Software Is The New Oil
I was with some friends this weekend and one of them was talking about an investment committee meeting he attended and there was a discussion at that meeting about some of the threats out there in the macro investment landscape. One of them was “vanishing liquidity” and the significant change in net cash flows from the global oil sector. Oil producing regions have gone from being a massive cash generator to a relatively small one in the past few years. Now this could well be a temporary thing as the oil market adjusts to some new realities. This post is not really about oil, even though that word is in the title of this post.
As I pondered that, I thought about oil’s role as the thing that captured the economic surplus of the industrial revolution. You can’t run factories, railroads, trucks, etc without carbon-based products and in particular oil. So oil has been a cash/capital magnet for the wealth that the industrial revolution produced. Those that owned oil producing assets (or better yet, oil producing regions) sat back and collected the economic surplus of the industrial revolution and that has been a path to vast wealth and economic power.
What is that same thing in the information revolution? And where is cash piling up around the world? On tech company balance sheets, of course. Apple has $200bn of cash on its balance sheet and produced $53bn of cash in the six months ending March 2015. Microsoft has $110bn of cash on its balance sheet and produced $30bn of cash in the year ended June 2015. Google/Alphabet has $70bn of cash on its balance sheet and produced $14bn of cash in the six months ended June 2015. Facebook could have $20bn of cash in the next year and could be producing $20bn of cash a year soon. Amazon, the company that “will never make money” surprised Wall Street last week with strong profits and it seems to me that they are going to start producing cash like these other big tech companies now.
It makes sense to me that software is the oil of the information revolution. Companies that control the software infrastructure of the information revolution will sit back and collect the economic surplus of the information revolution and that will be a path to vast wealth and economic power. It has already happened but I think we are just beginning to see the operating leverage of these software based business models. The capex spending necessary to be a software infrastructure provider at scale has shielded the cash producing power of these companies (and many others) and may continue to do that for a time, but I suspect at some point the profits are going to overtake the capex at a rate that the cash will be flowing out of software companies the way that oil flows out of wells.
Full Disclosure: The Gotham Gal and I own a lot of Alphabet stock and also shares in several hundred other software based businesses. We are long software.
For the latest quarter, AWS was 8% of total revenue but 52% of Amazon’s profit. Game on!!
Q: how do you build the biggest “software infrastructure service in the world”A: hide it inside the biggest online retailer in the world until it is so big it can’t be ignored
Seems like AWS was really a by product at first
Absolutely. Just think what Amazon’s financials would look like w/out AWS. Without those margin contributions Amazon would be dead in the water. No one, including Bezos, had the foresight to know AWS would be this big.
Good analogy, but Walmart has a host of other problems, including questionable lease and inventory accounting issues and that Mexican bribery scandal. Sam wouldn’t be very happy today.
Great observation about AWS margins
That’s common though.
Yes – but still neat that they created a real business while building their infrastructure to run their core biz.
My view has been that, just for their internal needs, Amazon about had to do essentially just that, and Bezos was smart enough to see it and do it.Or, he clearly saw how much work it was to get the computing he needed and had to notice that with just some good fiber and one bit flipped, he could do the same for at least thousands of other companies, also struggling with that work.
Yes, see my reply to Fred. I read that Jeff Bezos has the idea to spin it out as a separate line of business.
Oh Amazon will also have one of the biggest hardware infrastructure services in the world too. As will FB…Both already have first-mover advantages before GOOG, MS, TWTR and others.Per my slide yesterday… PIPELINE EFFECTS.It’s fun being a dev, product person and strategist because we see and play with the tools and toys of the future before most people! LOL.
It’s even better than that. What they have done is raised an entire generation on their way of doing things and their way of naming things. AWS is greek to those that learned English but not those born in Greece.At this point many of those thus educated and raised on AWS are in the “now yous can’t leave”  phase. They wouldn’t be able to easily do things the other way even if they wanted or needed to. Another Bezos long term strategy that appears to have paid off.Amazon in Plain English:https://www.expeditedssl.co…… https://www.youtube.com/wat…
Clever AMZN. They took something dull (data warehousing) and made it a lot easier.Plus they made it modular and integrable with third-party tools (Docker, Chef, Aerospike etc.).
I am not sure I agree that they made it “easier”. Unless you consider having to learn an entire new language and process for doing things “easier”. If you already know how to do it the “old” way it is not easier, it’s actually harder. Because you have to figure out how to do things the AWS way. And once you are in that walled garden it’s not easy to get out.What they did, their “value proposition” (to use a 1990’s term) is give people an easy way to scale. If you don’t have to scale something or if you are running a “normal” site you most likely don’t need to get involved with AWS. You can just use some cookie cutter service or a VPS or a shared server or your own colo server and save money and have greater flexibility. After all hardware is cheap as is colocation. What is expensive is knowledge. Knowing how to do all of that. If you already have the knowledge (and once again don’t need the ability to scale (there are other reasons of course)) then you don’t need AWS.
Have own VPNs. Have built own dbs (hello SQL and NoSQL!). Have dived down rabbit holes of server security and front-back integrations too.AWS’s a very nice one-stop solution.
What do you think about more traditional infrastructure providers such as Rackspace and Softlayer? I have been building upon these at a very low level, avoiding dependencies with their internal plumbing when possible. How does AWS compare?
Thanks for question.Originally, I was interested in Rackspace for the polyglot persistence aspects they marketed in their roadshows. I thought, “Great! This means I can do Neo4J for data viz of clusters, say of movies that people are watching, whilst using Redis for user sessions and MongoDB for the actual product list of movies.”And indeed polyglot persistence continues to be a core part of Rackspace strategy — see ObjectRocket:* http://blog.rackspace.com/r…However, then I went through some technical bootcamps with AWS (all the way to Advanced) and discovered this… pls see image.There was a limitation with DynamoDB’s utility and what I wanted to do but this has been resolved. So AWS is pretty good.A key positive is that the documentation is thorough and if you’re lucky enough to attend their technical bootcamps, everything (how each db model fits into the whole) becomes a lot clearer.* http://aws.amazon.com/mobil…
One of my all time favorite movie scenes. My favorite lines are when he picks up the guys head and says: “look at me I am the one that did this to you remember me” and then when he skulls the guy’s head with a baseball bat and says: “you ruined my whole fucking lunch”I eat with an employee at a restaurant every day for lunch. It is my only standing meeting.
