Hacker News new | comments | ask | show | jobs | submit login
American tech giants are making life tough for startups (economist.com)
455 points by johnny313 8 months ago | hide | past | web | favorite | 161 comments

This article makes the case for government intervention [0]. The central thesis is that, had Microsoft not been sued for antitrust in the 90s and wary of yet another lawsuit, they would have killed Google in its infancy. Apparently anonymous sources in both companies confirmed this would have been a distinct possibility.

[0] https://www.nytimes.com/2018/02/20/magazine/the-case-against...

Hypotheticals don't work in law. You don't punish someone who hasn't commited a crime.thats common law. Now if it were Europe style civil law, it would be different. Also if Google had died in its infancy, there's nothing that guarantees that another search engine would not not prop up. All great ideas come to fruition when the time is right. 'Search' was ready in 1999. Similarly many falsely believe that the internet wouldn't exist without DARPA. Note that packet switching etc had already found private profitability and without govt regulation would have expanded to become the foundation of the internet

Isn't hypotheticals the entire basis of civil asset forfeiture? Civil forfeiture as practiced in the United States is based entirely on the supposition that all property is subject to seizure without any burden of proof on law enforcement and that the owner has to prove they bought the property with legal means. Hypothetically everything someone owns was bought illicitly. Any law enforcement agency can just claim it, though it's in practice they like to take cash because there's a low chance of a paper trail and dispute resolution in favor of owners with that. Property is always assumed guilty in 42 states and federal court. https://www.splcenter.org/20180116/civil-asset-forfeiture-fo...

Civil law is one thing, and has btw nothing to do with crime and punishment - that would be criminal law.

Protecting a market from becoming a monopoly is another thing entirely; punishment isn't a category here but the overwhelming interest to enable everybody to participate in a market within reason. Shielding individuals from the law of the strong (as in the jungle) and the natural tendency for monopolies is the entire point for a society to begin with.

There's two things called civil law:

1. The legal system [1], based on Roman law, that is common in former French colonies across the world (e.g. Quebec, Louisiana, France).

2. In common law legal systems, the branch of the law that deal with non-criminal matters [2].

[1]: https://en.wikipedia.org/wiki/Civil_law_(legal_system) [2]: https://en.wikipedia.org/wiki/Civil_law_(common_law)

IANAL, but this definition leaves a bit to be desired. I thought the term "civil law" is almost universally synonymous with the law other than the criminal, constitutional/non-law-of-nations law even in non Common Law countries. Specifically, it is used for law with roots in the Code civil (as opposed to Code pénal etc.) in those parts of Europe having Napoleonic law traditions, with connotations of Enlightenment and positive right theories.

> Now if it were Europe style civil law, it would be different.

We don't punish anyone for crimes that have not yet been committed either :-).

Microsoft weren't punished because they were going to be killed Google. They were punished for anti-competitive practices (specifically, for abusing their monopoly position), and they had to stop some of them. It's very likely that they would have killed Google in its infancy if they could have continued to implement these practices.

The NYC article isn't really about a hypothetical. It's about foundem.com's battle with Google over maybe being lowered in Google's rankings because they are a rival search engine.

Whether Google should be punished for that, assuming it's true, or left to rank things as they like is more the debate.

Perhaps they should be free to do as they choose, but make what they are doing very explicit. In the early days Google heavily promoted the idea that their rankings were independent and free of commercial biases, and we now implicitly assume that's true. If that's not the case, then perhaps we enforce explicitness about that - somehow.

I'm not arguing for government intervention, but it wasn't hypothetical.

Microsoft explicitly detected DR-DOS and pretended that the DOS crashed instead of starting Windows, which worked just fine on DR-DOS before then. This was ruled an illegal practice even though Microsoft had been far smaller at the time.

>Hypotheticals don't work in law. You don't punish someone who hasn't committed a crime.thats common law.

As much as I don't agree with it in certain cases hypothetical very much apply. All sort so things that have a small but nonzero chance of a bad outcome are violations of civil and or criminal law (e.g a speeding ticket).

Minor nitpick. Garden-variety speeding and most other traffic offenses are "strict liability" infractions that aren't crimes in most states. The differences include the severity of the offense and the absence of the mental state (mens rea) element for strict liability offenses vs. crimes.

>Minor nitpick. Garden-variety speeding and most other traffic offenses are "strict liability" infractions that aren't crimes in most states.

There's usually a speed threshold (15 or 20 over in most states) that make it an automatic misdemeanor. No mental state required.

I would call 20+ over garden variety speeding. There's several interstate and state highways in the Boston area where 20+ over is normal traffic speed. The traffic flows at 70-75 because that's the speed that everyone deems safe and reasonable and the highway is posted at 45-55.

Scott Galloway explained this point perfectly in his talk at Ignition 2017: https://www.youtube.com/watch?v=1ebKI4x_k8A. It's time to break up the big tech companies to allow competition (and capitalism) to flourish.

> Big tech firms have been known to intimidate startups into agreeing to a sale, saying that they will launch a competing service and put the startup out of business unless they agree to a deal...

> Big tech firms are able to shell out huge sums to keep top performers and even average employees in their fold and make it uneconomical for their workers to consider joining startups.

Reminds me so much of the show Silicon Valley. A great satire!

I’m not always convinced Google, Amazon, etc can start a product and compete in a new market and have it work 100% of the time. These companies have plenty of their own flops. Many things come to mind: google plus, amazons smart phones, MSs acquisition Skype...

Certainly they have scale advantages, but there’s also diseconomies of scale that can work against them.

