5 ms·

We hear about co-founder dynamics a lot in the startup community, and I wonder if the traditional business/technical dyad exacerbates conflict? I am a technica

We hear about co-founder dynamics a lot in the startup community, and I wonder if the traditional business/technical dyad exacerbates conflict? I am a technical co-founder, and that means I build things. In a conventional co-founding relationship, I am be expected to give my business-oriented partner the final word with respect to strategy or direction, and he defer to me on how to build the product. At the outset, this puts all the power in technology. Nothing is more important than getting the right product built and shipping on schedule, execution is everything. Unfortunately, at any significant scale it's all strategy. Technology increasingly becomes a cost centre; either it works or it doesn't, either it ships or it sinks. Strategic decisions begin to lead technology choices, and the CTO drifts away from the locus of power. Even if she's explicitly invited to participate in strategy, the buck stops somewhere she is not. Influence, and with it ownership and personal investment, whipsaws from one founder to the other with neither realizing it explicitly. Ultimately, technical founders almost always drift away (as we've seen most recently at Buffer) when the company hits scale and they're fully vested. And that sucks - any new blood won't be as naturally invested in the product without further dilution, will lack history with the team, won't know where the skeletons are buried, etc. I hypothesize that balanced founding teams - two business, two technical, sales & business, etc. would have more staying power in the long run. Cases where there is no default division of labour and influence. I believe I've seen this pattern, but without a neutral data set I can't trust my biases. YCombinator has enough data now to figure it out... (\nudge nudge\)

