11 ms·
Startup CTO's Handbook
- camel_gopher 3y ago“ and proactively paying down debt is a necessary investment in overall engineering health.” Sometimes it’s cheaper overall to default on certain tech debt rather than pay it down. As long as the product managers don’t try to collect.
- wlll 3y agoThe analogy only goes so far and it's not clear to me what you mean, what do you mean "default on tech debt"? Re-write? Give up and just accept the penalty when changing systems?
- camel_gopher 3y agoBuild quickly with abandon to see if you get product market fit. Throw one away.
- pastacacioepepe 3y agoWhen you finally find product market fit you'll have a hundred other things to worry about before you can even approach a rewrite. Result: you're now stuck with a ton of tech debt and a flimsy product.
- wlll 3y agoYep, this is the majority of my experience, and clawing a way back from that sustainably has been a fairly large part of my consulting/contracting career over the years.
- coffeebeqn 3y agoHaha ain’t that the truth. The rest of the company isn’t going to “throw away” and tread water for 6 months when they finally get traction
- danjac 3y agoI have never, ever in my career ever seen one "thrown away". Unless the project or startup itself dies, of course. Once something is built, and in production in the hands of end users, it is something that needs to be maintained and built on and fixed over time. It's never thrown away. Now if you are lucky, it has been well-written and designed so that parts can be refactored over time as new features are needed and old ones deprecated. If it's shoddily built and rushed because you couldn't afford or couldn't be bothered to do it right the first time, then that's going to be tech debt you are going to have to deal with for years to come.
- Jare 3y agoI've only ever been involved in something close to that "tech debt default" in my career. All other cases of throwing something out were either planned to begin with (so intentionally a throwaway prototype, not tech debt), or just giving up on the product (tech debt included, but who cares at that point). (the case I was in was the game Commandos. We completely threw away all source code for the sequel, which we wrote from scratch with different architecture, tech and even somewhat different coding standards. Getting rid of the tech debt that Commandos accumulated over the yearlong deathmarch was nice, but without the completely different architecture needs I doubt it would have happened)
- ericjohannsen 3y agoAgreed. Sometimes the project is small enough and the debt large enough that a rewrite is in order. However, a project of any size and complexity only reaches tech bankruptcy when the advice given in the book isn't followed; when tech debt is ignored for the benefit of shipping features, which become increasinly hard to release and increasingly prone to bugs.
- amdolan 3y agoStartup CTO’s Cheatsheet.
- gumby 3y agoShould be named "Software Startup CTO's Handbook" There's a lot of good management advice to be had that applies to tech startups in which software is a very small part of the technical product or infrastructure. Such a book can't be written by one person alone, but this could be the kernel of such a book. I've worked in the last 30 years exclusively in tech startups (or as they are called these days, "deep tech", bletch) and there are important model commonalities across software, cloud, networking, pharma, mechanical, consumer electronics, and chemistry as much as there are of course significant differences. The key is the word "startup" which implies "fast moving / high growth" compared to "normal" companies in those sectors, even when it takes five years to get to product.
- zgoldberg 3y agoAuthor here! Such a book can't be written by one person alone, but this could be the kernel of such a book Completely agree, and that's my dream for the book. Should be named "Software Startup CTO's Handbook" I struggled a lot with the title -- good point that it lacks the specific software reference, that should be perhaps be incorporated somehow. "Startup" as well as also important to me, as I agree, size matters and changes how you evaluate trade-offs. Implicit in the word "startup" is that time is your most valuable resource, and much of the advice in the book is based on that assumption.
- gumby 3y agoThese days the language has become so debased that any new business or small business is called a “startup” (and of course they have to be “tech” startups even if they just sell organic oats online — real example). I’d set the ground rules up front so people who don’t want a high-growth business (and I see complains about that on this site which is fine) can stop reading. Likewise if your business doesn’t need the CTO function (even if they are still a four-person startup) then you can stop reading. If your dream for your book is to be about tech startups (in the actual meaning of the two words) I’d partner up with someone else from a different domain relatively soon, then expand the scope again after you’ve got edition 1 done. And call it the “Tech Startup CTO’s Handbook”. What about the folks making semiconductors?
- tptacek 3y agoI don't know any of the companies or people mentioned in this. Does anyone else here? Before I sink any time into reading this: I have an innate suspicion of "chief technology officers" to begin with; is this worth reading?
- mixmastamyk 3y agoWon't know 'till we read it! Or at least a significant portion.
- CSMastermind 3y agoI don't see anything in it that will be new to anyone with industry experience who has made it to the level of entry level engineering management. The target audience seems to be "CTO"s who inherit the title by being technical co-founders at startups with little to no professional experience beforehand.
