My startup is broken
caseysoftware.com
caseysoftware.com
Schools in the US and perhaps elsewhere DO NOT TEACH STUDENTS about how to handle failure states and anxiety. It is the biggest blind spot the US has and it is biting us in the butt. A big part of the culture is about perfectionism (perhaps rooted in religion - perhaps rooted in kids TV shows - sports culture) - and that needs to be replaced with "blameless continual improvement".
The best performers are those that divorce failure states from emotions and anxiety.
XYZ isn't finished yet - no you're not done.... that's TOTALLY ok. Every homework assignment starts off with a grade of F - you work on it to get to a C, B, or A. A startup's work is never done, there really is no success state, and that's OK too. You're helping to make the product better, you're delivering features to customers. Your making the task pipeline faster with your automation tools. Things suck - but you're making life easier for everyone tomorrow.
The reason why things are so bad now, is because before, things were worse. Its a much more positive way of looking at things for me. Yeah, things can improve, but we've done a lot to make things better.
Also https://www.businessinsider.com/astronaut-chris-hadfield-in-... - there is no problem so bad that you can't make it worse. The best thing you can do is separate your emotions from the situation because they degrade your executive functions.
I disagree with the "quit your bellyaching" sort of conclusions. It gives off a very https://reddit.com/r/wowthanksimcured vibe.
>XYZ isn't finished yet - no you're not done.... that's TOTALLY ok. Every homework assignment starts off with a grade of F - you work on it to get to a C, B, or A. A startup's work is never done, there really is no success state, and that's OK too. You're helping to make the product better, you're delivering features to customers. Your making the task pipeline faster with your automation tools. Things suck - but you're making life easier for everyone tomorrow.
It's worth noting at the same time however that "effort" alone isn't enough. Some ideas really just don't work, either because it turns out it's not something the market actually values, is simply beyond any amount of effort with present technology, oneself/one's organization simply doesn't possess the capability to execute on it (often for perfectly good structural reasons), etc etc. If the homework assignment is "wacky acme item a grand total of 12 people in the world want that will cost hundreds of millions", "factor this 256-bit key" or some equivalent of "develop a working Unified Field Theory" it's unlikely most of us are getting an "A"! Obviously IRL it's not typically so clear, but non-domain expert stakeholders (be it management, users or clients) don't always have any great way to understand themselves that their "simple" requests may be a grabbag of "one of the interns can knock this off in an afternoon", "if the team focuses this can be done in a few months", "if this was Apple/Google/Microsoft devoting massive resources to a cutting edge unit they might pull it off in the next few years" and "ask our a strongly superhuman virtual intelligence descendants in another 50-100 years". It's a job for professionals to help inform what's possibly feasible and what isn't too, and an organization which just says yes isn't healthy. But of course not everyone has the luxury of just packing their bags even if an organization's sales side constantly writes checks development can't cash.
It's also a classic startup trap to spend too many resources on tooling/infra that will make life easier for a tomorrow which will never come because they go under first. Getting the balance right between ugly hacks needed to hit targets cheaply and scaling if the startup works is often super hard. In all this implementation is more important than ideas, and a lot of solid hard work on something junk is still going to fail.
So I think it's also important to emphasize that sometimes no amount of working on it will get from an F to a C, B, or A. There is a time to cut losses, and some initial effort should be expended on getting a feel for basic viability before too much time is sunk so it's possible to pivot away. All while not losing hope about trying again because you can definitely always get that C, B or A in a future effort. No perfect map for that either though. Is grand success just over that next hill, or just a slide down into quicksand? And sometimes a very valuable effort comes out of some side project or individual feature that customers love in an overall product which bombs.
That's an interesting and very important take and something I learned doing semi-professional sports: As a runner (e.g.), you can't go faster, you can only go less slow. Beginning from the very first microsecond after the start, you're loosing time. This especially applies to sports that count points against a maximum.
Same applies to homework: It's not a question about what you can miss and still get a B, the question is always _how much_ you need to do to go get a B. Different perspective.
Some teachers where cool with it and would have me go in front of the class and solve it in the blackboard in real time instead, that was OK. Others hated me for it as it somewhat showed a lack of committment or interest for the classes they were teaching, who knows.