Yeah. Also the one where Sonny went after the guy who hit Connie:https://youtu.be/_OEhiTHJ0E…And a few bodyguards are able to keep control of the whole crowd.I mean there are so many of these just writing this reply I’ve thought of so many others.What’s odd is that Carlo even hit Connie given who her father was.
Or until they extract it as a service that becomes a separate business.It must have also helped that they got to test the functionality of AWS (even if it was not called that then) at Amazon scale before spinning it out.
Why would less money going to oil producers lead to vanishing liquidity for investors? Surely the money they had previously captured from consumers is now simply being redirected elsewhere – for example shiny new iphones. Either way, the net money “in the system” remains the same. Only difference is who has it.
Hasn’t vanished from the system, only shifted within it from one sector to another.
Plastic is oil.
Not strictlyAmber, tortoise shells , and animal horns are all technically plasticsref http://plastics.americanche…
Plastics are polymers but not all polymers are plastics
Fred does have a little DH in him
Oil companies made lots of money because, everyone on the planet needed oil. Now software companies are making lots of money because software powers almost everything. Distribution for software has gone from millions to billions in the last decade or so. The companies that you mentioned – AAPL – owns mobile, GOOG – owns info discovery, MSFT – owns productivity, FB – owns social. All these guys have distribution in billions of people they reach and on a daily basis. And COGS on software and distribution certainly helps margins.
Growth rates at Amazon and Microsoft for Cloud biz 70-80%
And the ISPs control the software “pipelines”
It’s a very good analog. Oil in all its various forms (software)Pipeline/delivery (networking/access)Infrastructure/Drilling Equipment (hardware)
i love it
This analogy does undermine your argument that everyone needs to know code.I can put oil in my Suburban but I don’t know much about hydrocarbon cracking.Oil light on your dash must have an info equivalent….
Happy that the analogy has not surfaced an OPEC so far.
rights of way were the central issue in rockefeller’s ability to create standard oil. same for wired (local communities) and wireless (FCC).
excellent. extending the analogy. i love it.
I found the global warming “debate” equivalent, job loss.
Which is why you see Google making so many plays for open broad band (guess they learned a thing or two from Rockafeller)…in so many ways it’s the same old battle only the names have changed (Rockafeller, Tesla, Edison, Morgan, Carnegie, etc. and their related companies — Google, Facebook, Amazon, Apple, Microsoft, etc. and their related CEOs)
This is just a really great and thoughtful analogy at many levels. To follow up, I would suggest that electricity was the bridge between oil and software (but didn’t do the balance sheet analysis for GE, etc.)
This is an excellent expression of a theme I am looking to write a book on. It will be about the 10 technological themes which will leave those businesses which cannot become sophisticated technologically in the dust. It seems more and more clear the disruption that has happened to media and retail will happen everywhere. The value of the economic surplus being collected in IT, while in the past it went to oil, is a great metaphor to illustrate this, and reminds me of the “software is eating the world” article by Marc Andreessen.Areas that I am keen to pick as core topics include the blockchain, open source, artificial intelligence, the cloud, the internet of things, cyber security, the smart grid, the connected car, genomics, and data analytics. Feel free to let me know if you want to help on this, by taking a quick interview, as I suspect the A VC community will have plenty of thoughts that I won’t have even considered.
Was wondering where you were going with that title given finite v infinite resources. Think it is spot on.
Yes and no, but for the same reason: software can be replicated at zero marginal cost.Oil is a finite resource while software is not, and economic theory would say that something with infinite supply cannot command any price at all, let alone a premium price.But unlike a barrel of WTI, all software is not identical. So owning the IP of a successful piece of software does actually allow premium rents to be extracted, precisely because of the zero replication cost.So is software the new oil? I don’t think so, I think data is.
And I would say that the new oil would be people/talent…for one thing, to actually make sense of data.
Zooming out one more level I was going to say it’s real estate for the talent to live on. Some companies like GitHub and Basecamp have short circuited this by going “remote” but most companies are not fully remote and so worker elasticity of demand for real estate near their workplaces is pretty low. Landlords literally can extract higher and higher rents from them with little consequence.
This dependence on proximity of team members at internet and technology based companies has seemed ironic to me but it is also a reminder that a business is also a form of community which takes a lot more work to create without proximity. It probably helps keep the human element of a business more alive (ha, figuratively).
Yep. I think there are a bunch of challenges to remote work and I think you’re especially right about the community aspect. That’s probably why co-working and coffee shop spaces have taken off. Not to mention the fact that the communities in which these companies launch (SF/Boston/NYC/DC) are highly desirable places to live to begin with.
I’m listening an argument in the middle of the office over a new feature with people representing these departments: U/X, Database, Datacenter, and Tech Support. It is rather loud with some strong opinions on what doing this would mean to that.The interdependencies of your work affecting somebody else’s work is why it is so important.Now if you have a discrete function like Sales or something like that it doesn’t really matter.
Having been a part of multiple remote teams, you can get these arguments with remote workers, they just take more coordination.The question is, at what point are does the cost of real estate/rent/attracting talent outweigh the additional daily coordination costs.
This is correct and I agree. In places like Delaware it is a competitive advantage.
It is amazing how much you can interact with your team without being in the same building.Probably part of my support for remote work is that I see companies often not able to hire the best person for the job, strictly based on location. “Best person” is relative — for some “best” includes what the person contributes to the environment/culture by being onsite.