- xtracto 3y agoMy first reaction was to ask: who is the author?I wish it was someone with good CTO credentials under their, but it doesn't tell me anything. I've been cto at 3 companies over the last 10 years, and maybe I'm getting old and impatient, but I think as others say, this feels more of a book for the just-graduated CTO that finds himself in the role as a cofounder (I do technical advising to those) . In my experience the more you've been in the role, the more specific info you look for, like the Accelerate book.
- paulddraper 3y agoVery good material. (3 time CTO)
- strbean 3y agos/Troughout/Throughout
- zgoldberg 3y agoThanks! Fixed.
- appplication 3y agoIf you’re looking for that sort of feedback: leadershp->leadership in the repo description
- bcantrill 3y agoThere is stuff in here I definitely agree with (record every meeting!), but there is also tons in here that I personally disagree with (e.g., on performance management[0]). This isn't necessarily a criticism -- different problem domains likely demand different approaches and leadership styles -- just to say that I wouldn't take any of this guide as sacrosanct: if anyone reading something in here that doesn't comport with your personal experience, I would recommend deferring to your own experience. [0] https://twitter.com/bcantrill/status/1216491216356823040 https://twitter.com/bcantrill/status/1216491216356823040
- zgoldberg 3y agoJust read through your twitter thread, I really appreciate this perspective and I completely agree, a manager's job includes ensuring the right circumstances to foster intrinsic motivation. I think where I might quibble is there are often circumstances where there are genuine skill gaps and, a manager acting as coach, can accelerate someones growth and performance.
- bcantrill 3y agoSure -- but why drag the entire organization through a perf process to accommodate those folks? Especially when they are tied to comp, performance review processes are (in my experience) about justifying decisions that have already been made rather than actually earnestly trying to improve an employee's growth: decoupling review from comp and then additionally not having a one-size-fits-all review process allows that coaching and mentoring to go where it's most needed -- and it similarly allows for positive feedback to happen much more quickly, broadly, and earnestly.
- user_named 3y agoWhy record every meeting? Are you actually going to listen back to it? Waste of time and energy to me.
- esafak 3y agoIt settles debates about what was discussed, informs people that did not attend, increases discoverability, and promotes transparency. Especially useful in interactions with customers. It's not for your daily stand-up. To quote the inimitable "Yes, Minister" (https://www.youtube.com/watch?v=85fx0LrSMsE https://www.youtube.com/watch?v=85fx0LrSMsE): It is characteristic of all committee discussions and decisions that every member has a vivid recollection of them and that every member's recollection of them differs violently from every other member's recollection. Consequently we accept the convention that the official decisions are those and only those which have officially recorded in the minutes by the officials, from which it emerges with an elegant inevitability that any decision which has been officially reached will have been officially recorded in the minutes by the officials and any decision which is not recorded in the minutes has not been officially reached even if one or more members believe they can recollect it, so in this particular case if the decision had been officially reached it would have been officially recorded in the minutes by the officials. And it isn't so it wasn't.
- SMAAART 3y agoInteresting open source (freemium) publishing business model.
- ronnmico 3y agoReading this book and so far really enjoying it! Easy to follow and very concise. Will update as I finish the book~
- wiradikusuma 3y agoCongrats on launching the book! I'm writing Opinionated Launch (https://opinionatedlaunch.com https://opinionatedlaunch.com) to share practical ideas based on my experience as CTO for a small startup and VPoE in a publicly listed company. When I started writing, I realized there are two distinct topics: management/team/people side of things, and tech side. I ended up focusing on the tech side since that's what I'm more passionate about. I'm glad someone did the other part!
- travissteffen 3y agoI had the privilege of working with the author across two companies, and I can’t tell you how big of a difference it makes when everything he’s written about here gets implemented. The impact is felt across the organization - not just within product or engineering. Like with any system or set of recommendations, eventually you’ll need to make them your own in some way, but I’d definitely recommend utilizing what Zach teaches as much as possible.
- monero-xmr 3y agoBoring technology is so crucial. As a private sector startup company you are already living a fraught existence - why risk it even more with unproven tech? As an example, mongodb inexplicably became a de facto DB for a bit between 2013 and 2016. It seemed like 1/3 of startup companies switched from MySQL or Postgres to Mongodb. What an absolute unmitigated shitshow clusterfuck moronic disaster
- kaptainscarlet 3y agoI remember the days Mongo or NoSQL in general. It's all in memes these days
- coffeebeqn 3y agoEvery company I’ve worked at uses Dynamo. It counts as boring. Most also use SQL - it’s really not difficult to choose the right tool for the job
- monero-xmr 3y agoNoSQL in general is a niche tech. Data is inherently relational. It is very hard for me to imagine a situation where basing everything on Dynamo or Mongo is the correct move, vs. only using NoSQL in specific problems that make sense. Dynamo and mongo generally smell of developers enamoured with specific tools over solving problems efficiently.