Most of us made an attempt at homework, but it certainly wasn't because it was richly entertaining.
Researching is done best by mostly hearing what your teacher has to say on the topic, staying engaged (until it got repetitive and boring for me) was a way to just get by.
Read the book, Write the thing. I don't know, Lot's of F's and be cool with that
Yeah I wasn't a great student tbh. Best left to my own to do whatever I want
Granted, it's been a long time since I was in school (in the US - as that's what we're talking about - just to be clear) but sports always helped me learn to take failure/loss better than anything else I was taught in school. As far as my son, he's a little young, but it doesn't seem hugely different.
In the real especially high growth world, Theres often no sense of a beginning and an end, I’m always “assigned” more work than I can complete, and end up dropping things. Best is subjective and I may not have time to “complete” it (docs, etc.) not to mention, being taken off for something else halfway through and leaving a 3/4 finished pr behind.
Stop blaming the teachers and start parenting your damn kids. Single parent families broken homes are the leading causes of poor outcomes for children in education and lifelong success. If the parents in America got their shit together, children would do a lot better in life.
It’s hard to blame specific parents when the laws and the participants in family courts (including but not limited to the parents) encourage broken families.
> If the parents in America got their shit together, children would do a lot better in life.
First we need to fix the laws to encourage families to be stable rather than require families to constantly struggle in opposition to the incentives created by the law.
(Not disagreeing I’ve just never heard this perspective before and I’m curious)
For general context, I think it's the case that most relationships will have some disagreements and dissatisfactions from time to time, that are usually temporary and could be resolved in most cases assuming both partners are psychologically healthy and have reasonable conflict resolution skills.
I also think that both partners will from time to time notice what appears to be a more desirable partner than their current partner, sometimes that's an accurate impression and sometimes it is only a temporary impression. In the past, social and legal incentives discouraged married people from breaking up their families at low points in their marriage, and ideally they'd stay together long enough to find out how to work out their problems.
That's not to say that anyone should be stuck in a marriage. Divorce should obviously be possible for lots of reasons, and possibly for no reason, but it shouldn't be *incentivized* in my opinion. If a couple merely wants an unstable, temporary relationship, that's perfectly fine. No one is demanding that they give up that lifestyle if they prefer it, but for those who prefer the idea of marriage as a way of encouraging each other to commit to the relationship in the long term, current laws have removed absolutely every disincentive to break up the family and actually made married relationships more unstable than unmarried ones by creating financial and other incentives for breaking up the married relationship.
1) Spousal support awards to a lesser earning partner encourage that partner to break up the family rather than put effort into resolving disagreements or dissatisfactions in the relationship.
2) No-fault divorce means that anyone can end a marriage at any time for any reason, which no longer seems like what "marriage" was in the past or the vows that are made. If it's just a temporary relationship that can be ended at any time, why call it "marriage" rather than "girlfriend/boyfriend/etc."? Importantly, if there is no fault, but both parties walk away with 50% of the assets (and in practice much more usually goes to one of the sides) then the lesser earning partner in particular, and/or the partner that started with less assets, is at least "not disincentivized" from breaking up the family.
3) Child support awards in excess of the amount needed to raise children encourages the lesser earning parent to break up the family rather than resolve disagreements cooperatively. Note that I wouldn't want to encourage anyone to stay in a truly bad relationship, but specifically the award amounts should not be in excess of the needed amount. To the extent they're in excess, it's effectively a financial reward for breaking up the family.
4) Child custody awards are something like 80-85% of the time to the mother, so if a mother disagrees with the father on how to raise the kids or is having some other reasonable disagreement with him, she can break up the family and know that it's highly likely she'll get the vast majority of time with the kids, and in almost no cases would she get less than 50/50. So rather than both partners needing to learn to cooperate, one has a nuclear option they can use at any time, which imbalances any negotiation around personal boundaries or child raising.
If the government wanted to encourage stable families they would do things like:
a) Let divorce remain an option, but the party breaking up the family should be disincentivized rather than incentivized.
b) Make pre-nuptial agreements ta default that people need to opt out of, so everyone negotiates the terms on the way in instead of expensively on the way out.
c) Promote education about healthy conflict resolution strategies for all couples considering marriage, and possibly in all levels of education throughout life, as it would be beneficial to society as a whole.
