[adamsmith]

> I've tried to explain this to VC firms. Instead of making one $2 million investment, make five $400k investments. Would that mean sitting on too many boards? Don't sit on their boards. Would that mean too much due diligence? Do less. If you're investing at a tenth the valuation, you only have to be a tenth as sure.

This turned out to be absolutely right, and the < $2M checks have ballooned in volume. They are invested by firms that do not take board seats and do comparatively less due diligence.

That said with the benefit of more hindsight I think the reason there are no more Googles is that FAANG has a stronghold on the largest technology markets.

[benlivengood]

> That said with the benefit of more hindsight I think the reason there are no more Googles is that FAANG has a stronghold on the largest technology markets.

And the FAANGs are willing to pay very large amounts of cash to acquire promising startups.

EDIT: Also why is it that the majority of the large tech companies are essentially marketplaces? Is there really that much friction in traditional markets (physical goods, movies, books, ads[!], apps) to fund their spectacular rise?

[PakG1]

I would say yes. Ronald Coase did some excellent work to explain how transaction costs get in the way of optimizing economic activity among and within firms. Traditional marketplaces would clearly have higher transaction costs than these newer marketplaces. Search costs are much higher, coordination costs are much higher, and data analysis is much harder. Tech has a huge advantage in making all three of these things much lower and thereby making transaction costs much lower. Heck, just by not accepting cash, I bet they make transaction costs significantly lower than the traditional marketplaces. Heck, just by not having to operate point-of-sale card swiping devices, I bet they significantly lower the transaction costs too.

He won a Nobel Prize if that matters. I think his two seminal works, “The Nature of the Firm” and “The Problem of Social Cost” do really well to explain why tech is winning so much compared to traditional marketplaces. Especially “The Problem of Social Cost”, which explains how externalization of transaction costs creates advantageous and unfair effects.

[brownbat]

I like the Coasean analysis. Now someone needs to build the Uber for unifying affected populations to demand rents for suffered negative externalities.

[pbronez]

Yes.

Although that’s basically just a competent, empowered, uncorrupted government. Still working on that one...

[ClumsyPilot]

I would be happy with just competent. Would make a nice change

[adventured]

> Also why is it that the majority of the large tech companies are essentially marketplaces?

They're not.

Non marketplaces: Apple, Microsoft, Huawei, SAP, TSMC, ARM, Intel, Samsung, SK, Cisco, Nvidia, AMD, IBM, Texas Instruments, Salesforce, Oracle, ASML, NXP, Netflix, Zoom, Dell, VMWare, Adobe, Applied Materials, Broadcom, Micron, Qualcomm, Intuit, Foxconn, Snapchat, Ant Financial, PayPal, Square, Stripe, Fiserv, Ericsson, Tesla, SpaceX, Activision, Electronic Arts, Nintendo, Sony, Twitter, Lenovo, HP, HP Enterprise, Lam Research, ZTE, Xiaomi, Western Digital, Seagate, Workday, ServiceNow, Veeva, Analog Devices, Canon, Spotify, Datadog, Cloudflare, Akamai, CrowdStrike, Elastic, DocuSign, Okta, Palo Alto Networks, Atlassian, Twilio, Fortinet, MongoDB, plus a hundred other companies.

Mixed marketplace conglomerates: Google, Facebook, Tencent, Baidu, ByteDance.

Marketplaces: Amazon, Alibaba, eBay, Etsy, Shopify, Craigslist, MercadoLibre, Uber, Lyft, Didi, Pinterest, Booking, Upwork, Fiverr, Redfin, Zillow, Expedia, Trip.com, Yelp, TripAdvisor, Groupon, and a few other small players.

It's not close at all, marketplaces are the modest minority of large tech companies, both in terms of market value and particularly when it comes to sales (giants like Sony, Dell and HP have relatively smaller market values and massive sales and large numbers of employees).

