The Golden Age Of Open Protocols

Protocols are a geeky topic. It’s way more interesting to talk about applications. People will go on and on about why they like gmail or some other email application. But not a lot of people get excited about SMTP and IMAP which are the underlying email protocols that allow the gmail application to talk to other email applications.

Open protocols are at the heart of many of the most important systems that we have. The Internet works because of TCP/IP. The web works because of HTTP. Email works because of SMTP. These are open systems that developers can build applications on top of. There are plenty of proprietary protocols out there too. But proprietary protocols tend to lock in users and drive value to the owners of the proprietary protocol, like Microsoft, Apple, Google, etc.

One of the problems we have had in tech is that there aren’t large monetary incentives to create and sustain open protocols. If they are open they cannot be easily monetized by traditional means.

However, that is changing with the emergence of blockchain technology and crypto-tokens. My partner Albert wrote an important post about this last week. Here are a couple paragraphs from that post:

Now, however, we have a new way of providing incentives for the creation of protocols and for governing their evolution. I am talking about cryptographic tokens. You can think of these like the tokens you might buy at a fair to get on a ride: different operators can have their own rides and set their own price in terms of tokens. You only need to buy tokens once (in exchange for fiat currency) and then can use them throughout the fair. With blockchains we now have a way of issuing and redeeming these tokens digitally (the underlying blockchain can be Bitcoin or Ethereum or possibly its own as in the case of Steemit).

A for profit company can now create a new protocol and create value for itself (and its investors) by retaining some of the tokens. If the protocol becomes widely used, the value of the tokens will increase. For instance, think of a decentralized storage service (a la Amazon’s S3). Anyone can implement the storage protocol in whatever language they want to as long as they meet the protocol spec. They can then get paid in the relevant storage tokens. The original creator of the protocol will make money to the extent that it is adopted and to the degree they have retained some of the tokens (so they can sell them at a higher price later on). This is not hypothetical as there are a variety of such protocols out there, including Storj, SIA and Filecoin.

This is super important because the more open protocols we have, the more open systems we will have. If Twitter had been built and monetized this way, things could have played out very differently. In the early days of Twitter, there were third party applications (Summize for Search, Tweetie for iOS client, etc). These were all built on Twitter’s API. If Twitter had imagined itself as a protocol instead of an application, these third party applications would not have had to compete with (or get bought by) Twitter. But at the time, there wasn’t an obvious way for Twitter’s founders and management team to benefit from a protocol-based business model.

In this emerging model, Twitter could have adopted a protocol-based approach and issued a crypto-token, Twokens, that users could earn from things like amassing followers, reporting abuse, etc. Twokens could also be sold by the Twitter founding team to finance their operations. Crypto-exchanges could make a market in Twokens so that anyone who wanted to speculate on the future value of the Twitter protocol could do so.

I am not criticizing Twitter for the approach they took. I was there in the early days and was completely bought into and supportive of that approach. It was the one that made the most financial sense at the time. I am also not saying or predicting that some new Twitter competitor is going to emerge using this new model. What I am saying is that we now have a new model that supports a protocol-based business model and I believe we will start to see innovative new protocols emerge that are based on this new business model. Obviously Albert believes that too and so does the rest of USV. As Albert mentions in his post, we are invested in a number of projects that are taking this approach. It’s a brave new world and we are grappling with what it means for us as investors and how we engage with and invest in this model.

But, as I have said many times here at AVC, I believe that business model innovation is more disruptive that technological innovation. Incumbents can adapt to and adopt new technological changes (web to mobile) way easier than they can adapt to and adopt new business models (selling software to free ad-supported software). So this new protocol-based business model feels like one of these “changes of venue” as my partner Brad likes to call them. And that smells like a big investable macro trend to me.


Comments (Archived):

  1. awaldstein

    Great stuff.Of all the major platforms today, are any really a function of model innovation? Seems like innovation around engagement and distribution but same old media for the most part around monetization.Am I missing something?

    1. Chimpwithcans

      How about Wechat? I just signed up for WeChat here in South Africa. Anyone know if their model creates such tokens (Wechokens??) Or is it only linked to traditional currency? Seems to be innovation around banking on Wechat….not tried it out yet though.

  2. kidmercury

    this assumes that people will pay to use the protocol/API. paywalls always reduce the size of the network. it may be advantageous to proceed in that direction anyway — i actually am more bullish on that type of a trajectory at some point — but it is a major change with many economic implications.anyway, exchange rate is the real issue here. it will get solved, but only when the dreams of anarchic utopia are relinquished.

  3. William Mougayar

    “business model innovation is more disruptive that technological innovation”.This is key. That’s why big companies can never take advantage of the new innovation potential. Everything ties back to their existing business models which they don’t want to change.

    1. pointsnfigures

      Yup. But in many industries their balance sheets are great so they will buy innovation

      1. William Mougayar

        “they will buy innovation”. yes, then they will bury it or kill it.

        1. Twain Twain

          Well, then … VCs need to do better at helping that innovative startup to IPO instead of being “flipped” too soon?

          1. William Mougayar

            but that doesn’t depend on VCs. entrepreneurs decide their fate mostly themselves.