mstachowiak 10y ago
Interesting hypothesis. A quick look at YC's top companies by market cap seems to back it up: Airbnb: 3 design centric cofounders DropBox: 2 tech cofounders Stripe: 2 tech cofounders
- sagivo 10y ago
  the professional skill-sets are secondary to personal chemistry. if the founders can't communicate it doesn't matter that they complete each-other's skills.
- argonaut 10y ago
  Airbnb was 2 designers, 1 technical.
wuschel 10y ago
I would, too, like to see some hard data on your hypothesis, although I believe that there are too many inseparable human and enviromental variables at play here. At the end, the change in role and power is a very human development (see also: families, partnerships, child-parent relationships).
jacquesm 10y ago
I think a bigger predictor for co-founder conflict (and, conversely, eventual success) is a common background and similar financial situation. If founders are from very different backgrounds, have very different home situations or have strongly differing financial situations conflicts are almost certain to arise, much more likely than simply because of a tech/business split in responsibilities.
- finid 10y ago
  I think greed is a major issue.
agibsonccc 10y ago
That vastly depends on the kind of company though. Most YC startups especially are about automating some sort of inefficency in the real world (see: paper based workflows). I know YC is getting in to hardware and other areas, but most startups they are likely to fund are going to be an efficiency play. It's what they know best statistically. Moonshots seem like they would go through YC research. There are exceptions to this rule like Boom which was in my batch. By definition many startups aren't actually about "tech". Yes as companies grow they will have mild scaling problems, but what determines whether users stay or not ends up being design or a big focus on a problem being solved. The one caveat I would mention here would be infrastructure startups (disclaimer: I run one and am therefore highly biased) where you have things like databases,containers,AI platforms,.. where the technology vision matters in terms of appealing to a technical user base or where merits are based on the technology solving a real problem. That being said - even something as "hard" as an AI startup: The model building isn't even the bulk of the work, after you build that it's mainly about UX, which firmly puts them in the same bucket as most other kinds of companies, even to the point of these companies just "adding AI" so they can raise some money. That being said: Statistically startups are hard enough as it is, and the bulk of them to be efficient businesses are mostly going to fall in to the bucket I just described: using a bit of software to make a process in the real world more efficient.
- tedmiston 10y ago
  > The one caveat I would mention here would be infrastructure startups (disclaimer: I run one and am therefore highly biased) where you have things like databases,containers,AI platforms,.. where the technology vision matters in terms of appealing to a technical user base or where merits are based on the technology solving a real problem. Lately I've also heard these called tech-first startups (disclosure: I do contracted dev work for one). At first I thought it was a redundant, unnecessary label, but given your contrast to efficiency play style startups, it feels like a useful and concise descriptor.
  - agibsonccc 10y ago
    I'm mainly aware of this distinction because of a unique problem with AI startups. Technically an AI platform and an AI product company (we call this vertical specific) are both "AI startups". There is a larger trend in the industry that typically these "platform" companies are a solution looking for a problem (this is often the case). These companies are typically SAAS and get acquired. What you see more of now a days are AI "product" companies focused on a particular problem like say: sales ops. The "platform" people seem to have moved on to "chat bots" now. It seems like the same trend will play out though: "chat bots" for specific verticals. That being said: As a founder of a startup that somewhat bucked this trend (focusing on an oracle type business model in various markets rather than SAAS and selling to developers) we focus mainly on a few key areas to make money. The way we do this is "core technology" bundled with some sort of an "app", in our case being fraud detection solutions with a "land and expand" type approach. The thesis is we can see some of the benefits of being a tech-first startup (lock in and broad usage) but keep focused like a product company would be.
mattmanser 10y ago
Unless I'm reading the blog post wrong[1], you've got it totally wrong about Buffer, Joel made Buffer 1.0 and was the technically minded co-founder at Buffer, not Leo. Sunil might have been an early hire, but we're talking a year after company founding. Joel + Leo took a while to grow it. I remember watching a talk by Joel before he'd even brought Leo on board in Nottingham, UK (Joel was starting up in Birmingham, UK if I remember correctly), about his previous products/failures and how he'd just built Buffer and the problems it solved. I think I saw a post about Leo joining 3 to 6 months later, at the time a recent uni grad compared to Joel's more experienced dev + freelance dev background. I've always been pleased to see Joel succeed, as he's a great guy, but I have to admit some of the posts they came out with have ultimately proved to be utter b%!!$hit. I remember calling them out on a post about how the UK was a bad place to startup, I've seen a couple of $100+ million buyouts of startups here since then just in Nottingham. Then they claimed bootstrapping was best, then raised. Claimed a flat structure was best, then go a traditional management structure, etc., etc. I sometimes felt they were just echoing back to the startup community what the community wanted to hear, and to be fair, it really worked, they got lots of upvotes on HN, etc., so great marketing out of it, and good traction. There was a time when like clockwork they'd have a high-rated post here every Sunday (or Monday, can't remember). [1]https://open.buffer.com/change-at-buffer/ https://open.buffer.com/change-at-buffer/
- DelaneyM 10y ago
  Ah, thanks. You're completely correct - their CTO did leave, but he wasn't a founding CTO.
- borplk 10y ago
  Maybe it's just me but posts like that you linked fill me with a toxic feeling. I don't know how to describe it. It just smells so fake and diplomatic. Plastered with photos of smiling people in different countries and happy-go-lucky company culture bullshit. I don't know how I could move my body out of the bed if I were to work in environments like that. A "company retreat" is a nightmare in my eyes. Sigh. I'm a pessimistic loner of an introvert and I don't need your bullshit in my life. I wish there was some more room for us on this planet. Every corner you look it's either filled with extroverts or fake extrovert wanna-be rockstar-templayers. </vent>
azernik 10y ago
I'm co-founding with someone with that dynamic (I'm handling product, she's handling business). Luckily we know and trust each other from way back, so tension isn't a problem, but there has been an interesting power dynamic - originally I was more in the driver's seat, since all we were doing was product development and there was no real business decision to make. Now that we're actually selling to customers (and investors), she's more in control of things, ie is more qualified to decide what features are important for business reasons.
- finid 10y ago
  I'm in the early stages of starting a startup, but I think if people focus on whether a decision being made is good/bad for the company rather than who's in charge, things would be a lot easier.
edblarney 10y ago
Those are very good points. But I think that a 'division of responsibilities' issue may miss the most important dynamic and that is simply the 'relationship'. If you get a long well, can handle disagreement, share a similar vision for the company, and have a 'history' together (which I think builds trust) - then those 'divisions' become less obvious. Sure - maybe WRT some issues, 'one of the two' can maybe have more influence, but I think the dynamic is more personal, and blurry. I wonder if it helps to have one of the co-founders simply have more authority - i.e. a 70/30 type situation wherein if the 70% co-founder does not abuse that, and stays pretty 'equal' for most things, and then hints at 'final call' in only a very few things ... that might work because it could help the 30% co-founder to concede. It'd be interesting to see some data points on that.
mememachine 10y ago
My biggest take away is that you use "she" by default.
rorykoehler 10y ago
How is it possible to divorce strategy and direction from product/tech? This siloed approach is almost guaranteed to bring problems.A CTO is someone who is essentially a technical CEO. They need to be involved in strategy just as much as the CEO for there to be a healthy co-founder dynamic.