I went independent 2 months ago and all of my work is now remote. It can work if you have the tools and the team is dedicated.
Yes, definitely some roles are more conducive to being remote than others.I currently have a remote team and am having to think about whether some of us need to be housed together. I’m sure as we grow that at least part of the team will be together but I place a high value on having and offering flexibility.
The real issue is the informal conversations and overhearing discussions.If that doesn’t matter, then there is no difference.In some cases its better. Frankly having sales separated from development can have some benefits, development doesn’t get into sales, and sales doesn’t get into development.
Yes, I think in some ways totally remote teams are a better situation than some remote/some local. That way, all the watercooler conversation gets forced online–and much becomes searchable.
On that we can agree. I like having everybody local, you like everybody remote. Is one right or one wrong? No. Just like so many things in development. When people get into “religious wars” it prevents things from getting done.
That’s a valuable asset
Mark Essel in the house!
Swapping from Vimeo back to a startup downtown (Sp0n) has given me more time and tickled my startup curioisity. I was head down building for a few years (since April 2012).Hope to rekindle old habits 😉
In some cases its better.Yep!
That discussion could very well exist within live video chat, email or jira/asana comments. Live presence is great for discussions, not as good for focused productivity once the direction is set.
Sales is far from a discrete function. It depends on Marketing, Finance, Product, HR, Admin. I’ll argue that all six critical success factors for any business (Capital; Product; Distribution; People; Facility; relationships with governments) are inherently interdependent.
I have 5 management principles and consider myself even as the CEO first and foremost a salesperson (that changed over time).1. Find and keep the best people2. Orchestrate between departments3. Make people do the 10% of stuff they don’t want to do.4. Visit and understand customers and potential customers5. Set the strategy which most importantly is what we don’t do.That being said, salespeople can love selling internally when their only job should be external and coordinating internally. That is why they can work remote.
Not sure if I talked to you this at all before, however this is a societal issue we’ll have to deal with sooner than later – and worldwide.
This is due principally to the fact that after 30 years while we might be in a 1-way broadband world, we are still in a narrowband 2-way world. We need settlements and open access to get to a world of ubiquitous HD/4K 2-way many to many collaboration. Few can really appreciate or understand what it will be like walking into almost any room (they’ll all have a large screen that serves as a picture frame while not in use) and being virtually there with any other room on the planet. Yes we all know and use skype and facetime but that’s like using a horse carriage as compared with a Model T. Early days still in the modern digital revolution.
People have it all wrong when it comes to data, once it is cleaned up it is fairly trivial to make sense of (to the degree it is determinative). In the end, the mean and standard deviation rule the day.
Software is still the oil for it is the only the data has been generated
This goes back to the excellent discussion here on avc and in my post ( http://awe.sm/cMv2P ) about the the platform advantage that the incumbents have in a data driven world.
Great point Arnold. If anything, it’s digital IP (software, data & protocols) that’s the new oil. Combining this ownership will prime digital real estate leads to a cash generation machine.
Right on (and communities!)
Not sure.The fit of any oil to any energy demand is pretty strong (not perfect)The fit of diverse software to diverse problems is zeroSoftware is inherently tailored (expensively) and then reproduced (cheaply)Oil is recovered (expensively) as a generic then refined /tailored (cheaply)So in this sense Software is more fundamentally limited than energy – Thus software ownership generates lease revenue – but oil consumption is required to generate revenue.I thus agree Oil != Software – but I disagree on the reasons – they have very different exploitation patterns (One is about Land and the other IP)
I think Fred’s point is that controlling the supply of oil is (or was) basically a license to print money because of its pervasive role in modern economies — in the sense that without it most countries would simply grind to a halt.The question he’s asking is: can the same be now said of software?
Ok – I get that. And yes.At the same time I doubt that it is possible to control the supply of IP in quite the same way as it involves far more abstract or immaterial assets – So long term no
Controlling the supply of software, interesting. Can that be done?
Sure, at the control/addressing layer (principally of a network). But control can be derived in many other areas of the informational stack.
“Software is inherently tailored (expensively) and then reproduced (cheaply)Oil is recovered (expensively) as a generic then refined /tailored (cheaply)Very true but aren’t their respective “expensive-part” trajectories quickly moving in opposite directions ” ???That expensive “software-tailoring” will become ever cheaper as data collection/synchronizing/filtering/redistribution software-processes commoditize into reusable/remixable generic substrate components easily tailored via widely accessible assemblers taking marginal costs to near zero.That “software-tailoring” commoditization curve may take some time but its software-data “cash/capital magnet” could/should be much less enduring than its industrial age oil counterpart ???That expensive “oil-recovery” process, on the other hand, is ultimately and inexorably heading in the opposite direction.
>> Very true but aren’t their respective “expensive-part” trajectories quickly moving in opposite directions ” ??Yes – very true – All in all this Software to Oil thing is pretty tenuous.Often a superficially attractive analogy or model breaks down under some use cases. This does not mean the model does not have good explaining power – but it means it is limited and must be applied / interpreted with caution.
In my life new languages and libraries have gotten easier to use, but have also grown more complex as devices and library authoring communities mature and grow more complex.I’m not sure it’s easier to develop now than it was 20 years ago. Scalable web servers have gotten easier to deploy.Specialized tools have grown very powerful (theano for deep learning, opencv for image and video processing, etc).
So owing the IP of a successful piece of software does actually allow premium rents to be extractedWhat’s interesting about Microsoft in this respect is that they didn’t make a ton of money by making good software (IP) they made it because they controlled and cut in a vast ecosystem (“the tech guys”) that was able to profit from helping people use and deal with their shitty buggy insecure software. If the software ironically didn’t have so many issues and gotchas there would be less work and profit for an entire industry that grew up around supporting that software.