          2. Twain Twain

            It’s not an either or situation? It’s symbiotic and to do with founder+investor+market fits and synchronicity (right people at right time in right place).There are likely some “unicorns” (hate that word) that would like to IPO but their investors don’t support that strategy.Maybe because market conditions means they’re unlikely to hit the N times ROI they’d like.

          3. Chris DeRose

            Keep in mind that there are companies making money hands over fist here in blockchain – and none of them fundraised. This space is highly profitable – and completely unfundable.

        2. LE

          Maybe what appears to be “burrying it” is just an artifact of all of the fresh face enthusiasm leaving. Either physically or in spirit. Moving on to new opportunities and a bigger better deal. Or simply losing the drive (as a group) that comes with a large number of employees cashing out. Or a new corporate structure killing the look and feel of what was fun and exciting.

          1. sigmaalgebra

            A common explanation is that the existing big business has long had strong influences from stockholders to manage, fine tune, and focus on the existing business, that is, as defined by its products, production processes, marketing channels, brand name, information systems, customers, etc. So, in that fine tuning, focusing, etc., everything else is pushed out.Also, something new in-house is a threat to many parts of the big organization already set up. I.e., keep out competition, in any sense, e.g., if only for the attention of the C-suite, especially any just down the hall, the floor above, or the building next door.So, really, about the only person who can effectively champion anything disruptive is the CEO. But the CEO was hired to please the stockholders who just want more milk, cream, and butter from the existing cash cow.E.g., for the existing business, have what the accountants wrote. Then the stockholders and the stock market can use those accounting numbers to evaluate the business and price the stock. But for something new, the accountants will report little or no value. Accountants look at a baby and see mostly only soiled diapers, not the future founder of a big, new business.

        3. Chris DeRose

          OMG – you do believe in conspiracies. I’m guessing your the guy who believes that 100 mile per gallon cars weren’t released onto the market because GM and big oil ‘didn’t want them to be’

          1. William Mougayar

            You’re the guy who puts words in my mouth and keeps trolling me.

          2. Chris DeRose

            Everyone knows I’m just a troll, it’s very obvious that there’s no content coming out of BU. As for putting words in your month – I don’t think anyone reading this is dumb enough to think I quoted you in any capacity.But it is very evident that your platform requires conspiratorial motivations in order to maintain any consistency. (Which IMO, makes your platform rather absurd)

    2. Richard

      Aren’t these business models ripe for abuse ?? “OneAwards incentivize the success stories. We award such as Black Di[a]mond and Double Di[a]mond. Being part of this exclusive cirlcle [sic] comes with awards, luxury items such as Rolex watches, Cash Bonuses, and Luxury cruise holidays. Other members will be inspired by your achievement – and realize what huge opportunities One offers.” –

    3. awaldstein

      I can’t speak to fintech and blockchain but beyond that this is no longer as true.I think some of the largest changes that impact the largest amount of people will come from the top down through the enterprise.In mobility. In human capital. In digital transformation.Personally, I see this shift happening as that is where the reach, the resources and the breadth is.

      1. William Mougayar

        “will come from the top down through the enterprise.”- what are some examples?

        1. awaldstein

          Operational mobility as it touches everything from emergency management, wayfinding, and asset management is the one that jumps to mind.If you have 100K employees and can mandate app usage, drive mobility as a channel starts from the top not the bottom.Control is the power of change in the enterprise.

    4. Chris DeRose

      I’m guessing William doesn’t believe deep markets are efficient. Why don’t they want to change william, is it because you believe they are incompetent, or because there’s a conspiracy?

  4. kenberger

    Skype vs VonageIn Autumn 2003 I was interviewing to be CTO of Vonage, and this open vs closed protocol conversation was at the heart of a lot of the discussion. SIP was (still is) the standard protocol, and Vonage was proud to boast how they were dedicated to it.That same week, Skype launched and declared that they felt SIP wasn’t up for the job, and built their own proprietary protocol. Many wondered however if this was a gamble by a new entrant to try and thwart potential open competition later.

    1. Twain Twain

      Thanks for sharing your story. Around the same time, UBS had incubated a P2P chat protocol that was the banking sector’s equivalent of Skype. We spun it out and the other banks co-invested.That proprietary protocol later got acquired by Microsoft in 2007/8.In the closed vs open protocol and systems debate … Well … Facebook remains closed and is worth $350+ BILLION. It’s only open-sourced about 5 developer tools:*…Meanwhile, Google is open sourcing its AI frameworks under Tensorflow.As for Twitter … well … it’s challenged by all sorts of issues.

      1. kenberger

        Another interesting example.What was the name of the chat protocol, and how large an acq did it wind up being? And MS would up taking it, rather than a big financial institution?

        1. Twain Twain

          I can’t share the acquisition amount because it wasn’t publicly disclosed but I know how much it was valued on the books: $ several hundred million.I was responsible for co-managing UBS’s Strategic Investments portfolio and had 100% responsibility for reporting on the portfolio.

        2. Ernest

          It was generally an IRC inspired protocol geared towards group chat. The price was in tens of millions.

    2. PhilipSugar

      You know Skype would not exist without two major protocols, those being the numeric dialing system, and the email address.

