needed - we'll always want the ability to tell a CPU precisely what to do (and coders are the people who learn the complex systems which give precise instructions to CPUs).

2. Every no-code tool is written by a coder who has their own limitations and incentives. Their no-code tool will always be limited, therefore, by their ability to learn which features that users want and by the business case for each of those features (limiting the feature set to those which many users want).

In the very best case, the no-code tool will overcome its flexibility limits by giving you the option to add your own custom code wherever you need it. This gives you the best-of-both-worlds approach in which the tool does the heavy lifting and you add a little code for customisation. I presume that this is the future of every no-code tool (think, for instance, of how much code you can write into an MS Excel spreadsheet).

That future is already here, however, with any mature coding language - it's just packaged in a different structure. Every well-established programming language (such as Python or JavaScript) has an immense array of powerful libraries which allow you to perform complex tasks (sending emails, scraping the web, processing images, getting data from a database, etc.) with very little code. To use those libraries, you only need to set up your coding environment, find the libraries (search for 'awesome python', 'awesome javascript', etc.) and download them. Once you've done that, you need only a few lines of your own code to send an email, scrape a website, modify an image, retrieve data from databases and so on.

Clearly, then, I have an opinion on which is better for you. Learning how to code will give you much greater flexibility as an Indie Hacker. If you can code you won't be reliant on anyone else, particularly not the creators of no-code tools or a 'technical co-founder'. You won't have to guess whether something will work - you can build a quick version and test it. You won't have to obsess over business ideas - you'll be able to test them, learn something new and modify / improve / reject your first idea. The more you learn, the more flexible you will be and the greater possibility space that you can operate in as an Indie Hacker.

But that doesn't mean that you should code everything! You should always use the most effective and efficient tool for your next task, which might be a no-code or low-code tool of any kind at all (e.g. MS Excel or WordPress). Your next task is probably not unique and, therefore, It's highly likely that someone has already built a tool to handle it. Being a coder gives you the option to code, not the obligation to code.

So ... let's bring this all together, starting with two definitions:

Coder [noun]: A person who has committed to learning whatever it takes to build the robot they want.

Indie Hacker [noun]: A person who has committed to learning whatever it takes to build their own money-making robot.

By these definitions, you can call yourself a coder and an Indie Hacker right now.

As you go, remember these things:

• your goal isn't to be a coder, it's to build a money-making robot by any means necessary (with or without code),
• coders aren't people who know and remember everything about coding (they're constantly searching for commands and solutions),
• the quality of your code isn't as important as you think (learn about Pieter Levels whose very ugly code earns him over a million dollars a year),
• the biggest risk as an Indie Hacker is not failing to build a robot (it's building the wrong robot and, therefore, failing to make money),
• start by testing what customers want and will pay for (not by building a robot!),
• do invest a little time learning about any tool which might make you more flexible (but don't over-invest in learning any one thing - your primary goal is to learn what you need, when you need it),
• do invest some time learning about Indie Hacking itself (not just solving your next problem) from the Indie Hacker Podcast and email newsletters from people like Arvid Kahl ('The Bootstrapped Founder').

As for what you should do right now (and every other day), here it is:

1. Learn which task is the most important right now (hint: it's probably human-related rather than code-related).
2. Complete this most important task by learning as you go.
3. Rinse and repeat until you have a robot which makes money.

Commit to learning and go.

Appendix: How to Learn to Code

The only way to learn how to code is:

1. Commit to building a coding project which is tiny and specific.
2. Build that tiny coding project by learning whatever it takes to complete it.
3. Rinse and repeat with larger and more important coding projects.

That's it. Commit to learning and go.

If you need help with the first few steps of learning to code, do this:

1. If you have a friend who codes, ask for their help.
2. Search for information about which programming language to use and select one (I use Python and JavaScript),
3. Download and install your chosen programming language (except for JavaScript, which can run in a web browser console and doesn't need to be downloaded),
4. Choose a program for editing code (called an 'IDE' - I recommend the JetBrains IDE for your language, e.g. PyCharm Pro Edition for Python & JavaScript),
5. Find 'getting started' tutorials for your language to learn the basics (how to write a basic program and run the code),
6. Do web searches for whatever you need to code (e.g. 'how to send email with python'), and
7. Find libraries (search for 'awesome python', 'awesome javascript', etc.) and learn how to download and use them (look for library-specific code examples and tutorials).

Commit to learning and go.