They _are_ day cares, and they are responsible for helping children become adults, beyond simply training them in the basic Rs.
Do you think the schools and the teachers aren't a part of the community?
Even the idea that teachers, administration and students are part of the same community is pushing the concept, unless prison guards and wardens are part of the same community as the prisoners.
Or, do we want some acknowledgment of the imperfection of humans and aim to structure a thriving society in any case, with some sense of community and public good? If so, and we acknowledge school as one public good, then why would we arbitrarily draw the line at "academic material"? Couldn't we just as arbitrarily say that parents should strictly teach academic material, while schools teach other values?
I mean, you're here advocating for values that you believe would benefit everyone. Yet, you would want no public or community effort to encourage those values? Seems incongruent.
No. Parents are responsible for basically everything, and schools exist as a place parents can choose to send their children to be educated. Schools and parents are not equal, with a coin flip to decide who does what.
In the U.S., school attendance is paid for by the public and generally required by law. This isn't to say there aren't exemptions for alternatives such as homeschooling.
But, this reflects that we have decided education, in general, is an important public good. So, teaching other valuable skills (like resiliency) is consistent with this principle.
There doesn't seem to be much value in arbitrarily circumscribing which skills should be taught. And, schools teaching a broader array of skills doesn't reassign ultimate responsibility from parents any more than sending them to school does now.
If there’s intense disagreement over the skills and values that should be taught declaring the areas of disagreement to be matters for parents and guardians seems wise. Avoiding exacerbating conflict generally seems wise.
And being from poor background you only get to meet unsuccessful people - so at best you get examples of what you shouldn't do if you want to succeed in life.
(Also, some people worry that writing honestly about a startup will hurt their career, but my book only helped my career. As it became more and more popular, I got more and more offers to consult with various entrepreneurs.)
A start-up is pretty much guaranteed to be a constant source of major stress either because of the success or because of the lack of it. Success will bring its own kind of problems and those can be very stressful indeed. I know lots of founders but I don't know a single one that did not have to deal with stress on a daily basis. Perseverance is often listed as a key requirement for success, I think stress resistance really should be added to that list, it is something you will have to learn how to deal with or it will cause you major grief.
On the flip side, I have also heard fast paced implied to mean "you will work 60-80 hours a week because we have unreasonable timelines and expectations". I passed on those, naturally.
bad answers:
* (interviewer guffaws) * “work hard / play hard” * implying significant overtime
While this is true, I can't help but shake the feeling that adding corporate-planning to a startup-environment would actually calm down things and would still hit targets. I do get a bit frustrated when startup leadership don't think about brokenness because 'that's the way startups are'. I have a friend, big corporate leader; I suspect he'd restructure any startup into something more performant in terms of delivery in less than a quarter.
A smarter founder will know to check assumptions, but it's a case of putting armor where the bullet holes are: sheer persistence is way more important, and where new companies are most likely to fail is in giving up on one or more types of roadblocks and trying to choose only certain fights when they need to persist and come up with some immediate thing that addresses every challenge and change, even if it's not a real or permanent solution or the specific thing they want to be focused on.
The other way to go is, well, the slow way: study the problem, build consensus, look before you leap. While older founders who are bootstrapping will tend to act more conservatively like this, it's not favored by startups of the VC money sort, not because it's actually a worse way to go, but because it isn't playing the game - it doesn't result in huge investment opportunities where a giant market share is captured in a short time frame. If you go in aiming to build consensus you may actually benefit society, but you aren't going to "capture" things. After all, capturing means "I win, you lose".
Thus, a startup that does get to the point where they are rapidly soaking up market share is likely to have a dysfunctional quality of being dissatisfied with the thing the business is currently about: there always has to be a next thing to conquer. Thus another re-org, new hires, repeat until you've attained tech giant status.
And yet we’ve still fallen into some of the same traps any startup in technology falls into. We recently decided to exit the Atlassian ecosystem (prior to their recent issues) and have been doing so while we’re still under the grandfathered pricing for their cloud offerings. We didn’t want to wait until the last minute. We bought ourselves some time so we didn’t want to rush it either.