      1. kenberger

        the email address is the *protocol* itself, or the namespace convention part?

  5. Dana Hoffer

    Business model innovation… definitely the source of economic profits.

    1. @mikeriddell62

      The real innovation is to share the proceeds with the creators of value. Not that complicated but anathema to the VC world where biggest dog wins.

  6. LaVonne Reimer

    I love the imagery of a seemingly humble “little” open protocol being the key to transforming stubborn old industries. I have been studying blockchain for a while simply because decentralized anything appeals to me. Albert’s linkage to monetizing protocols fills in another piece for me.

  7. David C. Baker

    Is there a recent introduction of an open protocol that you folks think might really take off (other than the blockchain, which Fred already mentioned)? It’s easier to see the open protocols that have flourished–in retrospect–but I wonder how well we’re able to pick the winners now.And from what I understand, Gmail forked the email protocol in a significant way, angering quite a few of the “open” folks.Interesting discussion, Fred. Thanks.

    1. Lawrence Brass

      Open organizations are often coerced by incumbents, a recent event in protocols is Google’s SPDY protocol turning into HTTP/2. From every angle you look at it HTTP/2 is a bad piece of design and engineering with no real benefits outside making ad tech work a little better for them, open advocates didn’t have the strength to fight against it.I think that one of the importatnt attributes that makes a protocol open is exactly that it is free to use and implement. The benefit comes from common use and simpler architectures that lower the costs among participants. I don’t understand profit at the ‘open’ protocol level.

  8. pointsnfigures

    Interestingly, also a way to create a defensible border for a business

  9. ZekeV

    I know USV has invested in the conventional securities of protocol companies. Do you have any thoughts about investors also relying on tokens for return, and not trying to monetize equity investments? For example, using USV capital to fund an open source foundation working on a protocol, while simultaneously buying up tokens in that protocol?

    1. fredwilson

      that is one of the things we are trying to figure out

      1. Demian Brener

        Coinfund is doing that

        1. ZekeV

          Does coinfund also sponsor developers, provide other resources… or do they only buy crypto tokens?

      2. ZekeV

        It’s an exciting idea, but a bit scary to do with OPM. Might require a special side fund with a different structure and tax setup. Timeline for returns might be different. Also would need to figure out how to have safe custody of crypto tokens. Maybe coinbase could provide custody services, but personally I would probably want to have direct custody.

    2. Tom Ding

      One way to think about this is a multi-stage rocket model. The VCs coud seed fund initial protocol startup lab(s) – which produces one or multiple decentralized projects’s prototype and bootstrap to the point of crowdsale (public token distribution). Once the crowdasle completed, then a separate software foundation is setup and with the incoming funds suppport on-going protocol development. As part of the arrangement, the initial for-profit startup takes a pre-allocation of the total tokens which could allocate by cap table pro-rata and hence rewards the VC investors.This also offers a quick exit path (1-2 year) for the VCs (if they wish) since tokens are traded after genesis state, compared to traditional seed investments of 5+ years.

      1. ZekeV

        I want a XEROX PARC for open protocol R&D, but with less clueless management.Hard to get the right team and atmosphere. Should be a group of strong math/CS thinkers and neckbearded hackers, funded by deep pockets with no expectation of quick return. No bitcoin celebrities, and no crooks. Otherwise conflicts would be too great.

  10. Jesse Walden

    One interesting angle on this model is the need for protocols to devise economic systems that convert the speculative future value of their token into a sustainable incentive for their adoption and use. Getting that right isn’t a light matter—it requires domain expertise in both distributed, cryptographic systems as well as a deep understanding of economic and financial systems. Adoption will also beckon new UX to mask the complexity of these systems—how many users want the cognitive burden that investors take on? Hopefully we’ll see more resources emerge to help developers get the fundamentals right in all these domains.

    1. fredwilson

      i wonder if there is anything to learn from how the current system works; early funds from friends and families, capped notes from seed firms, equity rounds by VCs. it took many years for that system to get figured out and evolve into what we now have.

    2. Guy Lepage

      Furthermore, this new UX layer will also require cohesive innovation in both UX and UI to assist with this masking.. We’re already seeing new UX/UI being developed with these new protocols ushering in a new set of UX/UI standards we’ve not yet seen before. I suspect users will not only be oblivious to the highly complex underlayers of the technologies but will see UX/UI like that will deviate from previous UX/UI design standards that we see on the web today. For UX/UI designers, this is an opportunity in time to move visual communications forward from what has evolved into the current traditional and mundane online experience. It’s an opportunity reteach users how to interact with the “new web.”