Bugs? They’re not bugs! Their features! E.g., tell XP users that bug fixes for XP will stop, should upgrade to Windows 7, etc., and bug fixes there might tip off hackers to unfixed bugs in XP! So, get rid of XP and buy a new version of Windows now!
Part of the reason this is possible is that for some reason when people are using computers they are willing to believe that it’s them and not you that is causing the problem. Imagine if people reacted to computer problems the same way they reacted to airline problems or car problems. I have to give more thought to the psychological underpinnings of all of this. The root of why this phenomena exists with computers. I think it has to do with the collective mentality of the users. As a result even if you are able to call bullshit on their bullshit you can’t because there are so few voices able to do that.
This bug, especially security, situation is serious and for some customers really, really serious. So, I have to believe that there are plenty of customers who really want some quite solid guarantees that their operating system and other platform infrastructure software is rock solid, safe, and secure. Some operating system vendors used to try to make some guarantees, but I’ve seen next to nothing since. Yes, were some efforts, some US DoD funded Ironman projects; IIRC Ada was supposed to be relatively secure just as a language; Multics was very careful about security; there were some supposedly severe DoD security standards that IBM’s MVS was supposed to be able to pass; maybe VMWare has some guarantees; there was supposed to be a version of Linux under development intended for fully serious production use in telecom that was supposed to be rock solid and secure, etc. But I haven’t seen much on such guarantees recently.If I get to be a serious customer of Microsoft, then I will ask some pointed questions about Windows Server, IIS, and SQL Server, etc.
Throw in a cost of termination to every session (regardless of how small) and you will drop security threats dramatically. Of course this requires knowing the identity (trust) of the originator of the session. That’s one reason we need settlements. The other two are a) price signals which serve as incentives for coordinated, rapid and ubiquitous investment (outside of silos) and b) conveying value from the top of the stack (software) to the bottom of the stack (hardware) and from the core (where the directories and controls are) to the edge, where the heavy infrastructure (access) costs are.
Data is part of the history of biz. What biz, w/ varying degrees of sophistication, hasn’t explored data collection and mining? Software has just made it easier to aggregate, parse and analyze, yielding greater efficiencies and effectiveness.Software, AR, IP, AI, Data, etc., are all subsets of Tech, which, like oil, has many tentacles and has spurned many biz and offshoot industries.I think yesterday’s “tech is the new oil” is a more apt descriptor.
The “scarce” oil or real estate metaphor will only go so far.source: https://prezi.com/xmzld_-wa…
Great overview of Economics as an evolving “Living System”.It’s a real perceptual-ergonomic “GEM” !Meaning it quickly rewired my brain by effectively amplifying my visualization/remix access-control to a large set of relationships I already understood in a less reusable mental syntax.My only disappointment is that it ended just as it started to get really interesting, just when I thought it was about to map the generic reusables associated with “Living Systems” onto economics as an instantiation of a “Living System”.
Fascinating post, Fred. David, a segue: the trouble I have always had with the “zero marginal cost” notion around software is that it turns attention to the cost of creation. But rather, one could argue that “manufacturing costs” are really just distributed in different ways. Hardware, at least before possibilities created by 3D printing, required a large upfront design investment, but then design was over and you just focused your costs on duplication and distribution. Software, on the other hand, is more like having a baby — it requires constant iteration, improvement, etc so your ongoing design and engineering costs are significant and usually grow significantly.
Hi Giff, long time no see. You’re right that software generally requires constant investment, but I would say that reflects the beneficial impact of market forces: iterate or die.But I don’t think it really makes a dent in the underlying economics of software replication. I remember a documentary from long ago in which Carl Sagan pointed to a massive cruise ship and said, “Imagine if we could clone physical things, just like we do software…”
“So is software the new oil? I don’t think so, I think data is.”Since data control/ownership is accomplished via software maybe that is just semantics ??? So with that framing in mind I’d like to once again quote Eben Moglen.If you like you can change steel to oil and software to data :-)”Software is what the 21st century is made ofWhat steel was to the economy of the 20 th centuryWhat steel was to the power of the 20 th centuryWhat steel was to the politics of the 20 th centurySoftware is nowIt is the curial building blockthe component out of which everything else is madeand when I speak of everything – I mean of course:FREEDOMas well as Tyranny as well as Business as usualas well as Spying on Everyone for free all the timeIn other words the very composition of social lifeThe way it works or doesn’t workFor USThe way it works or doesn’t workFor those that OwnThe way it works or doesn’t workFor those who Oppress All now depends on software!” – Eben Moglen
Well done David!
So we go from a post yesterday, speculating that Venture Capital might be in trouble, to this today, saying that Software is like the new Oil and it’s taking over the world.Never a dull moment on AVC.
that’s my goal. keep it interesting. one way to do that is zig and zag.
It also creates opportunities for more people to comment with different kinds of thoughts and viewpoints. A bit like the Socratic method.https://en.wikipedia.org/wi…
Fred was actually quoted in the print WSJ yesterday which I read.It was this story, online version contains a link to AVC:http://www.wsj.com/articles…
Ah, didn’t see that. I’m logged into WSJ but it’s not giving me access.
Ok so what you do when you hit a WSJ paywall, any WSJ article, is simply google the title:”The Dangers Ahead if Tech Unicorns Get Gored”…then follow the link and it will give you access.Wow – I just tried the same trick with duckduckgo and it doesn’t work. Gabe needs to get on that one, eh?  (It also works from bing).
…and why the EU and China are on their guard about US websoft companies. it’s digital imperialism.
Well once that software had eaten the world, only logical that it would indeed become the new oil. Now the net cash amounts generated by the big 5 is staggering indeed.
It is always interesting that no one ever seems to ask “why” things like oil companies or banks or software companies exist. What’s the motivation for these things in the first place?
I take it you don’t think it’s obvious?
for the rest of us, I believe the most valuable companies in the future will be those where technology compliments human effort – a hybrid of human effort and the tools of technology and software.
