Hacker News new | past | comments | ask | show | jobs | submit login
Marc Andreessen answers questions from Stripe Atlas founders (stripe.com)
145 points by darwhy on July 13, 2017 | hide | past | favorite | 84 comments



"And so if a founder can’t navigate a network into a VC firm, it is unlikely that founder has the skills to navigate the other networks required to succeed in building a company"

So, as someone who's working on a startup at seed stage, outside silicon valley (Socal), what Marc is saying does not make any sense; networking requires time, and time is very valuable especially at seed stage; I would rather work on developing the product and refine it for market fit, rather trying to spend precious time on meeting with people to expand my network.


I've heard many VCs say this. It's one of the few examples of Marc demonstrating complacence, wrongheadedness and herdlike thinking in this interview.

But it does point to an underlying truth: the VC market has not found a good filter for startup quality beyond human curation. Irony, anyone? They haven't managed to gather the data and apply the algorithms yet, so they let themselves rely on a primarily inbound model that depends on the VCs building a reputation and network, promoting themselves in the media, and then waiting for founders to come to them through the filter of the Silicon Valley milieu.

While there ought to be a better way, there are also obstacles to collecting perfect data in a fast-changing private market. Ideas gestate quietly, startups are founded in stealth, most of the data is self-reported, and therefore spotty and unreliable.

That said, I do think there's a lot of potential to gather data and analyze it to make predictions about startup quality. Some investors have already done that, based on their many interactions over decades. And those solutions, of course, will remain private for some time. The smartest actors in the market have little interest in improving how the rest of the market makes decisions.


Yes, I think the real reason why VCs don't like meeting people outside of their networks is that it makes them uncomfortable.

Meeting new people all the time is exhausting and destabilizing.

People like consistency and a feeling that the world is a small place full of familiar faces.

I think that the herd mentality makes sense for small VCs who don't really know what they're doing and don't have any vision but it's disappointing to see it in these big guys too. They are the trendsetters after all.

That mentality is exactly why the rich get richer and the poor get poorer.

Also it is sacrificing technology/innovation for sales/marketing/networking.


I think you misunderstand what is meant by networks.

VCs constantly meet new people but generally it's people recommended by their current network or sourced by outbound research rather than people who cold email.


Quantitative traders have trouble teasing signals out of data. And there is so much data in the public markets: stock prices, volume, interest (and the rest of Bloomberg data), Twitter sentiment, satellite data. The private markets have so much less data, so the best indicators are who the founding team is, how big the market it, and other obvious stuff like that. There are some quantitative VCs in Menlo, SF, and NY, but I wouldn't hold my breath.

VC is probably speculation, but let's wait for VCs to show consistent back to back returns on funds and we'll find out. For hedge funds, it hasn't panned out so well - not sure it'll be different for Anyone Else.


Actually, it's worked out quite well for some hedge funds. Renaissance Technologies is a quant-driven HF, and it's one of the largest and most successful in the world.

The only quantitative VCs worth anything will be the ones who have been operation for many years and who apply the algorithms to the data they have collected during that time, on all the startups they interacted with and invested in. They hold enormous private data stores about those private companies.

I think the real problem is that the best and most famous VCs, like Marc, don't need the algorithms to get access to many of the best deals. They're already overwhelmed with companies good enough to merit a meeting. And maybe they feel as though their funds are doing good enough. But for funds willing to leverage their private data, there's an edge to be had.


Renaissance is always cited in these conversations, but as you noted, isn't representative of the thousands of quant shops out there. That, and I hear their quant edge is disappearing with competition and they are relying more on HFT over the last decade.

Quant VCs just started appearing in the last few years, and it's not clear there is an edge there. As I said, my opinion is it's not likely.


Unlike the stock market, experienced VCs should be applying quantitative techniques to private data they've collected. That's one way to keep their edge. You don't seem to appreciate the distinction. Of course most public-market quants will see their edge fade if their competition has access to the same data. That doesn't refute the initial success of those techniques.


Both public and private have access to prop data. Public has more data, since the companies they're covering are bigger and so produce more data.


Also "For us, at the seed stage, 90%+ of the decision is based on the pedigree and track record of the core team."

I have that network and pedigree, and I still think this (common) approach is disgusting.

Not because it reduces diversity (it does), but because it needlessly excludes those who could transform our biggest consumer and business populations.

If you went to a non-pedigreed school (say Howard, Morehouse, etc), if your first job & location wasn't in the valley or NYC (perhaps Dallas, Schenectady, anywhere in the rust belt), if you're in a group which makes networking for the sake of networking with tech bros particularly uncomfortable (women), you won't have access to that network. You may, however, have unique access to America's biggest industries and consumer segments.

