[elmarschraml]

Depends a lot on your situation:

Does "CTO" mean you are the tech lead of a small (single team) engineering organization? Then everything written for staff engineers applies. E.g I've heard good things about "Staff engineer's path" by Tanya Reilly.

Does "CTO" mean you are leading an org that is too large to be hands-on with tech, and need to build an effective structure and culture? Then I second the recommendation for "an elegant puzzle" by Will Larson.

Or does "CTO" mean that you switched from being an engineer to managing a team of engineers? Then everything for new managers applies, for starters I'd recommend "Becoming an effective software engineering manager" by James Stanier, or "Engineering management for the rest of us" by Sarah Drasner.

For some good general material, I'd also recommend the resources that Gergely Orosz makes available for subscribers to his "pragmatic engineer" newsletter. Those are templates for the kind of documents and processes you will most likely need - if you're new to the role, you will not go too wrong by using them, and if you want to create your own they are excellent starting points.

[skrebbel]

And maybe CTO means “the founding programmer of a new startup” in which case my advice would be to stop reading and start coding :-)

[946789987649]

There's always time to read, you can't only code. E.g. I run a lot so I get through a hell of a lot of audio books.

[ysavir]

I don't think the parent comment means only code, never learn. I think their intention was to not worry too much about the "CTO" title and instead focus on building the product at hand.

[skrebbel]

Yep. When building a startup you’ll find a lot of well-meaning people (in books or in person) who really do nothing but distract you. YC got it right, you should make something people want. Everything that’s not either making something or finding out what people want is just going to slow you down.

[leetrout]

Too reductionist.

You must "sharpen the saw".

This takes many forms including adequate learning / training for self improvement as well as investment in your tooling that will pay dividends on delivering faster or with higher throughput.

These are second and third system effects that require intention to monitor or measure but the effect is real.

[skrebbel]

Of course nothing is black and white, but if you have limited time/money then you want to get to ramen profitability fast and you simply don’t have time for business model canvases, the perfect employee option scheme, a scalable k8s setup or a perfect CI/CD pipeline.

There’s so much stuff that feels important and valuable, but so little of it really cannot wait until after your wheels are off the ground.

When you read postmortems of startups that didn't get enough customers, often it’s this stuff that actually went wrong. Too much time spent on other stuff than “build something” and “that people want”.

To my experience, it’s difficult to resist all that good advice that’s all over the internet, books, accelerator programs and the like, and saving it all for later. People will tell you “you should get $PETPEEVE right from the start” for every imaginable pet peeve (all the way from legal stuff to unit tests to SEO) and they’ll be very convincing. Trying to resist this is not reductionist, it’s super hard.

[ysavir]

Sure. But the point was that if you're a team of 1 engineer, even if your title is CTO, time spent learning "how to be a CTO" when you don't have any team whatsoever is time not spent building the prototype that needs to demonstrate $XX value/growth by YYYY-MM-DD in order to survive.

Learning is always encouraged, but you should be learning to solve the problems you're about to face. Learning to solve problems which aren't going to be obstacles in the near or mid future isn't helping you in your immediate circumstances. Sometimes the immediate circumstances aren't much of a concern, but when they are, you need to be sure you're learning with that in mind.

[dasil003]

Any advice is going to be reductive when measured against the reality of building a startup. While learning and continuous improvement are table stakes, I think the GPs advice is much better to get first-time founders overall. "Sharpening the saw" is likely to feed perfectionist tendencies, or send them into dopamine learning loops online. There's a reason that "bias for action" is treated as a highly valuable trait in founders.

[Quiark]

I'd say the time for sharpening the saw was before you started the company. The next opportunity to invest into best practices / long-term is when the company actually has some traction.

[endymi0n]

And honestly, pre product-market fit my advice would be not even coding, but iterating in an extremely scrappy way (read: Google Sheets, Jupyter Notebooks, no-/low code, handwritten invoices) until you have people paying for your product and not churning after the first month. Everything else comes secondary. For every company that didn‘t manage to scale their tech and teams fast enough, there‘s 100 that die by „building and they will come“ a polished app with great UX through 4 ecosystems that get launched to deafening silence.

[skrebbel]

Tbf I believe this depends a lot on the product category. Uber could do this in their first few months but eg Retool couldn’t.

[macNchz]

Obviously some ideas can’t be tested without writing some code, but it’s more of a mindset than a strict playbook–many programmers (myself included) tend to bias towards wanting to write code and avoid sales conversations until the thing feels “done”. Tech message boards are littered with programmers trying to justify why their app has to have a full polished build that will take a year before they can begin selling it.

When you get in the habit of asking yourself, before building, how you can learn the most about what people want with the least amount of code, clever ideas often come to mind.

[robertlagrant]

Heh I was going to mention Google sheets. Great database that doesn't scale, but before you're close to worrying about that you have a lot of answers.

[fullstackchris]

this is perfect advice which i can't agree with enough. that is unless the founders insist on a flawless UI... then you are in for a world of pain

never understood why teams of literally 1 or 2 devs are always forced to try attempt to acheive the visual quality of apps like instagram, whatsapp, etc

[liquidpele]

> never understood why teams of literally 1 or 2 devs are always forced to try attempt to acheive the visual quality of apps like instagram, whatsapp, etc

For the same reason those kinds of places have more managers than workers… it’s a vanity project.