They don’t have to be successful when the mere threat of their entry into a market can be enough to kill off (or suppress) potential competition.

I think the issue here is not the advantages or disadvantages of their scale.

Rather the chilling effect - the firms themselves (may be) slow to perceive the next iteration of consumer needs.

Startups on the other hand figure them out - and build a product.

The big fish/spiders, with their larger webs, sense the new upstarts and THEN build those features - and then go back to whatever it was that they were doing. The cycle then repeats.

As can be readily observed with the MS v Netscape battle over web tech (while called a browser war, my impression is that the browser was a proxy for getting a foothold in the server market).

Never mind that MS had pulled something similar vs Novell some years prior, by bundling a client with Windows and then marketing their Exchange server as a drop in replacement for Novell's Netware. Just set up a NT box in the server room, configure the PCs, and phase out Netware.

>>and have it work 100% of the time. These companies have plenty of their own flops. Many things come to mind: google plus, amazons smart phones, MSs acquisition Skype...

when it's your startup it's 100%. But they don't need to score perfectly always. They can raise your costs by 100 fold if they want, and if they fail, so what, they try another thing. Time and time again and still have tens of billions in bank. The startup has 4-5 months of salaries, if lucky.

The world doesn't need to be convinced, only the startup needs to be convinced of the possibility. It is intimidation not technical vetting of a prototype.

A lot of their power could come simply from raising risk. Cut the chances of success in half, and that nice investment opportunity no longer a good bet.

I almost quipped that vacation rental marketplaces may be the exception but then I remembered Expedia acquired a bunch of them, including a former employer so there went that.

Yeah I often joke I want to decline at least $6 Billion. (Following Facebook declining Yahoo’s $1.5B buyout offer and later Groupon declining $4B. Yeah our company’s that old.)

Anyway I met the lawyer of the Free Software Foundation, Eben Moglen, with one question: how can I prevent the tech giants from cloning my open source software, while the indie shops can? He said AGPL. That’s the poison pill you need for those big guys.

“You just watch, we are gonna copy your open source software, roll it out it to our 1 billion users and release all the source code to everyone so they can build on top of it.. oh wait.”

Yes! Do that! It will help everyone

You see... an economic system of collaboration instead of competition eliminates a lot of these issues. The best architecture and code rises to the top instead of being killed off, and the profit motive isn't what drives progress.

Wouldn't GPL v3 be as effective as AGPL in this context?

GPL v3 isn't a poison pill for *aaS where the software is not distributed but runs in a data center - unlike AGPL.

Don’t those points sound great for startups and workers?

1) We will give you a large sum of money for your startup. 2) we will pay our workers so much they won’t want to do a startup.

The sale in (1) is often on very unfavourable terms, because the valuation of a startup depends on whether the quasi-monopolist is trying to kill it or not. How much would your startup be worth if it was banned from the App Store, for example?

2) is fine .. unless there are secret no-poach agreements in place.

They do. The unstated statement by the GP is that what's good for an individual is not always good for society at large. Competition is good for society; monopoly is usually only good for the monopoly and those few who can threaten the monopoly but are happy to roll over for money.

I have a hard time figuring out what to be angry about. One day we're upset at the evil capitalists for paying workers too little. Now I guess we're upset at the evil capitalists for paying people too much.

What about "evil capitalists" using their advantage unfairly to stifle & crush competition, keeping their profits high, while keeping salaries artificially deflated for the majority of workers.

But ok, let's just create this fantastical narrative that "paying workers too little" and "paying people too much" cannot possibly apply at the same time to different segments of the population.

Hell, let's even use the loaded phrase "evil capitalists", a great term if what we're looking for is to deftly avoid any real criticism of capitalism.

What's silly is running around like Goldilocks complaining that this porridge is too hot or that porridge is too cold. Well, it's not your porridge. You're not making it, nor are you eating it. So what business is it of yours?

Yet us tech workers are some of the highest paid people in the world, and technological progress is advancing at a rate never before seen in human history.

I guess the current system isn't so bad.

> Yet us tech workers are some of the highest paid people in the world

...and inequality is skyrocketing.

> technological progress is advancing at a rate never before seen in human history.

Compared to the growth between 1900 and 1950, we should have teleportation now.

I'm not making a joke: life-changing progress keeps slowing down. E.g. refrigerators, washing machines, cars, airplanes, x-rays VS internet and smartphones.

> I guess the current system isn't so bad.

Compared to what? That's a false dichotomy fallacy.

> technological progress is advancing at a rate never before seen in human history.

I think over the last 20-30 years, the useful and helpful-to-society technological progress as a fraction of the technological industry is fairly low. Most of Silicon valley is figuring out how to extract as much ad value from consumers as possible; except for the single percentage renewable energy fraction, most of the advancements in the energy industry is a net negative for humans; medicine is stagnant because of the increase in complexity and economics; and the list goes on...

This is nowhere near the fastest technology has progressed in history.

Yes, you need an all-out mobilization of WWII and the beginning of the Cold War to move even faster. Not in the consumer segment, though.

It's almost like there are different sets of people out there, who have different experiences with things.

While I agree with more efforts to thwart tech industry anti-competitive behavior, and I hardly think the prominent tech giants are innocent in this, I also think it’s disingenuous to act like modern start-ups are stereotypically “the small business owner” or that their own behavior isn’t at least partly to blame.

My interactions with start-ups have rarely been pleasant. As an employee, I was underpaid (and wildly so on a risk-adjusted, expected value basis when comparing the lowered salary with what could be earned elsewhere), overworked, and felt the management was not informed either on the business side or the tech side, but instead came off as people with wealthy parents who viewed starting a company as the modern day status equivalent to what socialite children used to think about non-profit work or humanities graduate school.