[wolco2]

Many of those are hardware/network companies. They don't necessarily fall into the type of startups being funded now or going forward. Companies like Cisco are important but they represent the past.

Apple, Nintendo is a mix.

Paypal, stripe, square are products for marketplaces.

Netflix is a pass to a movie marketplace. Same for spotify. Groupon is a marketplace.

[TheMblabla]

How is netflix a movie marketplace?

[TeMPOraL]

Because in a tech gold rush, nobody wants to dig out gold, or even sell shovels - everyone wants to get rich by renting shop space to shovel sellers.

[m-i-l]

> everyone wants to get rich by renting shop space to shovel sellers.

That would entail having to own and maintain physical assets (assuming you mean physical shops rather than virtual shops). Isn't the preferred model now to be a "technology platform" that connects people with available shovels to people in need of shovels?

[TeMPOraL]

I meant renting virtual shops - i.e. being the platform. This was a riff off a popular saying that during the gold rush, it's not the people who went looking for gold that got rich, but those selling shovels and pickaxes.

[laurent92]

Isn’t it Paul Graham who also wrote a blogpost saying “How to make a trillion dollars? Build a platform.” ?

[lotsofpulp]

>Also why is it that the majority of the large tech companies are essentially marketplaces?

They may earn money from a marketplace, but the reason is not because they have a marketplace. They earn money because one they did or do something that others can't or were too late to, such as creating hardware and software people want, creating and operating online services people want (Maps, Cloud Services), infrastructure for retail, or a network where everyone is and can provide some level of proof for identity or credibility.

[PaulHoule]

I think supermarkets are well run because they are highly competitive and you go every week so if you see it as a prisoner's dilemma where they "fulfill the brand promise" or "don't fulfill the brand promise" they get punished then they don't fulfill it so they have a reason to fulfill it.

Thus Amazon is going to have a really tough time in that sector, but other parts of the retail sector tend to have you come into the door infrequently enough or are in uncompetitive markets so that they don't get punished for failing.

It's too common for people to blame the victim here, or say that people are asking too much but that's just plain wrong.

Some examples:

* That time I went to K-Mart to buy three numerals for my mailbox but they didn't have the "2" in stock.

* That time my wife bought a bra at Target that had a nice print but had a completely nonsensical design that wouldn't be comfortable on anyone

* Buying clothes at Target can be really hit or miss. I bought a shirt there once that was great but so often I find things that don't really fit, are poorly constructed, I can't stand the style.

* Same with shoes at WalMart. At best you might go there and find a $20 shoe that is similar to an $100 shoe from Brooks, not quite as good, but certainly not five times worse. Sometimes they only have crap

* Inevitably in the early spring I still need cold weather equipment such as electric space heaters (sometimes desperately, as in to keep animals alive) but Home Depot, Agway and all of those other stores quit stocking it.

* Small storeowners decide they won't get any customers in their last hour so they go home early; so they've decided that the customers that do come will go to a big box store next time

* Years ago Staples used to stock only junk brands for cordless phones and similar things (e.g. vtech); today Amazon pushes no-name products from Chinese suppliers which are occasionally excellent but often junk

A common factor in those failures is that the stores don't get feedback. K-Mart never knew they lost the sale of a "4" and "7" because they didn't get "2". The small storekeeper doesn't realize that "nobody will come in" is a self-fulfilling prophecy, etc.

People also discount the idea that "the customer is always right" to normalize bad behaviors such as not stocking space heaters when you need one right now for your chicken house -- somehow it is my fault for needing heat at the end of the cold season instead of the beginning.

[lotsofpulp]

>People also discount the idea that "the customer is always right" to normalize bad behaviors such as not stocking space heaters when you need one right now for your chicken house -- somehow it is my fault for needing heat at the end of the cold season instead of the beginning.