It's a terrible waste. And all because VCs can't be bothered to open their email.

As a total aside, I'm rather surprised that nobody in the valley has yet thought to create an "anti-valley" fund. Intentionally seek out founders in underserved and isolated groups, and give them all an equal shot. I _don't_ mean a diversity fund; I'd be going after companies like Pigeonly, Darkvision, Veerum, etc.


I imagine the problem with the anti-valley fund is similar to the prisoners dillema- it takes more than one firm to see a startup through all of its funding, and if no one else bites you'll be left with an underfunded startup. It's probably why the pedigree rule is still in place- it's not rational in a vacuum, but it's a known bias of other firms, and so it makes sense for you to act in that bias as well.


Pedigree/track Record doesn't have to mean tech startups, deep experience in the industry vertical also counts.

If you look at say startups serving the oil/gas industry they're mostly founded by founders with industry backgrounds from outside of tech hubs.

If you're in that kind of scenario I'd recommend finding people who cross both the industry in question & the tech industry (i.e. angels with a background in the industry, other tech founders in your industry, etc.)

Although if you're based outside the major tech hubs it's probably easier to focus on local investors for early rounds.


I used to work in a startup where the all the CEO wanted to do was "network" and networking with VCs and people in the industry usually means go out drinking. It ended up that the team and I (I was the CTO) ended up building the product that made revenue that supported the business while the CEO's networking only meant he didn't show up a lot of times. It lead to absolutely nothing mostly just a waste of time.

Focus on building the product and focus on selling and serving your customers. I have since left that company due to issues with the CEO and started my own. I am now doing 0 networking and focusing 100% on building the product and servicing my customers.

People in the startup scene here say "you never go networking how's your startup doing, what's your valuation now" to which I reply proudly "we are growing quietly, my company has real revenue not valuations."


I think most investor intros that go anywhere are made by people who built their network in a much more solid way. Every time I've been introduced to an investor it was by someone I've worked with. So it's really this superficial networking that's completely dumb.


I'm of the school of thought that 'let investors come to you'. Focus on building your product to attract customers and the customers will attract investors. If you do it like that, you will always be the one deciding how much money to take for how much of your company you will give away.


Theoretically, once you have customers you could take loans instead of giving your company away...


Typically the trade-off is how much money you want.

Once you're an established company you can probably borrow around ~20% of your revenue while an equity investor will typically be putting in 3x-5x your revenue.


You sound like my kind of guy..

Just build the product! A company does not need someone full time yakking it up.


If you don't know how to program it's pretty hard to do a job that requires programming. Likewise if your job requires networking skills and and appropriate existing network, it's pretty hard to do your job if you don't have those things.


This reminds me of a story I tell about fundraising. It goes toward this question of time.

The way I raised my first money was I invited Evan Williams (Twitter) to coffee. He said yes. I told him my concept but didn't show a demo because I didn't really have one.

He liked the idea and offered to put in some money. Magic, right? Money just grows on trees and anyone can get funded here in Silicon Valley.

But the better questions is: how did I get that coffee with Evan Williams.

To me, it started in 1996 when I decided to get a computer science degree. That took four years. Then I took a sort of crummy job but impressed someone there who helped me get a job at O'Reilly where I impressed someone who introduced me to Evan Williams as someone he might want to hire. That took five years.

Then Ev did hire me and I did good work. Then Ev went off to Twitter and I cut my teeth doing a terrible job as a founder of something I bootstrapped. That took five years.

So really, it was 15 years to develop a network and trust where I could invite someone to coffee and get funded.

Yes, some people do this faster than I did. But I often see this real resistance to the idea of taking the time to improve your qualifications as a founder. Networking is obviously one of those.


If you're filtering 2,000 qualified pitches per year down to 20-40 investments in the hope that 2-3 of those investments put you in the black - then your filtering-signal-from-noise strategy clearly isn't working. You're throwing darts at a dartboard and assuming that your dartboard is better than everyone else's just because it's got more Stanford grads on it.

But why, particularly, is it arrogant? Because the logic that startup founders need to demonstrate the ability to network with customers by first networking with VCs is horribly flawed. Perhaps that skillset of networking and selling to customers is exactly what's missing from a startup which otherwise has built a highly valuable product at low-cost - and isn't this precisely what you would expect from the stereotypical basement-dwelling übernerd, a product with high potential trapped by the poor social skills of its creator? Perhaps that ability to network is precisely how a VC can help a startup turn into a unicorn?