- pastacacioepepe 3y agoLiterally every client I've had in the past months uses MongoDB, even though their data is relational. I can't imagine why would someone choose mongo in that case except to be slightly faster in the first few days of development, since you don't have to create a schema and update it. The problem is that you pay back that initial advantage with a lot of interests once your data becomes more complex and you need stuff like BI. Anyway, even if your data isn't relational, JSON in postgres still works better than mongo.
- dilyevsky 3y ago> Wikipedia describes DevOps as a set of practices that combines software development and IT operations… > I translate that as follows: DevOps is all the work that goes into making sure the business software runs in places other than your developers machines. Unless you've got a DevOps specialist on your team, you're probably deprioritizing DevOps to some degree… That seems like a weird interpretation of the original definition. Especially the “devops specialist” part.
- zgoldberg 3y agoGood feedback! I was trying to get at the idea that DevOps is broad, its often invisible and its therefore often undervalued/underprioritized. I'll have another go trying to convey that.
- dilyevsky 3y agoThanks for putting this together! I read the rest of this section and it seems solid, the into however does read like you’re suggesting to compartmentalize “devops tasks” which I think is antithetical to the original idea
- dijit 3y agothe problem with the "original idea" is nobody can decide what it actually means. The interpretation that I see banded about comes from the 10+ deploys a day talk from flickr, where they said having a team of diverse backgrounds was better than having an individual. Somehow in that zeitgeist sysadmins got rebranded devops, so in that case it makes sense to have a "devops engineer".
- coffeebeqn 3y agoIt’s the same as agile and communism. They’ll claim till the end of time that no one has implemented the truest version of it which would solve all of everyone’s problems with magic
- 3y ago
- cientifico 3y agoI recently became CTO, and one of the biggest struggles I have is communication with the CEO. (Sorry if covered by the book, just read the index) On 8 months so far, the CEO: * Don't want to have any alignment meeting. * Don't want to plan anything. * Don't want to drive vision. * Want to focus only in the killer feature. * Don't want to build mockups or prototypes to test with users. * If we do, they have to be beautiful. * Don't want to work on anything that takes more than a week. * Rejected to have any effective meeting. </Frustration> What I want to say in this message is that I miss exactly the things that are not in the book. Edit: we are just 3 cofounders!
- yowlingcat 3y agoI'll relay a quote from my mother -- "If you're feeling bad about this. Don't worry. It will only get worse! Hahahaha! (phone click)" In all seriousness -- what you're running into is exactly what makes the job challenging. Here are a couple places to start, having been both in your shoes, and the shoes of those on the other end: 1. Executive 1:1s -- who are the other executives you work with outside of the CEO if any? Try to get 1:1s scheduled so you can get some more context about what you're walking into. Get a sense of what their needs are, which will give you a sense of what the broader organization's needs are. Remember, you're being hired to make the CEO's problems go away /without/ their asking you. Being able to integrate into their executive team is one of the best ways to prove you know what you're doing, and all it takes is some consistency, diligence, an open mind, and a desire to earnestly ask helpful questions. 2. Executive alignment meetings -- Is your CEO and executive team meeting regularly? If so, ask to join. If not, take the initiative to plan it -- ideally after you've aligned with the other executives beyond the CEO. Even if the CEO cannot make every (or even most) of their meetings, they'll greatly appreciate the initiative and it will build credibility. I know you mentioned they don't want to have these meetings, so it may be required to work indirectly through other executives first, based on who has the CEO's ear. 3. CEO 1:1 -- if you're the CTO, your CEO likely wants to have some degree of a regular 1 on 1 catchup, if at least to make sure you don't leave. You can figure out a cadence to make this work on said CEO's calendar -- whether it's once a week, biweekly, monthly, etc. The purpose of this meeting is to ensure that you are able to align with the CEO and collect feedback from them about what areas you're doing well on and around what areas you can improve. If you can't get this scheduled, I don't think you're in a place where you're set up to succeed and I'd advise you to leave. If it's hard to get this scheduled, try pushing on the first two approaches first to get the rapport to get here. Most importantly -- if you can't get any of these scheduled, you should really re-evaluate whether this is the right role and whether you shouldn't move. If you are the CTO and you cannot get some kind of regular time, no matter how infrequent, with the CEO and other executives, you are not really the CTO and you should consider leaving. Just my 2c.
- TruthWillHurt 3y agoSo little about tech, and what that's there is mostly how to annoy your devs with trivial corp-like control nonsense. Too many people in startups try to "play house", pretending to be big shot managers/company rather than develop the product and actually moving the needle (i.e hiring an HR manager as employee #3-5). How about you do the work, keep to a pizza size team in a garage until you actually can't?