mstachowiak 10y ago

Interesting hypothesis. A quick look at YC's top companies by market cap seems to back it up: Airbnb: 3 design centric cofounders DropBox: 2 tech cofounders Stripe: 2 tech cofounders

sagivo 10y ago
the professional skill-sets are secondary to personal chemistry. if the founders can't communicate it doesn't matter that they complete each-other's skills.
argonaut 10y ago
Airbnb was 2 designers, 1 technical.

wuschel 10y ago

I would, too, like to see some hard data on your hypothesis, although I believe that there are too many inseparable human and enviromental variables at play here. At the end, the change in role and power is a very human development (see also: families, partnerships, child-parent relationships).

jacquesm 10y ago

I think a bigger predictor for co-founder conflict (and, conversely, eventual success) is a common background and similar financial situation. If founders are from very different backgrounds, have very different home situations or have strongly differing financial situations conflicts are almost certain to arise, much more likely than simply because of a tech/business split in responsibilities.

finid 10y ago
I think greed is a major issue.

agibsonccc 10y ago

That vastly depends on the kind of company though. Most YC startups especially are about automating some sort of inefficency in the real world (see: paper based workflows). I know YC is getting in to hardware and other areas, but most startups they are likely to fund are going to be an efficiency play. It's what they know best statistically. Moonshots seem like they would go through YC research. There are exceptions to this rule like Boom which was in my batch. By definition many startups aren't actually about "tech". Yes as companies grow they will have mild scaling problems, but what determines whether users stay or not ends up being design or a big focus on a problem being solved. The one caveat I would mention here would be infrastructure startups (disclaimer: I run one and am therefore highly biased) where you have things like databases,containers,AI platforms,.. where the technology vision matters in terms of appealing to a technical user base or where merits are based on the technology solving a real problem. That being said - even something as "hard" as an AI startup: The model building isn't even the bulk of the work, after you build that it's mainly about UX, which firmly puts them in the same bucket as most other kinds of companies, even to the point of these companies just "adding AI" so they can raise some money. That being said: Statistically startups are hard enough as it is, and the bulk of them to be efficient businesses are mostly going to fall in to the bucket I just described: using a bit of software to make a process in the real world more efficient.

tedmiston 10y ago
> The one caveat I would mention here would be infrastructure startups (disclaimer: I run one and am therefore highly biased) where you have things like databases,containers,AI platforms,.. where the technology vision matters in terms of appealing to a technical user base or where merits are based on the technology solving a real problem. Lately I've also heard these called tech-first startups (disclosure: I do contracted dev work for one). At first I thought it was a redundant, unnecessary label, but given your contrast to efficiency play style startups, it feels like a useful and concise descriptor.
- agibsonccc 10y ago
  I'm mainly aware of this distinction because of a unique problem with AI startups. Technically an AI platform and an AI product company (we call this vertical specific) are both "AI startups". There is a larger trend in the industry that typically these "platform" companies are a solution looking for a problem (this is often the case). These companies are typically SAAS and get acquired. What you see more of now a days are AI "product" companies focused on a particular problem like say: sales ops. The "platform" people seem to have moved on to "chat bots" now. It seems like the same trend will play out though: "chat bots" for specific verticals. That being said: As a founder of a startup that somewhat bucked this trend (focusing on an oracle type business model in various markets rather than SAAS and selling to developers) we focus mainly on a few key areas to make money. The way we do this is "core technology" bundled with some sort of an "app", in our case being fraud detection solutions with a "land and expand" type approach. The thesis is we can see some of the benefits of being a tech-first startup (lock in and broad usage) but keep focused like a product company would be.