Airbnb is a great example of this. There won’t be a “driver-less” apartment a la Uber anytime soon.
thanks for the reply Joah! Lots of examples, of course, but perhaps the paradigmatic example of this is, ironically enough, Facebook. Remove “humans” from the platform, and see what’s left.
We can all be modern and of the moment. Example: Architecture, design, and urban planning.
thanks for the comment Karen – I think you just defined the utility of technology = “modern of the moment”!
There is a lot more depth to this comment than meets the eye. The value of technology is that it releases humans to do what they do best and to do it better…and more of it in some cases. Technology is a tool.
and in a sense (relative to business and venture capital) technology and software enables humans to not only “do what they do best” and “do it better”, but presents us all with the best opportunity in history to separate ourselves from the competition… as economist Sam Wilkin so eloquently puts it, “don’t be the best… be THE ONLY!” — thanks for the reply Donna!
agreed w your pt, and this is a bit tangential, but how do you think about apple/hardware? the premier oil service / pipeline company that also benefits?
The new oil, with infinite possibilities of input and output.
Look at the title of this post alone and you can see why Austin (Texas in general) has been a much much better city for software / enterprise / SaaS startups built from the ground up than consumer startups. Even the rollup plays like HomeAway and RetailMeNot have a very oil-like approach to building consumer businesses where they see who has struck oil and buy them out. I can’t wait until we get out of this oil mentality and start thinking about alternative fuels. I mean consumer startups…
‘Software is the new oil’ was my favorite part of the comments from yesterday’s post.1) re. cashflows from oil drying up – how’s this for a side-effect: http://www.bdlive.co.za/bus…2) Interesting to me is the infrastructure side of things – at the moment software is still beholden to governments for licensing and large-scale infrastructure funding (fibre-optic cables, power, telco transmission towers etc.) Software cos. need to get over this hurdle for the true potential for cash generation to be met. Govt. won’t give this up easily though, especially in the developing world: https://citizentv.co.ke/new…
it was “tech is the new oil” in yesterday’s thread but i modified it to software. i think that’s more accurate
Ah yes – I guess my links are more referring to tech, than software specifically.
one problem with oil analogy is the payer. in that the number of users paying for FB, zero, number of users paying for Google, zero, number of users paying for Microsoft, falling. Imagine if oil were free and were ad based.
Owning oil producing assets is OK. Being able to exploit oil producing assets is where the money is made because that is where the value is added. Commodities have been crashing in price for 100 years. Its an asset with a declining value that has only gone up because of inflation. Your software analogy is too broad. Maybe owning a server farm is akin to owning an oil producing asset but thats not the google algorithm and data set or facebooks massive active user base.
Wondering aloud what the national security implications are – not only for the U.S. but other countries. Access to oil has been one of the drivers of foreign policy.
that’s a worthy thing to wonder about. i mentioned it in the third to last paragraph herehttp://avc.com/2015/10/winn…
How about this as a security issue – http://www.nytimes.com/2015… ??
That is awesome.
Access to data (new oil) is already under review … EU Safe Harbor and EU going after GOOG (which may or may not have affected its transition to Alphabet):* http://www.forbes.com/sites…* http://www.theregister.co.u…
In shooting wars, oil/fuel supply lines were heavily targeted.In the economic wars, you get threats like this:http://www.nytimes.com/2015…
This is a tough security problem. In the way we do things now, access to data and information equals lowered security barriers. More security barriers mean less access. If information is power, and data is oil, how can we use it without compromising it? Nobody is really designing that future, IMO. It’s all just bolted on security at this point. “Cover the pipeline with steel armor” works against some threats, but not all.
I think from an economic stand point, you have it spot on. However, in Big Oil, many of the really large cash flows giants are sovereigns, like Saudi Arabia, Iran, etc. They are entities who plow dollars back into public projects, infrastructure, etc. Perhaps the big Software barons, like Benioff and Gates and Zuckerberg will do the same, it will be tough to match the scale and consistency of a sovereign.
yeah, i was thinking a lot about that as i pondered this line of thinking. i didn’t blog about it because i’m not sure how that is going to play out.
Probably question of data versus software. Oil is data and refineries and Big Oil companies are tech.I’m a huge fan of Alberts thinking here. Would lead you to believe that if we can get control and ownership of data back to the people we could solve huge problems aka BIG
see my comment to sigmaalgebra above re settlements. as you know it’s something i think the net neutrality folks do not fully comprehend.
I would say google is doing something like this with Fiber
Not even close because the sovereigns have access to all of the profits (even if they shave off/steal some for their own benefit) while Benioff, Gates and Zuckerberg just have whatever money they can make by selling their stock or their other investments. Not even close in wealth terms/impact. Not to mention once again that software is not as secure as oil was or even is now. That is a big issue.And it’s quite obvious that public corporations are not going to (and can’t) be giving up the lions share of their profits to plow into pubic works although they do support, um, PBS and Susan Komen et al.
Anyone know what would these balance sheets look like if they weren’t avoiding taxes?
Can’t help but wonder if there are small oil companies like software ones? And if there are not, I can’t help but wonder if there is a future for them (small, independent software companies)? Okay… Maybe I’m taking this comparison a little too far. 🙂
is the barrier to entry in software as high as energy? i would have assumed no…
Oil is a store of energyWork is application of energySomething that does work (oil) is a substitute for any workerTherefore cheap oil eats labour.Administration is a form of workSoftware can assist administrationTherefore cheap processing eats labour.But also cheap processing eats oil.Technology can reduce reliance on primary energy requirements so given availability of sophisticated suitable software a correlation cost of commodity processing and energy prices is expected.This truism will be witnessed in transport, distributed manufacturing and anywhere that oil competes with labour. – So in pretty much any labourious or productive effort.The keys to the wealth are held by those who can deliver the optimisations (whether organisationally, politically, or technically)A recent economist article speaks to thishttp://www.economist.com/ne…
This explains the cash building / surplus analogy really well James.