[mooreds]

Haha, for sure. I wrote about that contrast here a few years ago: https://www.mooreds.com/wordpress/archives/2555

[belter]

His productivity will shoot trough the roof, when it moves into deleting code.

[techbiotic]

:-)

[Aurornis]

> For some good general material, I'd also recommend the resources that Gergely Orosz makes available for subscribers to his "pragmatic engineer" newsletter. Those are templates for the kind of documents and processes you will most likely need - if you're new to the role, you will not go too wrong by using them

One word of caution: Engineers and EMs read these newsletters and even management books, too.

It’s not hard to see when your leadership is just parroting things they read from a book or newsletter and trying to pass it off as if they had deep leadership experience. Engineers are good at seeing through cargo cult leadership.

I suggest that if you embrace a practice found in a book or a newsletter, be transparent about the source. Encourage people to read more about it in the book or newsletter where you found it. Don’t try to pass it off as a management technique you invented or pulled from deep background experience, because it feels disappointing to the team when someone realizes their CTO is just parroting things from a newsletter or blog and trying to hide the source.

Also, don’t get into the mindset that having read a lot of books or blogs or podcasts or newsletters is a substitute for experience. Some of the worst leaders I’ve had were those who would lord their book knowledge over everyone as though it made them the confident expert on everything, while we could all clearly see that the extents of their knowledge started and stopped with what was contained in the books they read. Books contribute bits and pieces and give suggestions to add to your corpus of knowledge, but if you treat them like definitive, all-encompassing sources of wisdom then it’s going to be clear to people that you don’t really know what you’re doing. Be honest and stay humble.

[onetimeuse92304]

That's a very good point. A "CTO" can mean any mix of managerial, leadership and technical responsibilities.

It is my assumption that a CTO is usually coming with technical background already except when he/she is a cofounder and they got technical domain as a result of division of responsibilities. But I am suspicious of startups where none of the cofounders come from technical background.

That leaves managerial and leadership. IMO, if you are a CTO, managerial is something you can hire other people to do for you as your technical organisation grows and leadership is something you should really be focusing on.

You can have flourishing technical organisations with the CTO being a poor manager but a good leader but very unlikely with a CTO with good management but poor leadership skills.

[mac-mc]

Even if you're effectively 'just' a tech lead of a small team, like I am, it's still a very different job than being a tech lead of a small team in a larger company. You are at the top level of the small group, so you end up being involved in a lot more non-tech things than you think you would be.

Your peers are not other engineers and engineering managers like you would be in most situations, but marketers, product people, designers, PMs and so on. Who is the founder and ultimate authority (CEO) and if you are a co-founder or not also really changes things. A founder CTO is also very different from a non-founder CTO. Co-founder conflicts is a huge thing to manage, (see https://flocrivello.com/co-founding-considered-harmful/ ) because at a small size even as a non founder, you're pretty close to it. You should decide upfront if you're going to approach the job as another effective founder or employee and communicate that to founder at the start as part of a conversation.

I made the transition from staff eng to eng manager to CTO at a small startup, and each one is still a very different job even though the skill sets transfer greatly. I code way more at the startup than I did as an eng manager too. Several things I would suggest:

1. You're an exec, even if your company is tiny, and how it works is different than being a manager. Concepts like the 'first team' https://www.michaelvizdos.com/resources/first-team and 'getting on the balcony' matter way more than when you're in a purely product & eng organization: https://www.bettermanager.co/post/move-away-from-the-dance-f...

2. Don't be shy about getting exec coaching, and having the company pay for it. We did that and it was very helpful.

3. Read Radical Candor (updated edition!!!), but also read it with a bit of a grain of salt, realizing the title and advice was chosen because the writer wanted to balance their general high agreeableness with a counterweight that goes too far for the typical person IMO, which the writer writes about that exact dynamic in the second edition.

2a. Learning to be properly direct is super essential, especially at small sizes where its more make or break based on how emotionally deluded you are. The people you are working with need be able to handle you being real with them too. If they cannot, move on, it's that important.

4. Expand your skill set, do everything technical. If you're a backend engineering manager, expand your stack to all parts of the business. That means the back end, the front end, data, analytics, observability, devops, AI models, performance tracing, etc. I started as a mobile eng with previous backend experience, and I wish I expanded sooner so the company had those things on lock sooner, especially in the data analytics side. Engineers hate analytics, embrace it fast to counteract that tendency.

5. Gergely, Will Larson, etc are good resources, but also note they come from what I call the "Uber strain", and thus structure a lot of their experience and advice based on formative years working at Uber. I also worked at Uber when they were there and I recognize a bunch of Uber-isms in their writing.

A good chunk of what they talk about is how to be successful at a place like Uber. Something I've realized later with them is that other companies can be pretty different, and what they espouse sometimes might not match. But on the other hand, Uber's work culture came from previous strains of silicon valley tech culture, especially google's due to the top engineers being from google and setting up their promo system to match googles (vs lets say apple) and the knock on effects of that. Something about that place made people get really good at understanding what is needed to succeed in tech leadership and write about it, I guess it's kind of a paypal mafia effect.