  11. Matt Zagaja

    I am going super contrarian on this post today, because I don’t think the idea of tokens is compatible with what I would consider an open protocol. Why?Tokens are only needed or valuable if part of the protocol is inaccessible without the tokens. Bitcoin sidestepped this issue because the tokens themselves are functionally merely money and they gave it away at the beginning. Their audience was also geeks. If you need tokens to tweet and the only way to get tweet tokens is to setup a distributed tweet server or pay someone to be in their tweet server, then you’re chocking off the oxygen of adoption from the masses. Conversely if the token isn’t related to anything valuable on twitter then I won’t really care nor try to acquire them.Networks whether they are twitter, Facebook, or GitHub tend to follow power law distributions. A tweet from Taylor Swift or Fred Wilson has much more impact and is read by many more people than my tweets. Unless creators are receiving revenue from their activity in a protocol I doubt they are likely to throw money or additional labor into it (assuming equivalent substitutes exist like Facebook is an ok substitute for Twitter). People are less likely to throw money at a new and unproven market or network.In many ways YouTube is sort of an interesting model here because there is a barrier to entry in time and money, but creators that take off then receive a share of ad revenue. Since creators receive part of the ad revenue that YouTube receives, their incentives are aligned (although the collapsing ad prices are causing the YouTube middle class to hollow out).Finally one of the largest benefits of the Internet is we removed scarcity and gatekeeping. When the risk of trying software or content is zero or nearly zero, we aren’t preventing underprivileged people from being able to participate. The magic of WattPad or Soundcloud is that anyone can upload something which can catch fire and lead them to something like a publishing deal or to have advertisers want to sponsor them. If I need SoundTokens to participate and SoundTokens either cost money or require me to put in resources that I do not have, then my success might never be possible.

    1. LE

      If you need tokens to tweet and the only way to get tweet tokens is to setup a distributed tweet server or pay someone to be in their tweet server, then you’re chocking off the oxygen of adoption from the masses.From the way I read what Fred said, I don’t think that was the idea. I read it as an enhancement of the existing system not a requirement to tweet or anything like that.Specifically:that users could earn from things like amassing followers, reporting abuse, etc. Twokens could also be sold by the Twitter founding team to finance their operations. Crypto-exchanges could make a market in Twokens so that anyone who wanted to speculate on the future value of the Twitter protocol could do so.”Could earn” does not mean “have to earn” (at least from my interpretation).

    2. Cool Beans

      Great Points1/ one of the largest benefits of the Internet is we removed scarcity and gatekeeping. 2/ Tokens are only needed or valuable if part of the protocol is inaccessible without the tokens.

    3. David Semeria

      Good points. At the end of the day a protocol is simply a specification. There have always been issues with how to enforce the monetization of a specification and blockchain won’t change this. The fairground example is different because it pertains to multiple services (in which the underlying protocols may or not be a selling point).In my view this is a discussion about micro-payments for services, for which the blockchain is remarkably well-suited. Protocols have a role to play here because they promise choice (eg choose between x storage servers that all the implement the same protocol).However, the point remains – you’re monetizing the service, not the protocol.

    4. William Mougayar

      Think about what the tokens can enable, via the protocols.

      1. Chris DeRose

        You don’t know that moogie. Thus far the tokens have all been useful for the purpose of pump and dump penny stock scams. The opportunity cost of these tokens is that they encourage a fleet of speculators to move in and dump the protocol as soon as they can make a buck. Also, it prevents the protocol designers from using their ability to understanding the opportunity costs to make unpopular decisions by the clueless and uninitiated (which are the only holders of these tokens)

        1. ZekeV

          Financial speculation is just part of human nature. If you’re designing a p2p protocol, you have to do so in a way that it benefits from nature rather than fighting against it. Your users, on average, are going to be morons at best and more likely fraudsters and thieves. Bitcoin is elegant in part b/c it does not require anyone to trust other users. Of course many still do, and come to regret it…

    5. fredwilson

      i am not sure that is true. you could easily implement a freemium model like SoundCloud or Dropbox where anyone can upload but at a certain level you have to start paying

    6. ZekeV

      Tokens have clearly been successful in bitcoin. I would agree in the specific case of STEEM, that it’s going to fail in the long run b/c there is no need for trustless tweeting. Bitcoin does not solve any problem that twitter has. Alternatively, if one desires a p2p microblogging service with no DMCA or policies, that can be done without any built-in token incentive structure. A project called zeronet is pursuing a very interesting model for this, a p2p publishing network. No tokens, but every site is a bitcoin public key.

  12. John Edgar

    Folks are also not building interesting things on top of the low level exchange protocolling that you noted, for example UDP is amazing and now look at Aspera Inc. One thing that is important to note: We have evolved to the point that I would call some abstract services like WebRTC a protocol, Firebase was also effectively a protocol. If you look at it in that light, generating revenue is super easy, as a modern “protocol” could simply be defined as: A Standard + HTTP + JSON + REST. Ask M.Maltz about how we’re thinking about this.

  13. LE

    This post is a watershed moment for me in understanding the future business potential of the blockchain.This example was great and perfectly illustrates the potential going forward:In this emerging model, Twitter could have adopted a protocol-based approach and issued a crypto-token, Twokens, that users could earn from things like amassing followers, reporting abuse, etc. Twokens could also be sold by the Twitter founding team to finance their operations. Crypto-exchanges could make a market in Twokens so that anyone who wanted to speculate on the future value of the Twitter protocol could do so.

  14. Nick Tomaino

    A key nuance here that may be obvious to people in the space, but not obvious to people new to the concept of crypto tokens, is that the crypto tokens must have utility in the context of the actual protocol.A crypto token sale where tokens represent traditional securities won’t work for legal reasons among other things. The token must empower users to do something new that they couldn’t do before (write dapps in Ethereum, curate content in Steemit, report outcomes in Augur, etc).