TY – It is something we have been looking at for a while. (And not all my thoughts by any means)
True. From the beginning of history we see ideas as the chief product and networks upon which those ideas were both promulgated and fed as key to success/failure (ie those who optimize).Early times = agrarian, physical force (food fed human capital; control over river and road networks and city hubs key to success)Modern times = industrial, mechanical force (oil feeds the machines which empowers the human capital which needs more food; control over transportation networks (particularly pipelines, but also railroads) key)The future = digital, software is the new force (power (increasingly from diverse and more sustainable sources than oil) drives the software, which drives the machines which empower the humans which need food; wired/wireless networks key)Energy translated to output is still central in all instances. History does repeat itself. Knowledge creation and network effects (and the boundaries and laws that govern them) are pretty much the same but relative and absolution consumption and value do change, as well as the velocity of information and ideas.
When is Google going to realize they made a huge mistake calling their company Alphabet?That is all I want to know.
I don’t think the software is the whole story…it’s maybe the refinement process (and thats very important)…but really the knowledge that the combination of software and data produces is where I think the real power and value sits.Or put another way, it’s not about the machines…it’s about the people…
You and @annelibby:disqus are on the same wavelength.But you know the saying…great minds.
This point may have already been made. In economic terms I think software is more akin to the assembly line than oil. Oil is an input to production. Software is a know-how or expertise in production, where the software serves to create the product. I think that this is generally true whether we look at production systems such as Google or Facebook or consumer apps like Uber or Evernote (where the front end is just access to a sophisticated infrastructure).
Oil may be a raw material input to production (as in plastics), but more commonly it is a highly substitutable transport medium for energy where alternatives include – grain ethanol, gas, hydro-electic, solar, nuclear, wood, wind, waves etcUntil we are clear about how we define oil and what it is for, any analogies with software will be brittle at best.Example – You say >> “software serves to create the product” – I think that fuel in a blast furnace equally “serves to create the product”So while your thinking is interesting it is not very robust.
If [ ] is the most valuable commodity in the economy, then those who create the tools to discover, extract, refine, and distribute it for consumption will naturally rise to the top of the pyramid.I don’t see any reason why OIL can’t be replaced by INFORMATION in that blank.EDIT: exchange “commodity” with “unit of value”, based on good commentary below.
Information is a better analogy for oil than software. Maybe software and information are fungible though?
A train timetable provides wholly non-fungible information unless you really don’t know where you are going.Oil is however pretty fungible until we compare Sweet Brent Crude with Heavy Sulphurous Bunker Oil
Information shares the same supply/demand drivers as any resource. There is indeed a (sometimes huge) cost to information production and consumption. That said, we are basically unbounded by how much information we can produce and consume (yes, I will have AI agents in the future consuming for me). The key, given the enormous variability of supply and demand, is efficiently matching both the supply and demand at some point in the not to distant future so as to generate an ROI on rapidly obsoleting software and hardware. That said, there is the aspect that information in one person’s hands maybe more valuable than in another’s based on how they can process and distribute the resulting output. Oil’s not the same in that sense.
This is a fair criticism of the concept of commodity to be applied to information as a unit of value in the economy. I think we can probably separate the necessity for something to be fungible, for it to be valuable in the economy… so maybe substitute “commodity” with “unit of value” to make this work.
I have trouble seeing information as a commodity – it is by its very essence discrete, Data can be thought of as a commodity (though it is not, the means to gather it are) – Information is however always unique – if repeated it does not inform.
I think you’re right that it’s a mistake to apply the concept of a commodity to information. maybe substitute “unit of value” to information to make it more interchangeable with oil in this analogy.point being, a unit of value need not be fungible to make it highly valuable in the economy.
Ok – fully with you on that !
a reason why solar power is not well liked by vested interested and the technocratic elite class.
Most software companies spend 10-16% of revenues on capex. That doesn’t really change, whether the company is small and growing fast or large and growing slowly. It applies just as much to Yelp or SAP or Workday. At first sight, it looks like software R&D expands to match available resources and has little scalability.And actually it fits the oil analogy – oil companies don’t make any profits either, they just sink the money back into drilling more holes in the ground.
Like this approach, interesting how few VCs speak to industry’s financials
We should not conflate the organisations that operate software with the IP that they operate or the market context in which they do so.If you remove the commodity cost of a technology stack , and assume pseudo efficient financial markets the remaining value is directed by barriers to competitive entry.This means IP and network effects, and perhaps regulatory friction which generally protects incumbents but has (I think ) little in the long term to do with governance structures.So I suppose the board members of most tech. companies and the software structures they provide are non-irreplaceable (if not quite easily cloned).So I think your argument has short term merit – but I believe very strongly if they “sit back and collect” as you suggest – they will disappear in a Kodak moment
Humans will be the new oil once robots take over earth. The more humans you .. err.. own /control, the more wealthy you will be.
it was ever so – Ask Genghis Khan, The Roman Empire etcEqually these structures are non-sustainable – Ask Genghis Khan, The Roman Empire etc
It makes sense to me that software is the oil of the information revolution. Companies that control the software infrastructure of the information revolution will sit back and collect the economic surplus of the information revolution and that will be a path to vast wealth and economic power. It’s not clear to me how you can make such a broad statement like that. Owning a patch of land that contains oil or being in an established legacy business that serviced the oil industry is quite different than being in anything related to technology. The vast number of businesses operating in tech are not printing money in the way that the Beverly Hillbillies did. There are of course some that you mentioned (and many that you left out). But the comparison really ends there. Oil was as close to a sure bet for many participants that you could get. Not the same with tech. For every company like Google that made it many others failed.Apple is printing money now. And the fact that it didn’t print money prior to Jobs return (while Microsoft did) just goes to show how different Oil and Software are for a host of reasons.By the way if Alphabet was as secure in the future as you seem to think they are they wouldn’t be twittering away their time on self driving cars and other moonshots. They’d just continue pumping the oil which is what Oil companies had been doing for the last century.