No wonder nobody's managed to fix the interviewing process for software engineers. If you only interview engineers who you're personally introduced to, and then filter out anybody who isn't an alpha MIT grad, I mean, does it really surprise anybody anymore that this is a failed strategy? But if VCs filter out companies in this manner then why is it any wonder that companies filter out engineers in this manner?


So basically "I always throw away half of the resumes I get at random. I don't want to hire anyone who is unlucky" except for startups.


I hate the point but for better or worse it's a standard. I view it sort of like tipping in the restaurant biz, "fuck it, let's let someone else pick up the slack".

Stepping outside of how to change that (and there are ways) if you want to be in a position to have Mark or another top VC fund you, then you can:

- apply to YC or 500. They do the legwork and look into all the opportunity that comes there way. Accelerators get you there.

- Raise an Angel round from someone "in" the circle. If you want to raise, you'll have to network and hustle. It sucks, but you can go for money AND/OR smart money. Hopefully you can get both.

- Mentors and customers. If you get a customer that is "well known" or have a fan that is a founder, it's a great way in. This community, your community or similar places can get you a mentor or advocate.

It sucks. I get where you are coming from. It's like an Ivy league or a great job. They are willing to write off people who would be top shelf because 80% will find a way to reach them. It's not even that absurd, if you're dream is to get funded by A16Z then you'll show up at the office, get into YC, find some one they funded or get so big they can't ignore you.

If you don't like it help change it.


> rather trying to spend precious time on meeting with people to expand my network.

He's not saying that. He's saying if you can't network in your domain or find people that do, it will be extremely difficult to scale a business. Or in more "hacker terms", get close to your customer and hustle.

I know more than a handful of very successful software founders who have made a ton of money scaling software businesses with dogshit technology, simply because they had a strong network in their domain. I think this one of the few instance Marc actually concedes that it's not simply about "software eating the world". Professional networks matter.


Joining an accelerator is a way to break into that network while focusing on your business. Depending on the accelerator this provides a pre-vetting too (your startup was picked as one out of thousands or more for a reason).


Yeah, I hate the insular attitude as much as anyone, but it doesn't really make sense to look at VCs as gatekeepers to the Silicon Valley world. They are mostly investing in companies that have already established some traction and gotten early stage investment.

If accelerators and angel investors all thought this way, it would be much more of a problem, but thanks in large part to YC blazing the trail, there are a lot of possible ways in for unknown founders--you just have to settle for climbing the ladder and proving yourself instead of having millions thrown at you because you're in the club.


Marc isn't saying you should focus on networking at the seed stage. You should focus on building + customers and just not try to raise your seed from a16z.

a16z isn't a seed focused fund. They prefer waiting for more proof points, and they're willing to pay a premium for that (Series B).


Outside of a major startup hub you're probably better off finding local investors or relocating.

The single biggest bottleneck at any VC firm is the partner's time. If the partner has to fly out to meet you (multiple times during the investment process + for every board meeting) it means you require significantly more time than a local investment. It's also a lot harder for a remote VC to provide support in areas like hiring (i.e they're unlikely to have many good candidates in their network who are located near you).

All of this means that the threshold for a VC to want to invest in a startup outside of their key hubs is higher than for local startups.


It's a lazy filter - and I don't mean "lazy" in the entirely negative sense.

If you get 4000 pitch decks per year and ultimately make 15-20 investments, you need to whittle that down fast. Having half that volume be "cold" and just tossing it gets you down to a more manageable number quickly. Might you miss some gems in there? Sure, but you don't need 'em all, you just need 15-20 solid ones.

Another VC might think differently. This just says don't go to this particular VC to raise your seed.


I'm not a VC but I think the quote makes sense.

Either you have a track record or a product with some customers. Cannot get funding without one of those.


I think it's the fuzzy definition of "track record" that people take issue with. This appears to mean anything from "has been a key member of a successful business in the past" (reasonable) to "went to an Ivy and has lots of friends" (kind of silly).


So... "It's not what you know, it's who you know"

That's what I wanted to get away from.


Surprise!


Not only time but network requires empathy too. And VCs/Silicon Valley has a huge problem with empathy toward some large groups, so his reasoning is wrong OR prejudice.


Listing influential books:

"The Baroque Cycle—a work of rigorously researched historical fiction with only the slightest overlay of science fiction—tells the story of the emergence of the modern world and its systems (democracy, the scientific method, financial markets, etc.) in a way that is wholly fresh. These novels make me think about what a Neal Stephenson of 2300 may write about our times and us."