mattmanser 10y ago

Unless I'm reading the blog post wrong[1], you've got it totally wrong about Buffer, Joel made Buffer 1.0 and was the technically minded co-founder at Buffer, not Leo. Sunil might have been an early hire, but we're talking a year after company founding. Joel + Leo took a while to grow it. I remember watching a talk by Joel before he'd even brought Leo on board in Nottingham, UK (Joel was starting up in Birmingham, UK if I remember correctly), about his previous products/failures and how he'd just built Buffer and the problems it solved. I think I saw a post about Leo joining 3 to 6 months later, at the time a recent uni grad compared to Joel's more experienced dev + freelance dev background. I've always been pleased to see Joel succeed, as he's a great guy, but I have to admit some of the posts they came out with have ultimately proved to be utter b%!!$hit. I remember calling them out on a post about how the UK was a bad place to startup, I've seen a couple of $100+ million buyouts of startups here since then just in Nottingham. Then they claimed bootstrapping was best, then raised. Claimed a flat structure was best, then go a traditional management structure, etc., etc. I sometimes felt they were just echoing back to the startup community what the community wanted to hear, and to be fair, it really worked, they got lots of upvotes on HN, etc., so great marketing out of it, and good traction. There was a time when like clockwork they'd have a high-rated post here every Sunday (or Monday, can't remember). [1]https://open.buffer.com/change-at-buffer/ https://open.buffer.com/change-at-buffer/

DelaneyM 10y ago
Ah, thanks. You're completely correct - their CTO did leave, but he wasn't a founding CTO.
borplk 10y ago
Maybe it's just me but posts like that you linked fill me with a toxic feeling. I don't know how to describe it. It just smells so fake and diplomatic. Plastered with photos of smiling people in different countries and happy-go-lucky company culture bullshit. I don't know how I could move my body out of the bed if I were to work in environments like that. A "company retreat" is a nightmare in my eyes. Sigh. I'm a pessimistic loner of an introvert and I don't need your bullshit in my life. I wish there was some more room for us on this planet. Every corner you look it's either filled with extroverts or fake extrovert wanna-be rockstar-templayers. </vent>

azernik 10y ago

I'm co-founding with someone with that dynamic (I'm handling product, she's handling business). Luckily we know and trust each other from way back, so tension isn't a problem, but there has been an interesting power dynamic - originally I was more in the driver's seat, since all we were doing was product development and there was no real business decision to make. Now that we're actually selling to customers (and investors), she's more in control of things, ie is more qualified to decide what features are important for business reasons.

finid 10y ago
I'm in the early stages of starting a startup, but I think if people focus on whether a decision being made is good/bad for the company rather than who's in charge, things would be a lot easier.

edblarney 10y ago

Those are very good points. But I think that a 'division of responsibilities' issue may miss the most important dynamic and that is simply the 'relationship'. If you get a long well, can handle disagreement, share a similar vision for the company, and have a 'history' together (which I think builds trust) - then those 'divisions' become less obvious. Sure - maybe WRT some issues, 'one of the two' can maybe have more influence, but I think the dynamic is more personal, and blurry. I wonder if it helps to have one of the co-founders simply have more authority - i.e. a 70/30 type situation wherein if the 70% co-founder does not abuse that, and stays pretty 'equal' for most things, and then hints at 'final call' in only a very few things ... that might work because it could help the 30% co-founder to concede. It'd be interesting to see some data points on that.

mememachine 10y ago

My biggest take away is that you use "she" by default.

rorykoehler 10y ago

How is it possible to divorce strategy and direction from product/tech? This siloed approach is almost guaranteed to bring problems.A CTO is someone who is essentially a technical CEO. They need to be involved in strategy just as much as the CEO for there to be a healthy co-founder dynamic.