If we’re being honest, there’s a lot of dysfunction with the way start-ups are operated, and plenty of misaligned incentives that lead to running a business more to drive hype than to actually find a product-market fit or empower employees to be productive.

Before I would start thinking, “won’t somebody think of the start-ups!” — I’d expect them to be a lot better behaved, a lot more concerned with ergonomics & healthy work places, more sincerely interested in productivity, and more eager to pay competitively. If they’re doing all this right, then I might have more sympathy when a bad actor displaces then for real anti-competitive reasons.

Google: we're so righteous we won't take Pentagon contacts.

Also Google: let's make the click through EULA as uninteligible as possible when we record our user's locations.

> we're so righteous we won't take Pentagon contacts.

I think you mean "won't renew". They already took them.


It's kind of strange how the query 'penatagon google' returns pages of uniform results, only articles containing variations of 'Google won't renew Pentagon contract'

Yea, joke is on us, because they'll still develop the thing they told the pentagon they would develop, all while appearing to be 'good guys' to people (which is apparently quite a few people) who don't understand how contracts work.

New ReCAPTCHA: select the correct person for a drone strike to proceed.

I don't think it's original to him, but Joey Hess (joeyh on Hacker News) posted an image of that concept:


I like to pretend that google is asking users to answer captcha questions to help waymo cars drive. "Which images contain a stop sign"...

I also thought they were training image recognition for self-driving cars. I also thought it was an ingenious way to solve two problems at once.

Very probably they actually do.

.. but in real time.

What stops Google from forming a shell corp to handle government contracts? They can profit off the shell corp while allowing them unfettered, royalty free access to their internal technology. They can staff the shell corp with smart people who aren't anti-government, and even continue to employ them at Google so no one knows what's going on.

I don't see why working on your country's DoD contract is for "bad guys".

Remember the guys building ARPANet for, well, the military's Advanced Research Projects Agency.

Those existing contracts are about see quite a few "scope changes"

I doubt that, and you have no way to prove it.

Whether google does it or not, China will.

The tech giants have learned to think on longer time horizons than capital markets have. Which means that big companies and VC's (their vendors) ultimately end up dominating most startups, which in turn means that the original core essence of the company that made it successful in the first place gets weeded out for short term gain. The companies that can resist these powerful dynamics are very rare. They tend to be tiny, or extremely successful (eg. Airbnb, Uber, ...). It's the companies in the middle that have to worry the most.

Good point. We may be in this position because private investors on average are naive buzzword chasing sycophants while Google etc have scale, leverage and knowhow to direct and control the market over the long term.

Another reminder, alongside the apparent Microsoft GitHub acquision, that regulators need to step up and prevent large tech firms from acting anti competitively.

From another Economist article ("How to Tame the Tech Titans", https://www.economist.com/leaders/2018/01/18/how-to-tame-the...):

> Two broad changes of thinking would go a long way towards sensibly taming the titans. The first is to make better use of existing competition law. Trustbusters should scrutinise mergers to gauge whether a deal is likely to neutralise a potential long-term threat, even if the target is small at the time. Such scrutiny might have prevented Facebook’s acquisition of Instagram and Google’s of Waze, which makes navigation software. To ensure that the platforms do not favour their own products, oversight groups could be set up to deliberate on complaints from rivals—a bit like the independent “technical committee” created by the antitrust case against Microsoft in 2001. Immunity to content liability must go, too.

> Second, trustbusters need to think afresh about how tech markets work. A central insight, one increasingly discussed among economists and regulators, is that personal data are the currency in which customers actually buy services. Through that prism, the tech titans receive valuable information—on their users’ behaviour, friends and purchasing habits—in return for their products. Just as America drew up sophisticated rules about intellectual property in the 19th century, so it needs a new set of laws to govern the ownership and exchange of data, with the aim of giving solid rights to individuals.

The problem is that regulators rarely have the farsightedness to see which acquisitions are indeed anti-competitive. And I don't think it's always obvious, especially on the bleeding edge of tech. The farsighted ones are probably part of the acquiring company (or at least acting as consultants to the hiring company).

Definitely. Instead of a regulatory body proactively thinking about practices as anti-competitive, there appears to be a very reactionary approach of punishing large incumbents that pursue m&a out of necessity or approaching anti-competitiveness as partisan politics.

Yeah, I think regulators realise we have a monopoly brewing once they see people buying Google branded milk from Googlemarts, having driven there in Googlemobiles, using traffic directions from... Google.

That's not a monopoly - a monopoly would be if for at minimum one of those things, you could only find a Google-controlled version.

I think it is monopolistic. If one company has so much influence in your life that they can direct you to their own products, so that you habitually start relying on them for almost anything, then they have the ability to develop a wide range of monopolies by influencing your choices. E.g. Google has a monopoly as a trusted source of all information - what's even stranger, is that they are seen as a non-interfering third party by most users - because of this, many other monopolies can emerge, because they have the ability to control "truth" and "popular opinion".

By that logic, Walmart would already be a monopoly in the perception of quite a few people in America, which is clearly not the case to much of the rest of the country.

That's not true, but it's close; at least according to US law. Microsoft was deemed a monopoly in 2000 (which is actually legal) and abused it's monopolistic powers (which is not legal) while other OS's existed in the market: Apple, BeOS, Novell, Linux, OS/2, etc.


As you know, Microsoft had more than 90% of the OS market worldwide. While other OSes did exist, they didn’t provide meaningful competition. The person to which you replied meant that Google Milk is only a problem if Google abuses its search monopoly to push the market away from other milk vendors, not that Google isn’t a monopoly as long as another search engine exists. Their main point was that it isn’t abusing a monopoly to launch a broad array of products if each of those products compete in a fair market.

For the legal notion of monopoly, you don't need to be the literal only one.

Definitely the political angle in most of these regulatory institutions is frustrating. And money in politics as an extension.

It’s further complicated by the fact that a lot of this stuff functions more like a public utility than a line of business. It may well be that nationalizing the firm makes more sense than cracking it up.

> Such scrutiny might have prevented Facebook’s acquisition of Instagram and Google’s of Waze,

Then again one of the main reasons people start companies like Instagram and Waze is so they can be bought by these giants.

I think people forget how inconsequential Instagram seemed at the time Facebook bought it. It was popular and growing fast, but the notion that this 12-person startup was worth a billion dollars was laughable. Read contemporaneous accounts and many reporters were incredulous. It was prima facie evidence of a bubble.

I don't know if a regulator would have had the foresight to see Instagram, which was still as much known for its filters as its social network as anything but a compliment? Look at how Instagram was seen as one of many similar products at the time:



An early rival, Hipstamatic, was named the Apple App of the Year in 2010: https://techcrunch.com/2010/12/09/apple-top-apps-2010/

Insta has taken on a life of its own over the last 2-3 years, but if they had chose to go their own way, whose to say Facebook wouldn't have clobbered them, in much the way they nuked Foursquare.

I'm not opposed to regulation, per se, but we should also be careful not to indulge in revisionist history.

This is totally NOT correct. When Facebook bought Instagram, Justin Bieber was already posting pictures there. I was already posting pictures there. My friends were already. It's not users, it's growth rate and engagement. And they were through the roofs. Also, thinking that Andreessen, Peter Thiel and Mark Zuckerberg just go and throw away $1B for the sake of it... it's nuts. Also, 30M MAU mostly in EU and US it's an extremely high number. For comparison, it has grown x20 in the last 5 years since the mentioned number has been reached. Approx. it kept its growth rate relative to scale, give or take a multiplier - depending on how you do the math.

On another plan (not really related to the argument- but more as to what really made IG stand out), when almost identical services as you mentioned are similarly growing, the only difference is the endorsement behind closed doors. Instagram was endorsed by Adam D'angelo, Jack dorsey and Sequoia w $50M in its war chest

Yeah this is definitely correct. I was on board believing FB would become bigger and bigger. But the Instagram acquisition seemed crazy. How many users did Instagram have when it was acquired? 30 million? How could anyone think Instagram would get to where it did (Instagram is supposed to do $6-7B in revenue this year).

Also, FB itself wasn’t what it is now at the time either.

The big platform companies see engagement and growth metrics for all startups, not just the numbers they are pitched. The public didn't expect Instagram to be a success - Facebook had more information.

This information asymmetry is a big issue. It means big platforms can detect competitors early and know when to offer generous acquisition deals to neutralise them. The people working in these competitors don't have this comparative basis to tell how well they're doing.

As a small player, it's very hard to see how successful your competitors are. For example, Google and Apple only provide vague statistics about number of downloads in their appstores, rounded very loosely. But they internally can track detailed metrics. Facebook and Twitter are used as distribution mechanisms, for login, and as sources of contact info. They can glean insights from that.

Facebook has even better insights through the Onavo VPN, which as it tunnels traffic from other apps, can even show which features are doing well https://techcrunch.com/2018/02/12/facebook-starts-pushing-it...

At the time it was bought, Instagram was a major thing for myself and my friend circles. It had been for a little while.

I'd love if such companies would say this up front. "This startup aims for an exit." "This is a product demo for facilitating acquihires." It would make it easier to avoid those.

I would argue that's also a problem. Startups who's eventual exit plan is acquisition are more willing to burn VC money giving stuff away for free, rather than coming up with a plan for profitability. If more companies had the idea that they had to survive more and more on their own revenues, I think we'd see stronger companies.

Seeing that Github is reportedly unprofitable, if they aren't acquired by a profitable company, what's the alternative?

What do you think Microsoft would do to harm Github? I doubt it would be as bad as what became of SourceForge.

"Seeing that Github is reportedly unprofitable, if they aren't acquired by a profitable company, what's the alternative?"

Lay off employees until they're profitable.

Tech companies tend to invest ahead of growth, because tech markets tend to be winner-take-all and if the market grows when you don't, somebody else will take the market. They're effectively using their employees as insurance to make sure nobody else can catch them in the market.

As a result, they often have orders of magnitude more employees than is necessary to keep the lights on, the product operational, and the money flowing in (at least in the short term). Google, for example, really just needs datacenter ops technicians and SREs to keep the product up, which is maybe 2% of their current headcount. If they laid everyone else off, they'd eventually become vulnerable to a competitor or lawsuit or public opinion or unhappy/indifferent customers. But until that happened, they could maintain quite hefty profits.

I doubt thier investors were looking for a no growth botique business that never went public and never got acquired. The investors were looking for an exit strategy.

Well, for an investor, there's another option: put more money in. They only have to think about laying people off if there's no profitability and no further investment and the acquisition falls through.

If I had a stake in Github, I would be delighted to sell it to Microsoft.

I don't have a stake in Github. I am a user with 2 distinct roles. In one role, I work for a company that pays for Github's services. Github has done a very good job hiring people who have maintained the service as it scaled to very high levels. I am concerned that Microsoft will be unable or even unwilling to keep those teams intact. They may wish to restructure the organisation. They may wish to change middle management in an attempt to make it profitable, etc, etc. Github has had a good track record with us. Organisationally we have no experience with Microsoft, so it's a risk.

Personally, I spent a chunk of my career working in the MS world (working with gold/platinum level partners). This is ancient history by now (10 years ago), but my experience with MS support was abysmal. Essentially "support" amounted to having a telephone number where I could report issues. No follow up allowed. No notification if the issue was ever resolved. Upgrade to the next version when it comes out and check for yourself to see if it has been fixed. Hey, that's why we pay for those MSDN subscriptions! Though, if you have about 1000 seats in your license, you can get real support (if you are lucky). Maybe that's all changed and they are fantastic, but once bitten twice shy as they say.

In my other role as a consumer with Github, I have my personal repositories there. I still vividly remember MS's push to create a division between "hobbyists" and "professional developers". The idea was to provide different levels of access based on the idea that "professional developers" work for large companies that will pay for high licensing fees and extensive support contracts (which only really seem to count if you have at least 1000 seats... see above). "Hobbyists" are intentionally crippled by giving them just enough access to seem like they can do something, but denying them crucial things.

With the release of VSCode, etc, I can see that the company has changed, but old ideas die hard. 10 or 15 years is not enough time for me to feel like I can trust the company. Will they slowly screw me over by using their dominant position to "boil the frog" -- slowly crippling their "hobbyists" without ever doing anything overt enough to cause them to jump out of the pot? Only time will tell, but yes I'm quite concerned. To be fair, MS is not the only potential purchaser who would concern me. At least it isn't Oracle!

> What do you think Microsoft would do to harm Github?

GVFS. I believe Github will only deploy to Microsoft's GVFS sometime after acquisition. This is one of the more obvious plans, I thought shrug

Well they could change the terms of service and essentially drive certain types of projects away. They could limit access to older builds and versions to non-paying customers. They could limit access to verified/signed builds. They could reserve certain rights to your software such as they did with npmjs.com. They could run ads, offer IT staff skill matching and promotions, FizzBuzz-like services, or other LinkedIn integrations. They could come up with clever schemes for offering commercial licensing for open source. They could go after the enterprise package mirrors and policy checkers market Artifactory et al are serving. Not saying they'll be doing that (MS isn't stupid), but given MS is selling mainly to enterprises, there are many creative ways they could make money of it.

Seeing they don't do any of that today with VSTS there is no basis to believe they would do that for Github.

What do you mean deploy to "GVFS"? VSTS -Microsoft's existing host solution supports the standard git web hooks today to support integration with third party CI/CD tools and supports Azure and AWS deployments.

These would be horrible. You make it harder to sell your startup with random oversight of small acquisitions, you’re going to scare away small and medium sized investors.

If it is true that GitHub run on long term loss, then itself is important anti-competitive problem.

Cashectomies. Applied to wind fall profits, savings, estates, endowments.

Keep the money moving.

How is it anticompetitive if Microsoft buys GitHub?

Another example is Partner programs. We're relatively new entrant in the social media scheduling industry and are unable to compete with Buffer and Hootsuite because they have API access to post to Instagram via the Facebook Partner program.

In order to even apply to be a Partner we have to have a "backer" within Facebook who will vouch for us.

Weve been trying to find such a person since February and we're still not any closer to being able to apply.

It's a hugely anti-competitive approach and makes it impossible for non Silicon Valley startups to get a foothold in the market.

As a hardware engineer who admittedly has a bit of contempt for software types, I can't help but think that maybe software startups just aren't novel enough these days? As noted in the article, software has dramatically lowered the barriers to starting a company, so isn't it inevitable that major companies would exploit the same low barriers (powered by their ability to gather intelligence on the next big thing via their platforms)?

It's not discussed in this particular article, but economies of scale certainly favor incumbents in hardware as well. If you need confirmation, go ask Pebble guys how they fared after Apple Watch was launched.

I am not sure you can use one example of a hardware company failing as proof of anything. Just as you couldn't use the success of a software or hardware company proving the opposite, that there isn't a problem.

There are enough articles out there from hardware kickstarters that failed post-funding for me to believe that manufacturing new products on-price and on-time is insanely difficult. I'm sure Apple can design and manufacture 50+ iPhone prototypes for any yearly launch. But startups really only get one chance to get it right before their funding runs dry.

You don't even have to look at startups. Look at all the teething issues that Tesla has had ramping up production, versus an established car company. Look at production issues for things from established companies like the Nintendo Switch. Even the iPhone X had its production issues.

Certain segments of software startups aren't novel enough these days. It's pretty hard to start a new consumer web company or mobile app and make any money off it.

Software has the dubiously nice property of its foundations being invalidated every 5 years or so, though. You can still make a lot of money in crypto or AI, for example. Ethereum is full of smart contracts that are doing the same things microcomputers did 40 years ago, millions of times slower, that nevertheless have raised hundreds of millions of dollars for it because they redefined money in a way that somehow convinced people to pay for it. And now all of the nice properties of software like "can process more than 15 transactions/sec" have to be rediscovered, which provides plenty of fodder for entrepreneurs.

Hardware doesn't usually invalidate the laws of physics on a rolling 5-year basis, by contrast. That gives firms an ability to build up large stores of domain knowledge which can't be easily replicated by startups, so it's correspondingly harder to get a new hardware startup going.

Yes, also tall trees in the forest block the sunshine to smaller trees, at times. However, software and hardware need each other and although one may say that hardware is where the true innovation lives, our ability to reason and make practical use of <abstract> data is limited by our understanding of computing science.

why contempt?

I assume it is because he thinks engineering involves making physical objects, rather than abstract ones.

I am a software guy, and I have heard other hardware guys make similar statements as well.

An example is a physical car engine (hardware), vs a video game rendering "engine" (software)

Which some what ignores the fact that a 2 stroke engine has three moving parts in its simplest form.

and mech engineering doesn't have its clients demand 5 months before a new engine delivery you now make it a rotary Wankel engine as that's the new hotness :-)

Ex BTEC A1 (Thermo Fluids ) Eng Tech BTW ;-)

Is BTEC suddenly a notable qualification? And isn't A1 the lowest level (up to 5?). Last thing I heard was that Pearsons BTEC system was a gateway to University in the UK; akin to GCSE or A-Level depending on the level taken. And A1 would be year 10 (14-15 years old) ~ GCSE level grades.. no?

If this is no longer the case my BTEC's are going to get a prominent position on my CV.

A1 was the specialised version of a mech ENG BTECH aimed at the pump, turbine, hydrodynamic and aerospace industries it did cover all 5 levels up to the equivalent of a HND aka first two years of Uni.

All but one of us worked at the bleeding edge in research for the scientific civil service or in related civilian organisations based at Cranfield.

I did do a spec pitch a few years ago to a small engineering company RBR racing - I recon I could saved them a F1 win using the basic skills I learnt at my first job

Sorry, I'm looking everywhere and I can't find any evidence to support your claim.

BTEC A1 is introductory level from all sources I can find.



As I explained A1 is the Version of the course that went from level 2 - 5

This is also responsible for geographic inequality throughout the country.

The large tech companies are based out of Silicon Valley and Seattle. The majority of their employees and economic activity will be in those areas.

That has gradually changed a lot in the last ten years.

A decent house is $150,000 to $200,000 where I'm at. That's 2,000-2,500 sq feet, 0.75 to 2 acres of land. The murder rate here is almost zero, violent crime is very low, and I have access to high speed Internet and normal shopping / consumer goods. I can day trip to multiple major metros. Public housing here is actually nice, safe, clean. The homeless rate is nearly zero. And I spend zero time in traffic.

Put your $50,000 down on the home. Your mortgage payment is now $725 to $750 or so.

Make $6k, $8k, $10k, whatever per month doing Go, Python, JavaScript, etc. remote contracting. Or just take a decent remote job.

Your take home is ~$54k on $85k income contracting, before any deductions. Your bills are $2,500 per month. You can easily pay your house off in seven or eight years.

And that's not a crazy scenario or a high-end outcome in terms of contract work. That's doable for anyone in the top 50% skill wise.

You can pull off that scenario all over the US.

Or make $150,000 in Seattle, and that same house costs $1.2 to $1.5 million. Somehow figure out how to come up with your $325,000 down payment. Now your mortgage is $5,000 per month. Your property tax is over $10,000. Your take home pay is $9k to $9,500 per month. Your house, for just the mortgage and taxes, is going to cost you around ~66% of your take home pay, versus ~16% in my scenario. That house is going to cost $2 million total - minimum - over 30 years of mortgage payments, plus another $300,000+ in tax payments.

I feel pretty bad for engineers making ~$150,000 or less in Seattle or San Francisco. Telecommuting will set you free.

Engineers in sf make a lot more than 150k in their late 20s/early 30s even outside big 5

And housing is even more expensive in San Francisco than in Seattle, general cost of living is a bit higher, and CA has income taxes, adjust the figures accordingly.

The $1.3m house in Seattle will cost you $2m in SF.

Now you need $500,000 for a down payment. Your mortgage + taxes will cost you $110,000 per year, out of your $158,000 take home on a $250,000 income. You'll burn $3.8 million over 30 years on the equivalent of a typical middle-class house, before accounting for any repairs/improvements or tax increases.

You have $4,000 per month left to cover basics like: car, utilities, homeowners ins, food, home repairs, savings, life.

That's all assuming a $250,000 salary, which few people make in SF. The average is a lot closer to $200,000 total compensation, the median is even lower.

$250,000 salary is a helluva line to have to perpetually stay above for 30 years just to barely afford a modest middle-class house, while saving almost nothing besides that.

That's not how the math works at all. 25-30 yr olds making 250k don't live in 2m homes. They usually share an apartment in SF and pay ~$2000-2500 on rent. Then when it is time to get married or move in with a gf, you get a starter home for 500-600K, putting 10% down. Now in a couple of years, your starter home has gained $200k in value. Sell that and use the proceeds for a larger single family home around 1-1.5mn. Meanwhile, since your partner is in the Bay Area, your household income is between 400-500K. Now you are 35 and you get a couple of promotions and are either a senior Engg manager or an architect/tech lead at L6 or L7. You then start pulling 200K just in RSUs. Meanwhile your house also appreciates in value. At this point, said engineer has 300-400k in liquid savings and another 300-400k in home equity. There is a reason why people move to Seattle and SF as the math largely works out

That path is as rarified among software engineers at bigco as software engineering at bigco is among software engineering at large as that is to the general population.

No, they move there because they hope that math works out that way, and they think they will be the one guy out of 10? 20? 100? where it actually does work out.

I don’t understand what is the unlikely thing to happen in the scenario described. If you’re talking about striking millions in startup exit/IPO money, I def agree with you it’s unlikely.

Plenty of engineers hit their cap at L5 or thereabouts (I'm sure if you look at the distribution of levels at your local bigco you'll see there's a sharp decline in numbers at some point), and not everyone is going to be part of a DINK engineering power couple.

But at this rate in 30 years that modest middle class house will be worth 2.6 billion dollars!

(kidding. I think).

As is apparent here, most (software) engineers can only understand things as linear models.

That is after you include stock.

Not a bad deal today, but I'm still questioning if 1 trillion dollar companies exist or are overvalued.

That is a fair point. Liquid salary doesn't usually go much beyond 200k; all of the additional comp is in RSU's. If there is a huge downturn, a lot of engineers will lose a great chunk of their assets.

My current "quarter life crisis" is around this number, it appears to getting from $150-200k in liquid salary is MUCH more difficult than $100-150. So if liquid income is your "north star metric" job wise, what do you do?

RSUs in a public company are liquid, they can just be volatile.

Correct. And a lot of folks deal with this volatility by immediately liquidating a certain fraction of their stocks and investing in e.g. Mutual Funds and the like.

Doctor's and lawyer's have been noticing this sooner and taking advantage of this too.

It's called geographic arbitrage or geoarbitrage, and factor's in more than just salary. You can actually make less in many places and live better.

The problem is, telecommuting is still not common. Hell, the running joke is that Slack, a company who's entire purpose is improving communication among teams to enable more remote work, does not allow remote work themselves.

Which city/state are you in?

I'm in the midwest. I've had zero luck with remote work. Zero. The rest of what you said is spot-on.

Not just startups. Indie/one-person shops too. Everybody’s getting drowned out.

Euh, you mean Amazon

I don't know, maybe I'm not enough for the startup but I don't feel very intimidated. If I've ceded control of my board, surely they can sell the company. If I make a product that Microsoft or Google or Apple can copy in the next 6 months. then surely that is my fault.

Added: I will say that I am not comfortable with some of the state powers that allow other businesses to mangle the legal system around you to sabotage you, but the solution to that is less state control (thus less reason to lobby), not more.

Software folks are quick to write off companies which are not technically difficult to build (eg all of social media), but I think Slack is a very interesting case to think about.

The software is simple, the idea has existed forever, there were competitors in the space. But the thing that really drove the space forward was Slack raising hundreds of millions and marketing the shit out of their software. Only after that have the big guys gone "oh, shit, we have to get in there".

I'm not really impressed by Slack as an engineering/innovation company, but it provides a very interesting perspective on business.

I'll admit I'm sitting on the fence, and perhaps ignorant about anti-competitive behaviour in tech. These companies are winning over marketplaces by providing the highest quality, and lowest price services to consumers? Hypothetically, where they slip up in providing the best services, startups could thrive? Or are there elements I'm neglecting as to how large company domination negatively impacts innovation?

In-app payment policies are a major restriction. Apple and Google have unlimited access in offering digital goods and payment methods. Third-party apps are limited in scope and taxed 30%.

If two songs are sold on an iPhone for $1 each, one by Apple and another by [insert music sales app], the small app makes 0.70, while Apple makes 1.30. Lets say each has to pay 0.50 to the label for the sale. The small app profits 0.20 while Apple profits 0.80. That's a 4x advantage for essentially the same purchase.

Although it doesn't relate to the acquisitions, I'd definitely agree, that's anti-competitive behaviour. Was not aware of that policy.

A different look may be like this: I made an app and I gladly give away 30% (it can be less, btw) for not having to care about setting up and supporting payment, distribution and updates infrastructure, dealing with PCI and chargebacks, etc.

You didn't get to make that choices. It was forced on you.

Nobody forced me to program for iOS.

The fact that iOS and Android are the only smartphone OSes did force you to do that.

Unless you mean that you could have made a competitor, but that's inane -- that's a completely different market, with huge barrier to entry.

Apple is using its huge market position in one area (smartphone OSes) to give itself a huge advantage in another (music apps). That's the sort of thing companies have got huge fines for in the past, or were broken in parts for.

You can have alternative app store on Android there's no technical limitations, right ?

And at some point you could jailbreak iPhone and iPod Touch and crack apps but no real alternative market rose up

Aside from setting up payment, all of those issue are still present with Apple / Google payments. The difference is 3% (MC, Stripe, Visa) vs 30% (Apple, Google). That 27% can easily cover those issues and far, far more.

In fact, it's worse. Payment & fraud disputes are also bound to the marketplace. What are the chances that app devs will end up being the ones eating big losses just to continue to exist?

They buy out companies long before most of them can become a threat. Google used to buy a new company every week a few years ago.

Also, when a smaller company does somehow manage to get known in the industry, it's usually under pressure to raise money or be acquired in order to compete with some other huge competitor.

I think this is mostly a regulatory fail in letting some companies grow so large and acquire smaller players. I think past a certain point, they shouldn't be allowed to do it. Finding out what that point is might be a little difficult, but probably not impossible.

I understand their strategy of neutralizing threats through acquisition. However, for this strategy to be effective, the companies have to consistently turn a profit on their acquisitions. Although it seems absurd for Facebook to spend 19 billion to acquire a 14-person company (Instagram), failing to keep Instagram popular amongst consumers will turn it into a 19 billion dollar writeoff as some Instragram clone captures consumer attention. IBM has a long record of acquiring competitive or smaller companies, but they're doing relatively poorly in the markets as of late. So while it seems clear that these companies might make life a nightmare for a lot of silicon valley entrepreneurs, they're still creating the best net value for society, no?

> Facebook to spend 19 billion to acquire a 14-person company (Instagram)

Do you mean WhatsApp? I think they "only" paid $1 billion for Instagram.

Woops. Yep. It's hard to keep track of all these billion dollar Facebook acquisitions.

It’s about control, not size. IBM controlled powerful segments of the market, and would defend or compete by buying technology.

This scales down to small things like parking lots. Real estate guys will buy and operate little surface lots at a loss for years to deny competitors the ability to develop properties that compete with them.

>IBM controlled powerful segments of the market, and would defend or compete by buying technology.

Right, but IBM used to comprise almost the entire hardware and software market. Over time, they failed to consistently deliver the best experiences to consumers and clients, so they definitely declined from their glory days. I understand your comparison about real estate developers, but I don't think it applies here. IBM can buy competitors, but if they can't make sure their acquisitions remain the best in the market, then the acquisition loses profitability, or even becomes a writeoff.

As an anecdote, I used to work on one of their analytics products, which they acquired in one of their biggest acquisitions ever. Nevertheless, we were competing for clients with smaller, leaner, and sometimes better teams in the same field. We had no inherent advantage in development over our competitors either. Sure, IBM could always buy them, but then they'd have another x billion dollar investment to turn around, and still no way to stop new competitors from taking the clients, other than creating the best product - which would be net beneficial for society anyway. In many cases though, IBM's attempts to exert control by acquiring other companies backfired because they failed to provide client value, and the company has had to write off a lot of losses by trying.

Big companies do have access to a kind of data that is not quite public, and gives them an edge Im not sure is fair. For example, a Reddit's founder says he picks which company to invest based on reddit activity by companies in their site.

Im not comfortable with the edge places like facebook, google and reddit have over other companies, knowing what they know about with loosely named propietary data.

Im not too concerned with them buying companies: microsoft used to do that and didnt stop any of these.

I once got a legal threat from facebook when I was setting up an app because I registered a domain to serve my app assets with a name like myappfacebookorigin.com and it said that the word facebook is famous and I need to cease and desist. I wish they had paid me back for the domain or forced the registrars to warn me I would get legal threats for using their trademarked name.

It’s not obvious that using a trademark you down own might result in the trademark owner asking you not to?

This is not anti competitive behaviour. In fact, this IS competition. Competition means all sorts of actions that lead to me having more customers than you. The only problem is when the entrenched business use govt/lobbying/abusive patent law to get this done. Buying out competition and trying to enter a low barrier to entry business is the very definition of competition. Suck it up

Agreed, I'm having a hard time seeing how this is anti-competitive. It sounds like they're just saying that tech giants are competing very successfully by virtue of being really big and having lots of money. Paying people really well, buying out startups with good tech, competing with startups, etc.; none of that sounds illegal, or even unethical.

There's also an argument to be made that big tech giants have some benefits for startup ecosystems. The option of being bought out, for example, acts as a safety net, so that even if your company doesn't have explosive growth that lets it go public, well at least you'll get something out of it. Similarly, having tech giants nearby means that even if your startup just implodes, you can probably still get a decent job in the area (going from big company to startup to back again is really common in the bay area, for example).

Disclaimer: I work at Google, my opinion is my own, etc.

This. If there has to be a regulation, I'd require companies to give more freedom to users, e.g. to choose default search engine, install alternative app stores, allow access to public data without account (e.g. LinkedIn) etc.

The article is an accurate assessment of the landscape today, but is it going to be true next year or five years from now? You can never be too sure!

Using a bit of imagination to think of risks these giants face :

* Antitrust - could happen when it's politically advantageous to publicly bash tech giants

* Stock market - a recession could reduce tech giants’ leverage by a lot

* China - competition from Chinese companies or the Communist party imposes an embargo

* Technology - another AI winter or maybe some other company gets the edge in quantum computing for instance

* Demand - their products today are better versions of what they were ten years ago, not completely new things. What if users turn away from these tech companies as they become the ‘old way of doing things’?

* Startups - Google grew on Microsoft's watch, Amazon on Walmart's, Facebook on Google's, Apple on Nokia's, Snap on Facebook's. What's next?

* Crypto - ethereum, bitcoin, etc. could usher in a whole new internet experience where giants on the 'old' internet find it hard to gain a foothold

While working on a startup today might feel like picking pennies in front of a bulldozer, tech giants face a number of risks of their own. Startups would better focus on internal excellence than fret about tech giant monopoly.

Please, for the sake of all that is holy, let's not let Coinbase get sold to Facebook.

Reading this article makes me think Microsoft acted to acquire GitHub before the other company-hoarding giants could acquire them.

> Big tech firms are able to shell out huge sums to keep top performers and even average employees in their fold and make it uneconomical for their workers to consider joining startups.

Perhaps startups should look at changing that. Offering more actual money, better options, and stop with the fuckery that makes most people believe that their options would be totally worthless even when the company is acquired/goes public.

Not the tech giants, but GDPR too.

No. The GDPR just says that you've got to care about user data. If your startup can only survive by selling out user privacy, then perhaps it doesn't deserve to exist.


How is this news? Yesterday’s immigrants make life tough for future immigrants. Likewise, yeaterday’s startups make it hard for new startups. It’s a circle of life. Evolution at play. Capitalism at play.

Yesterday's young (and today's old) make it hard for future generations. Are you defending that?

That view is pure social darwinism:


> It’s a circle of life. Evolution at play. Capitalism at play.

I'm not sure what that implies. We could say war and the strong robbing the weak aren't news, but we've developed societies and rule of law to limit those catastrophes and promote justice and opportunity for everyone. As a result, people have time, freedom, resources, and security to acquire skills and develop software, among other things, rather than having to fight for survival.

Applications are open for YC Summer 2019

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