    1. William Mougayar


      1. Chris DeRose

        Not exactly. ALL of the tokens issued thus far have attempted to create a market for services that are currently ‘free’ or unneeded. The only reason to buy tokens is to invest in a pitch that can be sold to the greater fool.

        1. William Mougayar

          Are there any good examples of tokens issued, in your opinion?

          1. Chris DeRose

            I think they’re *all* bad investments. Bitcoin included. But, I think bitcoin has had obvious utility, and now that most of the speculation in bitcoin has subsided,I think it’s at least a good time to be involved with that one.

          2. @mikeriddell62

            I kind of like where you’re coming from on this, Chris. Here in Manchester, we’re developing our own social currency which can be earned for contribution to community (where activities are defined and outcomes are measured) and which is linked to a marketplace that will become a ‘social stock exchange’ that links under-used resources such as young people not in education, employment or training, to unmet needs, such as work that needs doing but where due to austerity policies there is no money to pay for it. By valuing the people that carry out the work we can inject value into the points that store that value, account for that value and enable that value to be exchanged with those that want to support the creation of that value by enabling their fair rides to be used during quiet times by the kids that have done good.The new business model that we’re talking about is co-operative. Ok, not that new. Shared purpose, shared vision, shared interest, shared dividend, shared values and principles.

          3. Dominic Wörner

            Hi Mike,do you know any good book about community currencies and their history?

          4. @mikeriddell62

            I’ve read loads but the best is by Thomas Greco: “The End of Money and the Future of Civilisation” Greco is a genius in my opinion!

          5. Dominic Wörner

            Thx! I put it on my reading list.

          6. ZekeV

            Most speculation has subsided?! That is only true with a very short-term perspective. But if true, I would probably rather be involved in a project that is still highly speculative and no one has heard of.

          7. Chris DeRose

            RE: I would probably rather be involved in a project that is still highly speculative and no one has heard ofSure, if you’re highly specialized and was able to asses risk in ways that aren’t reflected by the general market. But you don’t have that ability here in the crypto space.RE: Most speculation has subsidedYes. of the $1.5 ish million USD entering the market every day, the estimates thus far have suggested that about half of that value is used for non-speculative purposes (mostly underserved remittance and value transfer)

          8. ZekeV

            What about the last 24 hours? How much of the $360mm in bitcoin/USD volume is explained by remittance and value transfer? I don’t know where you get your data from, but it is certainly false recently, and highly dubious as a claim even over a longer-term timeframe.In terms of specialized expertise, if you do not have that you probably should not be into bitcoin for more than beer money.

          9. Chris DeRose

            This paper (… would suggest that up to 650k USD per day is being us for darknet markets, and that number has likely been growing. There’s a lot of articles on this, and probably your best gateway to them is Considering that $1million USD or so Bitcoin is issued per day (12.5*6.24*$600) that leaves about half or so (probably less) of the issued coins being used for non-speculative purposes. (Don’t forget that backpage/bovada/nitrogensports/etc are a big chunk too)What data do you study?RE: if you do not have that you probably should not be into bitcoin for more than beer money.I agree. I’ve been writing articles/code doing video and podcasting for years now, and have yet to take a dime. Since you haven’t heard of me, I’m guessing you shouldn’t be speculating either.

          10. ZekeV

            Trolling aside, your argument includes a logical fallacy. You assume that all dark market-related spends are coming from the pool of newly created coins. I would suggest a more meaningful figure to use as your denominator would be total transaction volume. Even then it’s hard to make a refutable claim bc many tx are not attributable. For example, if I am a big wallet-holder and like to tumble my coins around, you would have to engage in pretty sophisticated analysis to say whether that is economic activity (speculative or real) or not. tries to estimate this but is probably not very good at it.In my earlier comment, I was referring to USD exchange trade volume. I don’t give a #$%! about 650k dark market estimate, which seems low to me in any event. What I care about is the total tx volume in use in some sort of economic activity, which is much higher, and can only be roughly estimated using self-reported figures by exchanges (much of which is fraudulent).

          11. Chris DeRose

            Everyone knows I’m a troll. That’s what skepticism is, trolling.I don’t assume that dark market spends are coming from the pool of newly created coins. I assume that the capital inflows are at half of the issuance rate. And yes this is not a certainty, but this ballpark rate has been confirmed by a number of studies. I welcome any studies to the contrary. There are some, but not many.You certainly care about the dark market. That’s what blockchains do – they serve the underserved. When you invest in blockchain, that’s the economy you’re betting will grow.

          12. ZekeV

            OK, trying to be more precise then. What is the total velocity of bitcoin transactions in USD, after estimating out non-economic transfers. What is the rate of increase in velocity, again denominated in USD. How much of that increase is explained by purchases of goods and services (whether darkmarket or otherwise), and how much is an accumulated effect of net long positions motivated by desire for financial gain? I’m afraid I have not done my homework. I think those are good questions to ask, but I couldn’t answer them with any rigor. And really I don’t care as whatever the answer is, I’m more interested in what we can state about bitcoin in an exact way by looking at network stats, and not by putting figures in a USD context.