Just deciding to step back and replace tooling can be difficult. It means you’re tying up a LOT of cycles in internal work, but sometimes that’s a good call to make. We had been using Jira for anything “taskable” but we have both internal/code work and external/billable work, and we weren’t managing either very well in Jira. It’s easy to say “we will lose a lot of history and velocity” but you can export data, and the velocity you lose continuously by not using tooling that works for you, can be much more harmful.
It can be difficult to SEE that you need to step back and reevaluate something as big as your tasking or documentation management. For us, we decided to use this as an opportunity to look with fresh eyes at everything and ended up replacing virtually all of our tools and infrastructure. We even replaced all but one core AWS service in our application and moved everything else to Linode. We replaced our billing system too. We moved from Chargebee, which was vastly over complicated for our current needs to the point where we hated using it, to using pure Stripe and project management with ActiveCollab.
We said no to some business while we did that, but from the inside perspective our lives are much easier, our revenue cycle is more consistent and manageable (and actually -happens-) now that our billables are in a revenue-aware tasking system (ActiveCollab) instead of a pure-task tracking system (Jira).
We’re OK with taking our time because slowly we’ve been seeing just how to build this business effectively, and for us that’s far more important than scale or quarterly returns right now.
I joined one of their companies because I thought exactly like you, that without the chaos the startup could probably be more efficient.
After 3 years in 3 different projects there, I had changed my mind. Without embracing the chaotic nature of a startup, leadership gets too attached to their original plans and ends up taking longer to find product fit. Hiring was fast because they would bring consultants for anything at first, but because their pay was not tied to company growth, they only cared about milking DT cash as long as possible, and so on.
Some of the people that worked there went on to create their own startups, but I am yet to see any big breakout story from T-Labs.
Ultimately, the amount of work you do does not matter if it doesn't get you closer to where you need to be. So, the most valuable thing you can do in a startup is figuring out the big picture. You can't fix all the issues. You are starved for resources and time. So you need to be super picky about which things you actually do and when. In fact, mostly this is about deciding which things you definitely aren't going to waste brain cycles on. Perfect is the enemy of good here. It's a given that you are going to have to take short cuts on some topics.
In practical terms:
- Identify what is really blocking the business. Start removing those obstacles. This requires that you understand both the business and the tech. That's the classical role for a CTO. A good CTO needs to be all over product management, business development, and sales so he/she understands what is being sold and how the product aligns with that. The tech is a means to an end and needs to serve business goals. Why are sales deals falling through? What's missing in the product that makes it not fit the market? Etc. Ask those questions and figure it out. Understand why this is not working out.
- Then write down the ideal end state of your company some years down the line. Basically, define the strategic goals and translate that into short term tactics to get to those goals. Shoot for something that is actually doable and realistic. At least on paper.
- Break it down into things you do not yet have in the product that would work towards that vision. Think in terms of themes that you can strive to improve rather than a long feature list. This isn't a backlog. It's gaps in your product where your lack of progress is holding back the business in some way.
- Prioritize these themes by quarter and year and argue why these are the most important things to tackle next. Basically, you plan a few quarters ahead and you front load the most critical things. This is your roadmap. Realistically, cap the number of themes per quarter to just a handful. "I want it all, I want it now!" is not a plan but a great line from a song by Queen.
- Organize your backlog of features, nice to haves, tech debt, enhancements, etc. by value and difficulty and align with your roadmap themes. Only do those things that are aligned, valuable, and doable and in order of priority. You will never get to the bottom of that backlog. But if keep chipping away at the most valuable things first, you should be making the company better over time. This is what a good product manager should do together with the CTO. In some companies this is one person: a technical product manager.
- Revisit the plan regularly and divide work you are doing day to day into "unplanned things that just have to happen" and "things that make progress on our roadmap". About a 30/70 split of time is ballpark a realistic goal. There's always shit to deal with but you want to keep on making progress on your real goals most of the time. Anything else, you don't waste time and energy on. The benchmark of success is clear visible progress on the themes that were identified as the most critical. If you have extra time, start working on the next most important thing. No reason to sit on your hands.