Exactly. Nobody is going to “sit back and collect the economic surplus of the information revolution”. Good luck with that strategy.Several aspects of this analogy of software and oil don’t work. At a high level of abstraction, I see Fred’s point…but thats about it.I actually hope Software in the 21st century does NOT play out the way Oil did in the 20th Century.
+100>> I actually hope Software in the 21st century does NOT play out the way Oil did in the 20th Century.It would be pretty dispicable to hope otherwise
Control of Infrastructure Software It makes sense to me that software is the oil of the information revolution. Analogies are weak forms of evidence, but, for an analogy, okay. Companies that control the software infrastructure … The infrastructure is usually thought of as, say, the virtual machine, hypervisor, container support, the TCP/IP stack, the file system, the operating system, the device drivers, system management tools, runtime software, compilers, software libraries, development tools, etc.For Microsoft, that’s heavily Windows, Windows Server, the compiled languages, the common language runtime (CLR), the .NET Framework, SQL Server, etc.But the Linux and open source software (OSS) side is a significant competitor to Windows.For the infrastructure for my startup, I’m using Microsoft, hoping that if I pay more then I will get more, but a lot of people are betting on Linux and OSS.Point: With “software infrastructure” as I just described it, no one has “control”.Vast Wealth … a path to vast wealth. I think we are just beginning to see the operating leverage of these software based business models. In simple terms, the “vast wealth” is due to massive automation. E.g., Google’s software and server farms automate the work of a lot of reference librarians and more.Amazon? Compared with the alternatives, so far mostly just brick and mortar, Amazon has bigger and better selections, better product information, better pricing, a very well done Web site, and really good execution, e.g., for paying, shipping, tracking, etc. Bezos is beating the Waltons (@JLM) “like a rented mule”, like the Cavs against the junior high hoopsters.Right: Google, Amazon, and Microsoft all saw long ago that just for their internal computing needs they needed essentially a cloud, one good enough that serving outsiders also would be just a flip of a bit. So, poor HP, IBM, etc. are standing around wondering how they missed out.The point is not really “operating leverage of these software based business models” but good, rather than nearly brain-dead, execution. Such poor execution is especially strange for IBM which long, in one form or another, ran the early versions of a cloud, a service bureau, facilities management, and long ago strongly encountered and did well with some of the main issues in a cloud — security, virtual machines, performance via hardware parallelism, reliable clusters, etc.Making Moneythe cash will be flowing out of software companies the way that oil flows out of wells.Yup: Software never sleeps and never wears out and is the key to much more in automation — “machines should work; people should think”, and enjoy life!Barrier to EntryClouds: Likely HP or IBM could pluck just the right half dozen people from the cloud operation of any of Microsoft, Google, or Amazon, say, for less than $10 million per person, less than $60 million total, i.e., chump change, heck, let’s be generous, $100 million a person, $600 million total, still chump change, and have them guide a team of otherwise readily available, low cost labor to bring up a competitive cloud operation in about the time the brick and mortar guys could have the building, power, HVAC, network connection, and raised floor ready.Why? Such a team knows how to draw the architecture diagram and what parts and pieces to get and how to plug them together and, then, how to do the system management, customer support, billing, etc.E.g., once I did a small version of that: While in grad school, as a small part of a part time job, I ran a time sharing computer center. It was good. Then (trying to have time to help my wife), I became a B-school prof and quickly, as just a small effort, did the same for the B-school, was made Chair of the computing committee, helped the university get a new CIO, etc., all for only about a day a week at first and then next to nothing.Why? I knew how to draw the architecture diagram and what parts and pieces to get and how to plug them together. E.g., get some Belden five conductor, general purpose signal cable and some certain, easy to use AMP plugs, get some plastic, carbon ribbons, get some A/C with a special valve to permit running in cold weather, get the secretaries using access control lists (ACLs — still the key to much of computer security), etc.Side Effects: Soon the staff got a port mapper box. Soon some of the student workers set up a nice system for the alumni office to send lots of beautifully done, personalized letters to formerly neglected alumni — the money raised likely paid for the computer center and staff many times over. The system was so good the university’s CIO shop took the design and architecture and rolled it out for all the schools on campus — 55,000 students. Making big bucks!Point: So, the cloud guys have no very significant barrier to entry.Google: IMHO, as a search engine, Bing is fine. Apparently Yahoo, Facebook, Baidu and more know how to build a search engine that can compete with Google.Point: So, Google has no very significant barrier to entry.Amazon: Likely people in Bentonville are now up early, wide awake, and awash in CapEx. There’s no biggie reason Wal-Mart/Sam’s Club can’t do what Amazon has done. In addition, Wal-Mart/Sam’s Club already has (A) highly refined purchasing with product suppliers, (B) warehouses, (C) retail stores, (D) trucking and more in logistics, and (E) home delivery.Point: So, Amazon has no very significant barrier to entry.Facebook: Sorry Zuck, each time I visit Facebook I’m shocked at how just plain sickly horrible the Web site and its Web pages are. Commonly the Web pages sent to my recent version of Firefox on my 1024 x 768 pixel, 14″ screen are just unusable. Bluntly, I have no idea how the heck to use Facebook and no desire to struggle to learn. Great candidate for the worst significant Web site on the planet. Really bad.Anyone with a foot into the social network, e.g., Instagram, PInterest, SnapChat, Twitter, and likely many more, could just start another Web page as the home page of a start on a replacement for Facebook, grow the functionality of the site, and quickly be one heck of a competitor for Facebook. Then Facebook would have to redesign it’s site and, thus, lose most of its present user lock in and greatly lower user switching costs.Point: So, Facebook has no very significant barrier to entry.NetThe keys to the future of big bucks in the information revolution are not illustrated by Microsoft, Google, Amazon, and Facebook. Instead, we need more including with some significant barriers to entry.