That last sentence is fascinating to me, and I think telling about what helps a Marc Andreessen stand out as a person.


Reading the Baroque Cycle is probably the only thing I've done more intellectually demanding than learning to code. It's ultra long and insanely dense.


I'll never forget the years I spent reading it.


We call this "what do nerds do on nights and weekends". This is the single most reliable source of new ideas in our industry that will ultimately be adopted much more broadly.

Right now we see huge nerd night and weekend energy in areas like cryptocurrency, biohacking, quantified self, synthetic biology, virtual reality, drones, and self-driving cars (!!).

Anyone car to expand on this list? I'd include applied ML perhaps, and IoT hacks.


If you're into what nerds are doing on their weekends[0], we get together and hack/show/chat about stuff every Sunday at Noisebridge. A lot of cool stuff happening there from crypto/bots/bio/ai/vr and beyond. We also live stream the event so it's easy to keep up and participate.

[0] https://www.noisebridge.net/wiki/Hack_Days


I am not terribly impressed by Marc Andreessen since he invested and advocated for uBeam, Soylent and claimed sleep is overrated. Laws of physics apply to wireless chargers and biology applies to human bodies. This all has a whiff of anti science and that is a stench you don't want around a technology evangelist, frankly.


We were excited about Atlas and just applied two days ago and got rejected this morning. We have sales ready to get going but now have to legitimize the old-fashioned way.

With the "network" of venture capitalist firms and startup incubators able to give you an instant invite to Atlas, it felt like it's made for them (the venture firms) and not the businesses.


Hiya, I'm Patrick and I work on Atlas. I'm sorry about that -- that rejection appears to have been a mistake; I'll fix it for you (check your inbox in a few minutes, please) and look into fixing the underlying cause more broadly.


Thanks, Patrick! You've made a fan out of me.


I looked at Atlas this week, in large part because I have a healthy respect for 'patio11, but waiting to be approved when I can just file in Delaware with somebody like HBS and have it done in three days didn't really make sense to me right now. ($5K in AWS credits would be awesome--for some ventures I've done. This one, meh, whatever.)


This goes for anyone on HN -- feel free to email me directly. I'm (my HN username)@stripe.com I'm happy to expedite things if e.g. they're time sensitive for you.


Thanks Patrick; this is the first time I've heard of Atlas and it looks like a great service. Is there a floor of what you expect from applicants, i.e., is a website showing some viable product required before an applicant would be considered competitive?


It's not a competition.

Investors have some limited pool of capital or bandwidth and want to conserve it for the companies they think are most likely to be successful. We are not constrained in this fashion.

We need to know what the business intends to actually do, mostly for compliance reasons. A working website or prototype we could look at makes this easier, but we know there are a lot of ways early stage startups get to market. We can be flexible -- if you don't have a website but do have a deck, we can evaluate that.

We (and our partners) have a notion of what companies are a good fit for Atlas, and it's not quite as broad as "every business we could legally transact with." It's described in more detail in the FAQ, but broadly speaking we're looking for businesses that are intrinsically international and/or ambitious and technology-enabled.

The world needs more hair salons, too, but we'd probably not work with a hair salon at the current moment in time. We're not confident that the thing we are building is appropriate to their needs yet, and want to stay focused while making something for our core audience.


That helps clarify things for me. I did find more information in the FAQ, but I would suggest making it easier to find. Maybe a simple Is Stripe Atlas right for my business? link in the "Ready to join Stripe Atlas?" section.

Look forward to applying in the near future!


Right on--cool that you're doing that. My stuff already turned around, but I'll give it another look next time around if need be.


That's interesting; I have an Atlas invite that I'm about to use but I do not have a workable business...yet. I didn't even realized they "rejected" people, I thought it was more of a waiting list type of thing.


With all of the venture firms in the "network" I'm wondering if there is anything suspect with incorporating through them (ie. Equity percentages) but I'm probably just being paranoid. Feel free to share any detail after you use your invite.

EDIT: was being paranoid. retracted


What's your route to "legitimize the old-fashioned way". Is it a case of incorporating the C-Corp and calling up Silicon Valley Bank yourself?


Yes but I'm not knowledgeable about how to incorporate so I have a lawyer friend who's gonna help. Atlas was mainly for convenience and the free AWS credits.


So what's your business?


Well I didn't think this was a place to market myself and it's not incorporated or anything so there isn't any benefit to share yet, is there?

EDIT: shared below


I'd take the opposite approach if anyone ever asked me what my business was - every interaction is an opportunity to refine and tighten the elevator pitch.

Just do it and ask forgiveness if someone moans about being OT.