Get consensus from anybody that matters in the company that this is what the company does. You'd be surprised how many startups don't even have that consensus and are just operating day to day on the whims of some erratic C-level types with short attention spans. Basically, insist on some realism. You're not looking for wishful thinking here but something that is doable and valuable. Beware of CEOs that suffer from magical thinking here that just gloss over all the obvious issues that will ensure the goals are perpetually unachievable. I've seen this in action a few times where they just assume a knight in shiny armour will come along and wave their machine learning sword around a bit and will turn them into the next Steve Jobs. Never happens without a real plan.
Stop reacting, start planning and strategizing. Figure out the hard issues. This is how I've been running my current startup. It's hard. There are so many things to react to and almost all of them are horrible time sinks.
I find year long plans in a start-up environment difficult mainly as the plan needs to adapt more than in a corporate environment.
The highs are really high, and the lows just feel stupid. My boss was in <very important magazine's top young professionals list> last year. The next day, I couldn't find an allen key because someone took the whole rack. One day, the product is on the cover of <well-known scientific journal>. The next day the `develop' branch doesn't compile because someone merged a change to a submodule without updating the other repos.
“ think of your company as a child:
When a kid is healthy, and growing, they outgrow their shoes. You don’t blame the kid. You just find new shoes.”
I have interviewed countless candidates who have a romanticised idea of what a startup is. Perhaps it didn’t help that HR kept using euphemisms like “fast paced” or “extreme ownership” in job postings, I don’t know. Reality is, not many people are ready to work in a constant state of uncertainty. When the company didn’t get the expects funding, uncertainty might as well be seen as almost chaos.
The author softens it a little bit by saying that, in startups, everything is a “variable”, which I believe is, again, an euphemism.
But I guess that it would be way harder to hire if you said the bad parts out loud: in their early stages, startups are messy, chaotic, and you never know if you are going to make it to the next week, regardless of how awesome the product is.
Will this get you to where you want to be faster? Dunno.
This is a key point.
As you do when you run startups, you wear many hats, and one of the hats I frequently wore was that of interviewer.
In that role you very quickly develop a smell-test for the corporate-type.
It is an inevitable fact that the majority of responses from any job ad will come from people whose immediately previous job(s) were in the corporate environment.
That's fine, nothing wrong with that. But you have to structure your interview technique as such that you try to determine if they are more the corporate-type who has become accustomed to the cushy life of the corporate world (teams, departments, tight job descriptions, guaranteed 9–5 life etc.).
The problem with those who have become accustomed to the corporate world is that they'll never change. The mentality becomes too engrained, they seriously struggle and cause headaches for their colleagues. The more years they've spent in corprate-land, the worse it is.
I would often find the classic "do you have any more questions for me?" at the end of the interview to be the most revealing of all as to the mindset of the candidate.
They ask about vacation, perks/benefits etc. ... basically all the fluffy intangible stuff.
The most classic one I remember was a guy had an interesting CV, interviewed reasonably well, then at the end of the interview his choice of question to ask was "where's my office ?".
The briefest of glances at his surroundings whilst being led to the interview room would have made it quite clear there were no private corner offices with a view on offer (it was a small open-plan office, the only two closed off rooms were the interview/meeting room and the kitchenette ... both had glass partitions, so it was clear to see).
Aside from the obvious, the other problem with the question was that the job ad did state this was a primarily field-based role.
Not sure I buy that high turnover is inevitable
> Not sure I buy that high turnover is inevitable
I guess it also depends on the context, e.g. labour market, geography, culture, etc. In my case, my first two were based in Europe, and many hires found the risk/pace to reward ratio unacceptable.
It’s also a matter of how good you are at hiring. I obviously wasn’t, if we look at the results.
To be honest I don't know why anybody would join a startup at today's inflated valuations. Though they seem to be coming down now.
You work much harder just to have a chance to make a FAANG equivalent pay it you get really lucky. I'd only join one again if ownership % was very high and protected from dilution.
I owned a fair amount of my company, joined at low valuation and still came away just roughly equivalent to mid level FAANG comp. Though I was young/dumb and could have negotiated much higher equity, in retrospect.
To be fair I came in with an untraditional background, and the journey/experience certainly was fulfilling in it's own right
A startup I worked in struggled because someone had been given too many shares and then the CEO wanted them out but didn't want to leave equity with someone who wasn't working there any more. If they had planned that in from the beginning and not been so quick to hand out large equity amounts, they could have much more easily fired them and moved on.