          13. Chris DeRose

            I’m not aware of this “total velocity of transactions” metric. Where did you get this? (Did you make that up? Did you mean to refer to monetary velocity of BTC?)

          14. ZekeV

            What other kind of velocity is there in this context? What I’m saying is that (a) your numbers are not plausible, and (b) the variable you’re estimating is not meaningful. I don’t claim any special status, I’m not a bitcoin “expert” for hire or s@#tcoin promoter, so this is merely of armchair interest to me. There’s really no point to continuing this conversation here. If I ever run into you in person I’d be happy to BS with a beer in hand.

          15. Chris DeRose


    2. jason wright

      Stem cells can be bio engineered to become any kind of cell.

  15. secretarycleary

    Healthcare – specifically health records – ripe for this sort of disruption.

  16. Frank W. Miller

    Your job is to make high risk bets I suppose. Smushing together the twitter sewer, your ongoing hope that blockchains will change whatever your latest pain point is, and a crazy idea about putting a leash on open protocol development has a very low probability of return. Just sayin.Here’s what you really need to do if you want “open” protocols. Take you money somewhere else and stay the hell away from the universities. When I was in my doctorate program, I wrote papers, had them peer reviewed, wrote implementations, distributed them as open source, etc. Then all you VCs appeared on the scene and started pumping money into it. Have you noticed how the rate of fundamental innovations has dropped over the last couple of decades in computers? Yes, the field has matured but now what happens is some freshman shows up at school X, says something that some VC (i.e. business trained vs engineering trained) likes. The VC or angel or whoever throws a few bucks at them and thats the end of it. Whatever potentially interesting basic tech work that was being grinds to a halt, or more likely gets locked up in some patent to “protect intellectual property.” Instead they’re spending their time worrying about, you know, monetization or some other equally non technically focused crap. Not exactly the kind of environment that fosters free thought resulting in open innovations is it?This dynamic was just starting in 1998 when I was coming out of school. I got recruited by several dot com startups at the time. Its basically de rigeur today. If you want to complain about why there isn’t any interesting fundamental tech coming out any more, look in the mirror. You’ve had a something to do with its demise.

    1. LE

      Instead they’re spending their time worrying about, you know, monetization or some other equally non technically focused crap. Not exactly the kind of environment that fosters free thought resulting in open innovations is it?Take you money somewhere else and stay the hell away from the universities. You’ve made some really good points. Unfortunately having a group of people with money stay away (as a group) from university would be as likely as telling a group of guys to stay away from the girls so they could “study and spend time on academics, not their relationships”.

      1. Frank W. Miller

        Oh I know, I’m just lamenting

    2. fredwilson

      you sound angry and bitter. that rarely results in a way forward. just sayin.

      1. Frank W. Miller

        Apparently you don’t like my post so you’re giving me you’re typical high road brushoff to try to marginalize what I’m saying. Too bad. I’m neither. I’m just a guy who knows something about your business from the entrepreneur side and likes to call you VCs out from time to time when you get too big for your britches.I see analogy on this with you and your brethren’s attitudes about housing in SF and silicon valley (and Boulder now too). You see, there are different ways to look at what you’re doing and in this case, you need to see your part in it and not try to minimize a different opinion.

    3. Pascal Aschwanden

      I think there’s more then enough researchers and tech people to follow up on interesting fundamental tech. Only a tiny percentage of potential products or ventures ever get funded. The ratio of talent to opportunities is extremely high.Now, the point about patents is valid – quite an egregious situation, but that’s a different matter.

  17. kevinmarks

    A protocol that is a monoculture (single implementation) or a silo (requires permission from one organisation) isn’t a fully open one. Multiple interoperable implementations matter.

  18. sigmaalgebra

    > crypto-tokensCute. Cute application of the blockchain ideas.USV, Albert Wenger, William Mougayar seem to be getting to be a leader in blockchain ideas.Ah, Twitter, Tweetie, Twokens — reminds me of Twinkies!> brave new worldAh, some Aldous Huxley!Huxley was writing fiction. Hopefully block chain applications can be analyzed more rationally. I believe that business model innovation is more disruptive that technological innovation. But crypto-tokens sounds to me more like “technological innovation” than “business model innovation”?Gee, used to be the best we had was 300 baud public switched telephone network (PTSN) point to point dial-up, and we’ve had some disruption since then! Now we have TCP/IP and the commercial internet, and that was “business model innovation” or “technological innovation”?Similarly for HTTP and HTML?The optical fibers and solid state lasers crucial for the backbone of the commercial internet?The chips for the high end routers for the backbone of the Internet?Border gateway protocol (BGP) for the core of the Internet.Cable modems with 100 Mbps data rates over coaxial cable in millions of homes?Giant magneto resistive hard disk heads to permit 10 TB (trillion bytes) in a 3.5″ form factor package?Solid state disks — now 14 TB — crucial for mobile devices and also with some astounding performance possibilities?The Intel 14 nm microelectronics manufacturing process for 2-24 core processors as in…

    1. James Ferguson @kWIQly

      Not so much Twoken implies Cryptoas Twoken imples Kripkehttp://bigbangtheory.wikia….

    2. Chris DeRose

      William Mougayar has no idea what he’s talking about. He’s the deepak chopra of blockchain.

      1. fredwilson

        you can disagree with someone without attacking them and calling them names. that’s how most people do it around here.

        1. Chris DeRose

          It’s a fitting characterization. He throws out word salad without any technical comprehension of the science. You should give bitcoin uncensored a try, it’ll probably save you some money there at USV…

      2. sigmaalgebra

        Naw! For that, William would have to be writing long, complicated posts about the ancient, Eastern wisdom, inscrutable, miracle, health giving, healing, and mind expanding properties of Quantum Blockchain while listening to brass bells ringing from swinging in the wind, in a hut of bamboo on tall legs with a thatched roof, wearing only a homespun loin cloth, sitting in the lotus position on a woven, bamboo mat, smelling burning incense, and singing Hooooommmmmm or some such.That’s not the way I see Canada and William. Besides, in Canada, usually it’s too cold for that.

  19. jason wright

    Is USV ‘donating’ in the ICO market?

    1. fredwilson

      are you talking about Open Ledger?

      1. jason wright

        He’s a pioneer, but his plans can be hard to keep up with.icoo.ioI was thinking more about the general flow of ICO announcements on BCT and whether USV participates – Albert said he put his own 0.5 bitcoin in The DOA. Or, is USV looking to take equity in the companies (but many are forming as Swiss foundations and some as not for profit gGmbH social enterprises) in the traditional VC way.

  20. Josh Rehman

    Hey Fred, I’m confused on one point: how would Twokens be spent? You talk about how they are earned, but you didn’t talk about why anyone would want them. Thanks.

    1. Tom Ding

      The typical model is that you create an on chain biz protocol (e.g. decentralized Uber, Twitter, etc.) that also contains a ledger of “voting token holders”. Whoever holding the voting tokens are entitled to vote on policy and upgrade decisions for the protocol, e.g. do anyone get paid for tweeting, upgrade to v1.1, add a new moderator for certain section, etc.In the case of a ridesharing protocol, easy to imagine that riders pay a fee for each ride, Uber today charges 25% commission, let’s say our ride protocol is less greedy and just charge 1%. That 1% goes to the “protocol” (i.e smart contract) that further rewards all of it voting token holders.Typically, all voting tokens are transferrable/exchangeable, which creates a liquid market, hence the incentive for early purchasers of voting tokens to buy early on during the protocol’s public token distribution event.P.S. In the case of twitter-like protocol, that commission fee could easily be advertising revenues that the protocol collect for sponsored tweets, and then distribute to voting token holders. In practice, there’re other ways to reward, but this is the simplest conceptual model.More thoughts here:

    2. fredwilson

      i mentioned that they would be traded on an exchange

      1. Josh Rehman

        It’s true that you said “Crypto-exchanges could make a market in Twokens so that anyone who wanted to speculate on the future value of the Twitter protocol could do so.”I’m still unclear about this. Again, why would you buy Twokens if you can’t do anything with them? With stocks, you can get dividends and vote them. If you amass a majority, then you can control the company. But that is not true for Twokens is it?A simple example would really help me. For example: I earn 1 Twoken for every follower I get on Twitter. I can then spend 1 Twoken to “push” one of my tweets to an unrelated Twitter user. (Or, more likely, some multiple like 10 or 100 users).

      2. JordanCooper_NYC

        i had the same question. i get they can be traded on an exchange, but i’m not sure how the value of a twoken is tied to the value/use of the protocol? if twokens aren’t an input to use of the protocol, they aren’t really tied to twitter right?

        1. kidmercury

          eventually someone will come along and deliver the whole solution at once: issuance of tokens and a network of storefronts where they can be redeemed. but this type of end to end solution requires greater centralization that is, at least for the time being, antithetical to the purported spirit of bitcoin et al.

      3. Josh Rehman

        Sorry if I’m being slow, but I still don’t understand. I understand there is a market, and I understand (I think) the *source* of twokens, but I don’t understand the *sink* of twokens. Do you just have them to trade? Like beanie babies? Or can you do something with them? E.g. what is the scenario where someone is like “OMG I *need* 15 million twokens for my evil plan to work, muhahaha!”?

  21. jason wright

    Buy the new venue

  22. jason wright

    I’ve got +9000 disqus votes. When can i exchange them for dokens?

    1. James Ferguson @kWIQly

      Now +8999 – When you find out please remit me the difference :)Upvoted your next comment just to show Im kidding 🙂

    2. @mikeriddell62

      When a disqus vote is worth something that people are willing to accept in payment. Till then, it’s just another worthless point.This whole conversation is really about currency. Currency that is earned into existence for generating societal and/or business value will be the dwoken/twoken of the future.Particularly if it is designed to pay down fiat debt.Debt is the problem. Community currency the solution. Innovative business model = a co-operative. Hard to capitalise. Easier to resource. People power is coming to Manchester.

    3. Alex_Hirner

      You can monetize indirectly by getting paid for influencer marketing. I think it doesn’t count much compared to other clout measures, but better than nothing.