- TeeWEE 3y agoAgreed, if the company is small keep the process light and focus on good people and a good product. This includes a good DevOps “product” side of things. Personally I do like having a documenting culture so things can easily by found. In the current startup it’s often said: code is the documentation. But to me that’s just lazy. Especially if the code doesn’t have a minimal set of inline docs.
- ericjohannsen 3y agoMy advice: Understand all the tools. Apply the ones called for in any given situation. I agree that you want to keep your pizza-sized team focused on fast iterations and on shipping ideas to see what works. There's a huge transition in management skills when that team succeeds. I've seen many companies flounder and fail when they don't recognize how things have changed. My current startup is larger (a little over 100 people) and almost at break even. We just acquired one of those pizza sized companies. It's been a fun challenge keeping their velocity high and enabling them to focus on shipping stuff while integrating them into our more structured ecosystem. Use the right tool for the right job at the right time.
- serial_dev 3y agoIt might be just me, but I like having meetings recorded, and not for having ammunition in my hands when the team starts the blame game, the he said she said. I joined a team where some meetings are recorded and I really like being able to re-watch meetings and discussions. Let's assume there is an important discussion, an alignment where I really stretch my capabilities. I can't pay attention to the topic and take notes/memorize everything at the same time. If I know there is a recording, I can fully dedicate the meeting (where other people are present) to paying attention, asking good questions, challenge their assumptions and so on. Then, the next day, I can listen to the call, pause, increase speed 2x,3x, take notes for myself, write minutes, write summaries for internal use (casual slack message), notice inconsistencies in our thinking, or run an AI note taker all on my own schedule. And let's not forget when someone is sick, need to pick up a mother in law from the airport, have a parallel meeting, or joined a week after an important alignment. If the meeting wasn't that important? No problem, just ignore it. Set up a sane auto delete rule, and that's it. In practice, I don't think people spend time snooping around where they shouldn't, so that isn't really an issue.
- dijit 3y agoThis, entirely this. I moved to google workspace precisely because of the ease in which you can record and retrieve recordings of meetings. (IE; it's embedded into the calendar event). If a meeting is not important enough to be recorded then it probably wasn't a meeting worth having. My personal position is that I put more hours into the day than other people, but it's not synchronous. I might be in a more important meeting when there's another important meeting happening; a lot of nuance gets taken out if you just read meeting notes (which people seldom do also) - but even there it's so helpful to record a meeting so that you can properly revise meetings notes quickly. I really push quite hard for having meetings recorded.
- ericjohannsen 3y agoWe moved away from Google workspace because we were acquired (miss you, Google!), but I'll add for the benefit of those in the MSO ecosystem that Teams can also caption recordings.
- diamondfist25 3y agoRecord meeting is such a pre-AI paradigm. Now u just need to transcribe meetings, summarize, and extract action items Done
- rolisz 3y agoI find the breakdown into 3 types of CTOs (tech, people and external oriented) very interesting. A very early stage startup (no product yet, just some basic demos, doing pre seed funding) is courting me for a CTO role. What should I expect is most needed for this role? A tech CTO? A people CTO? I expect that it should mostly be a tech role for the beginning, because we would have to develop the product. But I wonder how soon would the transition be toward a people CTO? Does anyone have any insight/experience with this?
- svilen_dobrev 3y agothe question might be the other way around. #1 What You want to be? (now, but do reevaluate later ~1year into it) The tech guy? The people guy? The <other aspect123> guy? Be mostly that one, keep other stuff more-or-less afloat, and when/if that work gets too much/critical, hire other people for other 2++ aspects. But there is another question which is harder to answer: #2 What do they/others CXO want you to be? Because this can be/become obstacle to #1. And, in some cases, it's CTO without the C. Just some tech guy who is told what needs be done, runs bragging-shows and is never told the "why's" about business. Here few links i've bookmarked on the topic "WTF is a CTO".. http://www.startuplessonslearned.com/2008/09/what-does-startup-cto-actually-do.html http://www.startuplessonslearned.com/2008/09/what-does-start... https://www.allthingsdistributed.com/2007/07/the_different_cto_roles.html https://www.allthingsdistributed.com/2007/07/the_different_c... and also the book Manager's Path by Camille Fournier. have fun
- rolisz 3y agoThank you for your answer! The founder wants me to both build the product and eventually grow the technical team.
- mleonhard 3y ago"Section 3.4 Testing" is missing an explanation of why testing is necessary. Some startup CTOs make critical errors like un-staffing testing or migrating to Service-Oriented Architecture without integration tests. We need the handbook to instruct such people on why testing is a necessary part of the software process.
- joestolte 3y agoSuper helpful, thanks for sharing!