> an opportunity to refine and tighten the elevator pitch

Thats a great insight, thank you. I'll just drop the website (work in progress). Gamma Fi: https://gammafi.com


IMO the first thing that you should have on your website is an e-mail collection mechanism. I want to sign up to know when you have a product because this looks promising.


damn, I have a hard timeline to get launched in August but will try to throw something up soon because you are right.


I feel I've been MVP-ed by your website (not that there's anything wrong with this.) Every link led to a 404 but since I clicked on every link to see if that were the case, I'm guessing my site visit didn't tell you anything about what I was interested in the most. Sorry about that.


Yup. I threw that up quickly and haven't had time to finish (been finishing up the product). I put it up just for Atlas


Way to you go, OP. That's four new people who never you existed who now care about you (as ephemeral as this may be.)


Your logo is quite similar to Google's, enough that I was momentarily confused opening the website.


When do you think you'll have a working demo? Would love to show this to someone on my team.


I've took the whole month of July off to finish this up with the goal to have it finished by August 1st


Looks great! Who did your website?


thanks. I did :)


how long have you been working on it? have you shown it to any potential customers yet?


I've been working on it since December. We've been showing it some people in relevant positions and getting feedback. I literally threw up a website to sign up for Atlas so it wasn't meant to be marketable just yet


"And so if a founder can’t navigate a network into a VC firm, it is unlikely that founder has the skills to navigate the other networks required to succeed in building a company." At what point does it make more sense to use your networking skills to acquire more customers than use those same networking skills to acquire VC dollars?


A strong network is generally fairly adaptable. The work goes into building the network, not tapping it.

Once you have the network in place, you can just as easily tap it to find a VC as to find customers. (Of course, this is assuming that your customers are generally in overlapping industries/areas, which isn't always true.)


Ask question the opposite way: at what point is getting an investment going to exceed the rate at which you acquire customers and revenue compared to doing it without the investment? Your default plan should be no investment if possible.


What an elitist comment. I'm sure they'll get burned eventually with that kind of attitude.


Can someone living in the US care to expand on this?

"...which is also when many VCs get back from their August vacations (sad but true!)."

What is sad here? Being Swede,my initial reaction was "ok, late summer vacation".


The VC industry is somewhat unique in the US in that it takes a long, almost-industry-wide vacation at a time of year when other American professionals do not have a coordinated vacation. People who have to interface with the industry (like e.g. founders) are occasionally annoyed at what is, from their perspective, an unannounced total work stoppage.

It is useful background knowledge here that American professionals culturally expect less vacation total than European professionals do (the norm is ~2 weeks) and expect it to be largely non-contiguous with the exception of the end-of-the-year holiday season.


Is this because American people genuinely think that 2 weeks a year for vacation is good enough, because of a culture of ambition, or simply because of the grip of employers is too strong to take the risk of asking for longer vacations (i.e. very cheap to fire people compared to other countries)?


It's just a cultural norm. There are strong economic and political and historical reasons why the cultural norm exists, but the fish doesn't see the water. From the point of view of an American, it's just "normal".


I know in Seattle its even more culturally tolerated because we get so little sun year-round and you gotta binge in the peak summer month.


> It is useful background knowledge here that American professionals culturally expect less vacation total than European professionals do (the norm is ~2 weeks) and expect it to be largely non-contiguous with the exception of the end-of-the-year holiday season.

The exception would be big tech companies where it's not unusual to see employees at all levels take 3 to 4 weeks vacations in the summer* , especially foreigners since they want to spend time to visit their families, and depending on their country of origin, were used to 7-8 weeks.

(* not every summer)


Some are saying that the "sad but true" comment is a commentary on American/Silicon Valley culture.

I interpreted it as a comment on the fact that VCs will pass on funding if it doesn't fit into their schedule


A month long vacation is a luxury most don't have here.


"Can we help you start your company? Join Stripe Atlas"

This made me think all the interview was an ad for Stripe Atlas. Marketing ploy to lure you into the culture(we help you make the company + tell you all the inside secrets for the second step - funding).

Irony is, we have applied to Stripe Atlas for a "Delaware" company as they say, and it got rejected like in a few hours. So IDK for whom this is actually intended.


This made me think all the interview was an ad for Stripe Atlas. Marketing ploy to lure you into the culture(we help you make the company + tell you all the inside secrets for the second step - funding).

Of course it's an ad (or more specifically, "content marketing"). That's generally what company blogs are for.


I thought several of the questions asked were unique and interesting. It would have been nice to hear Marc answer them in more detail.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: