Software Is the New Oil(avc.com) |
Software Is the New Oil(avc.com) |
"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."
Regardless, it's a good analogy, but be careful how far you take it. There's a _very_ important difference between the energy and other major industries. It's a difference that the software industry shares, but in the opposite extreme: The amount of capital needed to do business.
The energy industry is notoriously capital intensive. For my employer, a project has to be over 250 million to be considered a significant project. We have a _lot_ of those, including multiple projects in the 10-50 billion range. Sure, I work for a very large company, but even a "mom and pop" oil company (there are a lot, actually) needs tens to hundreds of millions to get started.
By contrast, the software industry prides itself on being able to get an initial product out the door with essentially no other investment than time.
At any rate, it's a fine analogy, but keep in mind that the oil industry is _very_ focused on building _big_ infrastructure. In that sense, oil companies have more in common with independent nations than with other companies. It's something to think about, at any rate.
We've already been through one. We may be on the high side of another right now. I think the mechanisms are sufficiently different from the oil world that comparisons aren't too helpful, though; the presence of some form of negative feedback in the system is often sufficient to produce oscillations. The negative feedbacks are quite different in character even if the results are the same at a sufficiently high level of abstraction.
(I'm not sure that we're in a "bubble" in the 1999 sense that people worry about, but still, if the economy ever becomes healthy again and interest rates go up, software will cease to be the only investment people can make that has any chance of paying of, which will mean that even without a "collapse" the investor money flows may dry up relative to today. However, I think the industry has a greater focus on real profits today, which will buffer the industry substantially vs. 1999. Yeah, we talk about the hyper-growth eyeball-selling companies, but we talk about they partially because they are the exception; there's a lot more "real value" companies out there right now.)
I'm not sure that we're in a "bubble" in the 1999 sense
that people worry about,
The access to cheap money has brought in a lot of players into the market who otherwise would have been sidelined. I think it's not only evident in the startup/unicorn sector but quite so in the building/construction sector as well. I see all around me skyscrapers & apartment complexes being built in an unprecedented rate and people buying half a million dollar+ houses where quarter of a million was too high (as you can see I'm not in Cal).
This is no joke. I grew up around big Oil coming from Houston. There was a joke that Chevron effectively had the second largest combat navy in the world because of how many resources they devoted to platform and tanker defense.
Tools and automation get better while all computer resources become less. oil operating costs likely dwarf a comprable SASS conpany in operational expenses.
edit
to clarify, opex in software becomes more expensive in aggregate. i can't think of a good business that has dropping opex as it scales except maybe Enron or a chainmail referral company. Energy companies do have high barriers to entry with a much lower opex after infrastructure has been developed, but software isn't neccesarily the opposite. If we compare a built out production application, outside of beta, with a decent userbase and its core purpose/features implemented, with a nuclear plant or oil refinery, the costs to rub the app is almost certainly lower than the energy comps and could possibly have opex held fairly linear relative to growth.
Oil production and infrastructure maintenance aren't cheap. It's not like we just turn on the taps after the wells are drilled. However, labor is cheap, even in the software industry.
To me, it's the top talent of "software PROGRAMMERS" that's the "oil".
Let's pretend that Google Inc opensourced their entire software stack. Now, anyone can just spend money on hardware and datacenters and "replicate" what Google does in a certain sense. But did you really duplicate their abilities? Would intelligent and visionary investors fund such copycat endeavors?
I say no because smart people would realize you didn't replicate Google Inc's "hiring pipeline" of the best minds from Stanford/MIT/etc. Yes, we may have gotten a snapshot of Google's source code but we didn't duplicate their ability to attract desirable workers who can build <<the next future thing that's NOT in that source code dump>>.
Even Bill Gates had noticed this point: Google's ability to poach top talent from Microsoft was better than Microsoft's ability to attract Google defectors.
Same analysis can be done for Amazon inc. A person could take all their source code for ecommerce and warehouse logistics and I'm not confident he could outcompete Jeff Bezos. First, you must prove that you're hiring a better pool of candidates than Amazon.
Lastly, the "software" advantage leaks outside of the organization because others copy it (open source) or ex-employees with knowledge leave and reimplement it (legally) at their next gig. On the other hand, it's not as simple to make top compsci graduates switch their career aspirations from Google/Facebook/Apple in SV to SmallPotatoesInc in Alabama.
Hopefully we'll be able to use digital oil to fuel more than just the great engine of personalized advertising, which seems to be where most of the revenue is coming from right now.
The other is entirely dependent on skilled labour, and once deployed continues to deliver value.
Software is much more like real estate: the first entrants erect a patent barrier around the best bits and use it to charge rent to everyone that clusters around the core. It's subject to gearing-driven bubbles.
I don't want to be dismissive, but the software industry might very well go through and M&A and commodification phase once it matures. It looks like oil now because those companies are big monopolies, will it last? I don't know, I wouldn't put all my assets in software, that's a bad idea in general anyway,
All of those involve very large software elements, and benefit from software innovation.
It's quite plausible that the existing winners are at the top of their ramp, so buying in today at high p/e might not be a good bet. But all the unicorn investing is betting on new software companies coming along and reaching ascendency.
It's like starting with observation that no business nowadays can run without email, and therefore building an email client and a server for business is the way to make it big in the software industry.
Energy, land, machinery, labor, transport - a few of the things you need to extract and then process oil. The same goes for creating and making a microchip useful. The first oil tower at Titusville didn't use oil as its energy source.
Is oil a layer two product, and software a layer three product? Perhaps, however oil is definitely not primal, it just seems like it is. The vast effort and industry required to make oil useful, along with a century of science, is mostly hidden away from view.
I missed this... any good articles about it?
http://www.computerweekly.com/news/4500256048/Amazon-turns-s...
Why did Amazon launch a semi-high priced phone (rather than stay consistent to the traditional low price strategy)? Bezos was hoping it would spit off a few billion a year in profit, and they'd finally have a big cash fountain to swim in like his competitors in tech.
In four to six years AWS will be a $150-$200 billion company in terms of valuation. It'll be capable of generating $4 or $5 billion in operating profit in that time frame. It'll likely pass Facebook in terms of sales within 36 months, and Facebook is worth $292 billion.
Amazon strategy is to become a leader in many fields, they got lucky and envisionned the potential of cloud early, for commerce of goods it is a thousand year old industry with big competition from all other the world, their goal is to cut costs, not to care about short term returns and once other businness collapse due to prices cuts that they cant folllow, then amazon will be the only leader in commerce and will rule the price of pretty much everything that is traded in the world.
Also, there are other wildly profitable companies in other sectors that have been around for decades.
Came here hoping to find a "Facebook could..." line when talking about revenue. Was not disappointed. :)
By comparison SAP is Europe's largest tech company and does $21 billion in sales. Facebook will surpass them within eight quarters (at a mere ~14 years old as a company).
Ask Western oil workers in Libya or Nigeria or Venezuela or Mexico (depending on the decade) how accurate they feel that sentence is. Oil concessions have been regularly expropriated.
you can't copy oil, you can't open source oil, software isn't limited
Can be found online at: https://hbr.org/2003/05/it-doesnt-matter
I think the difference is that the state of the art shifts so quickly, long term survival for platforms is questionable. There's always a new player.
I'll sell you some shares in my software company at the same multiple. We've even got better margins. It'll be great. I promise.
Specifically what's so amazing, is that AWS is set to produce perhaps $2.25 to $2.5 billion in operating profit in the next four quarters. With the growth curve it's on, it'll very rapidly become one of the most valuable companies in the world, buried within Amazon. They'll hit $20 billion in sales in just three years, assuming slower growth.
By comparison, AWS is already generating 2.5 times the operating profit of VMWare. It has greater sales than either Salesforce or Netflix, and is generating a drastically greater operating profit than either of those two.
Of all the suit babble I ever heard, this must be the most incomprehensible. Juggling stock options apparently wreaks havoc on the speech center of your brain.
Explain to me again how this is an analog for software?
And that execution comes from the talent powering the company, at least in the right positions.
WalMart is significantly bigger than Amazon. WM - 485.65B v Amazon @ 88.99B. Call it 6x.
I shop at WalMart, and I will quite frequently take their "in store pickup" option. I also have amazon Prime and the goods I pick for those are different, it seems.
So does that make all the rest of us plain programmers the grease?
This analogy is getting strange, but I like it.
Maybe ordinary software programmers are like crude oil because there is finite number of them, but still abundant. And fairly fungible.
Talented ones are like highly enriched uranium in that they're much more rare than crude oil by volume.
But they're still fungible because they can jump ship and decide to burn themselves at a different company.
decide to burn themselves [out] at a different company
Truth in humour, dude.
I think by this point we're accustomed to dealing with large software players who have scaled their businesses to the point where labor cost is indeed marginal to overall profit picture.
Back in the days attempts to build a small-scale software company (local Web design firm) or even a medium-scale software business (seller of compilers, industry-specific tooling, shareware) basically broke down because of unfavorable economics.
So we're dealing with a bit of survivor bias here - smallish and medium software firms for whom labor was a major cost center are either out of business, commoditized or both. In energy business, while the large companies still enjoy large budgets and economies of scale, it's still quite possible to be a small to medium size player and enjoy a steady cashflow.
You can't just go down and pump another few thousand barrels of raw software into your Strategic Software Reserve.
All the amazing software Google develop is to support their ad business, more or less.
I'm talking strictly in the 'online retail' space, which I believe was pretty clear.
Just like pollution from the fossil-fuel industry, you can pick high-intensity problems that are restricted to the areas nearest the source, or low-intensity problems spread over the whole country/world. For example, the housing pressure in SF/SV is very high intensity but affects only a small part of the world; non-STEM high school and university departments losing prestige and/or funding is less intense in any one place but more widely distributed.