This doesn't make any sense to me. Stores have limited space for inventory, and they have to make decisions about what the best use of that space is. It's very well possible it the probability of earning a profit on a heater at the end of a cold season is too small to merit stocking it. The store has no obligation to provide you with what you need at exactly the time you need it, so the concept of "fault" doesn't make sense to me in this scenario.

"Customer is always right" is a nonsense saying in my opinion. I always have informed my staff to show the customer the door if they're unnecessarily wasting their time.

[ffrr212]

> And the FAANGs are willing to pay very large amounts of cash to acquire promising startups.

But I think FAANGs also drain the talent pool for startups by paying very high salaries (& partially equity) to employees such that startups have a hard time to competing on this level. At least it feels it's getting worse than, say, few years ago.

[vlovich123]

Those high salaries though arguably just allow people to build nest eggs to go the startup route whenever they want.

The real challenge is that, rationally, startups an extremely large gamble on a lottery where your expected value is < 0. Most startups fail or take too long to get acquired and, unless you’re a founder/invested, you’re not seeing the really big amounts that make it really worth it (and sometimes even then). Fundamentally the challenge is that as a regular employee (even early) your stake is too small. I think startups can attract talent more easily by taking the salary disparity and matching it with stock (ie if you’re making 300k and the startup is paying 150k, for the first few years the startup should treat you like you’re investing at least 100k into the company because that is what you’re doing with your time).

As an anecdote, I have a friend who was working at a startup that got acquired by Google recently (he’s a Staff SWE so he’s making really good salary). He noted if he had just stayed at Google instead of working at a few startups over the years he would have been at least break-even or better based on salary + google RSU performance over that time. My personal experience with startups was a bit different. While the acquisition I went through was larger for me at the time, the biggest impact was that it pushed me into a far higher salary band. Startups and smaller companies can be a better way to “level up” some times than the traditional corporate ladder climbing.

[srtjstjsj]

By the time they have the nest egg, it's too late to launch a startup without hiring people who FAANG are competing for.

[dasil003]

Not following, why would they not need FAANG-qualified engineers now but they would later?

[vlovich123]

It’s not unreasonable by the time you’re 30 or 40 that you have sufficient funds that the team or project is more important to you than a salary for a few years, especially since it’ll always be there later anyway.

[pbronez]

My theory is that the big firms can do this because those relatively fixed lob or costs over huge user bases. Smaller companies just don’t have the leverage.

I also think this applies to internal applications at huge enterprises. The number of users for any given internal application is small, so you can’t spread the development costs out very far. You have to connect the initiative to something that impacts the whole firm, the whole customer base to really invest in it. Even then you’re dealing with a user base that is way smaller than the big tech firms enjoy, so you still get outspent... unless you have dramatically better margins.

[eismcc]

They also have a stronghold on a lot of the talent that would make the next big thing.

[mtgx]

As in they buy them out. There should be a more laws preventing this from happening for large corporations.

We now see that Google Search is turning more and more to AI. Perhaps if Google didn't buy DeepMind, but some other smaller competitor did, they could've turned that into an advantage against Google.

Either way, it was a loss for the market for DeepMind to be bought by Google instead of forcing Google to come-up with its internal competitor.

Companies with billions in profit should be "incentivized" to use that money to create their own competitors against threats - not buy them all out (often just to kill them). I think we can all argue that large companies buying out small competitors is a net negative for the economy at large.

[BlueTemplar]

No, companies that have billions in profit have grown too big : they can just use regulatory capture to ward off real threats.

The only thing that can be done is to prevent them from getting that big (economic incentives ? limited company life ?) or to shut them down completely (which needs enough popular support for a politician to do this kind of a risky move).

[NineStarPoint]

Regulatory capture only works if we vote in people who let them do it. It’s not a much different issue to enforcement of rules that stop companies from getting too big or shut them down. The same thing that does or doesn’t let regulatory capture happen also stands in the way of any other government attempt to control the companies.

[BlueTemplar]

I've noticed that regulatory capture by lobbies generally happens as a long-term process. Public opinion quickly gets tired/bored of a specific issue, while lobbyists never stop pushing.