Well, no significant barriers except branding, at any rate. For overcoming that, well, a certain logic puzzle comes to mind: https://xkcd.com/blue_eyes….Granted, that could just be a red herring. But I think it’s a useful way of thinking about behavior, and the threshold effects that show up when it comes to networks.
We have to look at the present in a trendline which starts with the telegraph in the 1840s (1.5-way digital communications) and veers off course in the 1913 (wired 2-way) and again in 1934 (wireless), where centralized governments erected unnatural information monopolies. Had we known then what we somewhat know now (and we’ve managed to forge a number of important lessons along the way), we might have started this glorious digital information revolution back in the 1930s-40s rather than waiting til the 1980s-90s for it to kick into gear. The past 30 years was all about playing catch-up and even then legacy institutional structures and regulations are impeding growth today while processing, transport and storage costs drop (the cost to bring a new product/idea to market and consume it) drop 30% annually. No reason other than these legacy mindsets and institutional structures that, given current technology, 4K VoD, 2-way HD collaboration, mobility first (not only), and IoT aren’t pervasive and cheap. Instead a world of silos that are either expensive or try to invade our privacy.
I like yesterday’s “tech is the new oil” analogy better. (Software is a large subset of tech, while oil, like tech, has many tentacles.) Tech and oil are both boom or bust speculative industries, although oil is far more capital intensive. In both industries everyone is drilling for a gusher, yet few hit pay dirt.Think Jed Clampett. (Love that Ellie May).http://i275.photobucket.com…
I wonder how much economic impact came to Beverly Hills as a result of that show. You and I were raised on that show. What would we think of that town if that show never existed? In a similar way it’s hard to question the impact that SNL has had on NYC. (Not to mention any other shows mentioning both of those cities and pounding it into our brains at an early age).
Television certainly had a much greater cultural (and likely economic) impact in the 60’s-90’s. No more. It’s been displaced. No doubt location based series (top-of-mind Twin Peaks, Frazier, BH, etc.) enlightened the masses about different geography and likely impacted tourism.
The connection between Thomas Edison and Hollywood is a great story.
Elon, Larry, and Sergey are the new Koch Brothers / T. Boone Pickens?
Been saying stuff like this for a long, long time. http://seekingalpha.com/art… But it’s always good to hear a pro’s take.The key is that software is limited. It’s still made by hand. That makes it scarce.
Actually, as we get higher and higher abstractions, more and more of it is made by machines. Compare file I/O C -> java -> Ruby.But there just appears to be insatiable demand for software, probably because it is so customizable.
I find this a depressing vision of our future, though I know it was not intended to be dystopian. This future is one where we all depend on commoditized software piped to our homes and devices through vast infrastructure that is controlled by monopolists. Ugh! Screw that!
Same as for the robber barons, there’s bound to be some legislative backlash ? Hopefully, something better than War on Drugs and banking legislation ?
I am glad you brought that up! In general, governmental intervention in private transactions (e.g., the decision to use gmail) is inelegant. However, I think there is a role for antitrust authority in software, particularly in the way that software standards are used. The old, bad version of Microsoft used to have a play called “embrace and extend”. That is, embrace a new protocol (such as .txt) and *extend* it so that no one can use it who doesn’t pay for Microsoft software (such as Word). That kind of conduct is arguably a violation of antitrust laws, at least in spirit, and I think an enterprising US Attorney General would pursue it through her trust-busting authority. I also would be kind of open to laws that force walled gardens to make their users’ data portable.
Innovation is the new oil from a cash flow perspective and Solar and wind are new oil from the energy side. Apple’s business is sustainable only if Apple retains its innovation edge. To sell its products at a premium, Apple has to keep innovating and leading with features. The same is true for Google and Amazon.
I’m utterly unconvinced that Apple is leading with features. They’re leading in branding, and getting that smartphones are like handbags now, but features and performance are a wash with others.
One major difference. Oil is an “industry standard” – everything runs on the same thing. In software, however, it’s all different. Rails, Grails, Ruby, Java, Python, etc…you’re betting that your language will be the prevailing language forever, whereas oil is the universal language everywhere. No?
there are standards of oil too, most notably WTI and brent.
Just wait until the taxman starts taxing software as it taxes oil. At one moment the European taxman got 300% more revenue per barrel of oil that the country giving up that renewable resource for ever… (and as for some of these the ad-blockers are just around the corner)http://teawithft.blogspot.c…
Software is the new Oil and Unicorn Valuations are the new Snake Oil.
Hopefully we won’t get into as many wars and support as many shady governments over apps as we do over oil ? Sorry, I meant “democracy”.
Too bad , Twitter is not an oil but an useless software today. Twitter down to 26
A slightly more accurate characterization of the two ages in question is perhaps (a) the power age, and the (b) information age.The power age was/is the industrial revolution. Power is required to harvest and transform resources and is the key resource is this economy.The information age is differentiated by the fact that power is incidental to information.It was never just oil, coal was and is a hugely important energy resource. And if you want to look for a transformational technology wrt the industrial revolution then look to materials science and fusion.The ability to deliver cheap power will never be unimportant. We have barely scratched the surface of the information age.
Oil companies do not have the volatiility of software company. With the rise of smartphone in a few years time frame, Microsoft is dethroned from the King of software. The next disruptive technology will also dethrone Apple or Google at unexpected time
Data is the new oil, software is pumping it…SW without data and processes has no value, except games and creative software like Photoshop or Illustrator.
After going through thefinancial figures, one has to admit that, to run a successfulbusiness, you need to have access to cutting edge software. For smallbusiness owners, having access to a good sales automation software isessential for carrying out all crucial tasks in a hassle free way.
Ask the FTC!
I like the example case. They do not own all the oil but they own the pipelines.