Same issue with people who can't do their job for whatever reason. If you don't realise that is a possibility, then you don't think of success measures, targets etc. and so they end up staying too long before you realise they are not working, at which point it is even harder to get rid of them rather than saying up-front, "we expect X within 3 months".
Another blog I read recently said your hiring process is number one priority bar-none and I think I agree. So much of success is down to the right people that you need to consider all of this before your first hire.
If you have any influence over a backlog, roadmap, or list of stories, advocate eliminating anything that's not an immediate concern (bugs are an exception - sometimes it makes sense to purge outdated bugs, but it shouldn't be done lightly). Delete the stories, take things off your roadmap. Almost without exception, they do nothing but generate stress about an insurmountable amount of things to do, more than half of the things won't ever be worked on (which is good! it means you learned something else was more important), and will just lead to people, both internal and external, getting frustrated that "things aren't getting done" even when they are.
One of the hardest things is prioritising dissimilar tasks, especially since the return on investment is not always easy to see or might take a number of weeks/months. Maybe the marketing stats really help marketing target more of the right people but maybe they don't or maybe something unrelated stopped the figures from improving after you did the work.
I suspect someone has done research on how to add real cost/benefit to some of the intangible things to try and objectively rank them.
You have to prioritize on what will a) move the business forward or b) unlock other people/opportunities that will move the business forward.
That could mean improving features, sales funnel, marketing, hiring, or a ton of other things. It really matters on what the risks/opportunities are at any given moment.
Is it that the uncertainty, stress, and more working hours are exciting? Is it actually just the lottery ticket aspect? I know startups have some advantages in terms of building something new and having more impact on the product, but it seems hard to outweigh the day-to-day unpleasantness.
I’ve worked at both and I’ll never go back to a startup. I’m really unclear as to why they appeal to people. Maybe those factors just don’t bother other people that much?
The lower pay isn't low - I'm making plenty for the Bay Area. The stress isn't too high - we have important customers with product expectations, so we test thoroughly before we release changes. I may not be in the weeds writing code every week because I need to wear the product and QA hats too, but overall I love being a high impact player on a team making an impact in a market I care about. Whether the company takes off or doesn't, I'm happier building something with minimum overhead and maximum flexibility.
I’ve never taken a pay cut moving from a larger org to a smaller one.
I feel way more stress being idle than I ever do from external pressure. It’s a personality type.
I think the best way to categorize the startup experience is a choose your own adventure game. Basically everything is variable and it's your responsibility to find what environment best suites you.
Yeah, but in my experience the infighting in small companies exists alongside the work, not in opposition to it. In other words you may have a nasty argument with a cofounder, but when it’s over, you both must sit down and do your work.
In a big company you may have a more mild disagreement with another department, but that disagreement stalls your project for a month.
It's simple. If you want to shape something and grow it from the ground up, doing it at a startup is one of the most likely ways you can do it. Because the team is small, you have to get your hands in a lot of things. Because things are moving quickly, stopping to gold plate things can be dangerous. Because sharing a stressful situation with people can build trust and lifelong relationship.
I've launched new products at established companies and every person has a role and every job has an assignee. If you step out of/beyond your role, you're stepping on someone else and could be ticking off the wrong person. That said, once your piece of the job is done, you can ignore the rest and "failure" isn't on you.
Startups or big companies aren't for everyone but each has benefits and tradeoffs.
Because that's an historical anomaly.
The big companies didn't used to pay 2x-5x more. And, if the current news is anything to go by, that appears to be ending. Get ready for a bumpy ride down.
Things move slow, and then fast. The benefit of so much being incomplete or just not there (lol what observability?) is that we collectively know it’s not there. We know no one can really change it today, so we put more effort and focus into the more pressing need — which is developing the product.
I find pleasure in that ability to focus.
You need best practices because countless other startups have faced some variation of what you are facing. Whether it's hiring, marketing, product, engineering. Collectively, these problems have been solved before. Learn from others.
You need first principle thinking because your situation is still somewhat unique, and best practices may not be best for you. Also, the best practices you hear or read about are riddled in all sorts of biases.
So the first thing I ask a startup I'm advising is around trying to identify where their bias is and help them fix it.
Growth is painful and chaotic.