On the other hand, a politician that makes one of his/her main goals to shut down (or block if not in US) the GAFAMs (or any other giant company), might be actually able to make a real and long-lasting change.

(Or the other way, to change a few of the most important laws that describe how ALL companies can function – also could be a main policy point.)

[ineedasername]

Yes, thought I think the smaller $ amounts have also been helped along by reduced startup costs. Heck, a startup can literally get $100,000 of AWS credit for free. In manufacturing terms that's like getting the factory for free, you just have to configure the assembly line to build your specific product. So $400,000 will go a lot further these days than it would have 20 years ago.

[jbay808]

20 years ago, you'd be competing with much smaller salaries from the big tech companies, which would make recruiting much easier. If you meet some smart people in college, you could more easily convince each other to try building something together, especially if they're averse to corporate 9-5 culture. Whereas now it's pretty likely that they'll have their eyes on FAANG jobs.

[ineedasername]

Good point, that's true. The lack of FAANG dominance at the time meant that labor cost would have been cheaper even before inflation. On the other hand, founders & early employees of startups often accept little or very reduced salaries on the prospect of large gains on exit, especially after the bubble burst & Y2K went away, leaving excess talent unused.

It would be interesting to analyze the difference between the lowered infrastructure costs against increased salary costs to see what the net $ effect was in terms of funding a comparable startup. If you were willing to ignore the increased cost of scaling if/when you hit the hockey stick growth and simply opt for a few co-located dedicated servers running a LAMP stack, maybe infrastructure costs would still have been low enough that the difference would be negligible. Although you might still have needed more dedicated expertise in dev ops to manage it, increasing salary... I'm not sure. I was in college 20 years ago and not very caught up in following startup culture until a few years later.

[baxtr]

This seems to be in stark contrast to some of Peter Thiel’s theses. Would love to know if someone can clarify that.

He argues since returns are distributed by a power law, more due diligence and fewer invests make more sense for VCs.

[CuriouslyC]

The truth is that regardless of how well put together a firm is, break-out success is a crap shoot. A VC strategy that maximizes potential revenue (by heavy due diligence) is almost always going to end up with lower actual revenue than a VC strategy that attempts to maximize actual revenue, by having a larger, diversified basket. If the terms of the initial funding give the VC first crack/options for later rounds, all the better.

[chrisco255]

Well, yes, but do they have that stronghold on the future of technology markets? I'm less convinced of that.

[adamsmith]

Super interesting question. I don't know.

I suspect they will dominate all the largest technology markets for the long run, unless the government intervenes. I might be wrong.

They may not dominate the future of "technology markets", though, if the sum of the long tail dominates. Right now FAANG is 42% of the NASDAQ Composite 100.

[andrewjl]

I don't think they do, at least in the long term. In the short and medium term its arguable either way. We're still at the very beginning of the era of truly distributed computing and what it tends to do when it spreads is replace centralized entities with protocols.

If this sounds like a very played out line of thinking, that's because we're now in the trough of disappointment of the hype cycle. You can't expect any fundamental new tech to live up to its hype only a few decades after invention. Also not all the pieces needed have been invented yet.

[praptak]

They filled the feasible niches for "internet-based monopoly". The next Google will have to find another way to get big.

[lixtra]

I remember a time when altavista was a the only serious search engine around. I recently switched from google to ddg on mobile because I got sick of clicking away the privacy waiver in private mode.

If ddg would deliver better search results then google search would be in serious trouble (although it may take twice as long as yahoo to die). Of course it’s unlikely because if the talent google attracts, but that was true for IBM in the 80s as well.

[ricksunny]

Which firms filled the $400-$500k gap Paul Graham outlined the niche for? Not asking for a friend.

[ekanes]

I would expect it's the category known as "micro-VC".

[VikingCoder]

Gee this sounds like the argument against managed funds, in favor of index funds.