  23. Jack Gavigan

    Isn’t the crypto-token model Albert describes pretty much exactly what the Ethereum Foundation did? Except instead of an open standard, what they’re monetizing is a distributed, decentralised, computing platform. And instead of for profit, it’s a non-profit foundation.

    1. fredwilson

      yes, but Albert is suggesting that it does not need to be non-profit

      1. jason wright

        Lisk is going gGmbH in Berlin for tax efficiency, and I think to promote a sense of stability in a very frontier market.

  24. chris dixon

    Great post. Maybe you’ve already made this point, but in addition to providing a monetization model. an additional benefit of (public) blockchain architecture is it allows protocols to be stateful. If you think back to the 2000s battle of open vs close social (RSS vs FB, Twitter etc), the open side was at a severe disadvantage because it had to try to build the name space, social graph etc via user websites (effectively trying to overload DNS–since it was the only pre-blockchain public database). Today, open source projects and open protocols can use Bitcoin, Etherereum etc to create persistent databases.

    1. David Semeria

      Good point re blockchain’s ability to store state (although capacity and throughput remain issues). However, I’m still to be convinced on how blockchain changes the game wrt to protocol monetization.Protocols are abstract notions – they can be specified via a dedicated and articulated website at one extreme, or on the back of a napkin at the other.Since digital services are always backed by code, monetization enforcement is straightforward. This cannot be true in the abstract world of protocols.

    2. fredwilson

      i did not, but that’s how Albert kicked off his post that i linked to and inspired my post. that’s a big deal

  25. Ronnie Rendel

    The concept of cryptocurrency is one step closer to my vision of ultimately our association with networks will supersede our association with nation states.

  26. Sebastian Aldasoro

    Such an interesting topic. In Argentina there’s social network called Taringa! with about 28 M registered users. Through the years they have grown consistently thanks to a vibrating community which generates content and traffic to the site. They were thinking of a way of paying some money to top members of the community by calculating how much revenue they were generating for Taringa! After lots of thinking they did an alliance with Xapo Wallet to give out wallets and start doing micro-payments. Even though it’s not protocol based business, I find it interesting to see how they were kinda going that way intuitively and, off course, with Bitcoin.

  27. static

    Albert’s post is pretty wrong. We don’t need the blockchain to bill for API calls- in fact it is massively more complicated than something like the Google Maps API token based billing scheme or the Amazon Web Services billing. The simpler solution for APIs is something like monthly billing for API calls used, with a freemium level to get started. What Albert is talking about is more like a secondary market for unused Google Maps API request tokens…Open protocols have absolutely nothing to do with this. Even authenticated endpoints have open protocols. Bitcoin itself is solving a different problem, low transaction cost micropayments. Focusing on models where people can pay very small amounts to consume without establishing a billing relationship with the provider or incurring transaction costs is still the real opportunity.

  28. Jeff Lunsford

    Companies with portfolios of assets tend to trade at a discount to net asset value – between 0.6 and 1.1x NAV. (see GSVC and closed end funds) Smart entrepreneurs will figure out a way to launch an open protocol and figure out a way to preserve a position in the emergent ecosystem where they can build a recurring revenue stream. Those trade at 3x-10X revenue, depending on growth rate, gross margins, etc. The finesse will be carving out that preferred position in an open protocol world. Many companies have successfully done this with open source software, a decent but not exact analog.

  29. Kevin Hill

    I hate to keep being a wet blanket…I do love the idea, but 8 minute latency kinda destroys the value in modern protocols. And that latency is baked right into the fundamental structure of the blockchain. EDIT: you might be able to get that latency down to 2-3 minutes, but not 200-300 ms. Doing so would strongly compromise the security of bitcoin.I would love it if someone would invest heavily in a system that could enable this type of service. Unfortunately, you guys aren’t and pretending that you are would seem to do more harm than good.

  30. sigmaalgebra

    Wait! Stop the presses — or Web servers as the case may be.This just in!On disruption via technology or business model, at…is in part:Scott Sheffield, the outgoing chief of Pioneer Natural Resources, threw down the gauntlet last week – with some poetic licence – claiming that his pre-tax production costs in the Permian Basin of West Texas have fallen to $2.25 a barrel.”Definitely we can compete with anything that Saudi Arabia has. We have the best rock,” he said. Revolutionary improvements in drilling technology and data analytics that have changed the cost calculus faster than almost anybody thought possible.As inEnders A. Robinson, Multichannel Time Series Analysis with Digital Computer Programs.from when the fast Fourier transform (FFT) was new, likely part of the “data “analytics” is more from the FFT!Ballpark ItLet’s round that up to $3 a barrel and, from…see that in 2015 the US consumed 19.4 million barrels per day.So, with $3 a barrel instead of, ballpark, $100, we’re saving $97 a barrel. So, the bottom line savings are, and may I have the envelope, please [drum roll]97 * 19.4 * 10**6 * 365 = 686,857,000,000dollars per year. [Simple arithmetic courtesy of my favorite text editor KEdit and a simple macro.]Does saving$686,857,000,000a year qualify as “disruptive”?Who is Fourier?The FFT? There was a good applied mathematician, J. Tukey, e.g., as at…and as inR. B. Blackman and J. W. Tukey, The Measurement of Power Spectra: From the Point of View of Communications Engineering.Exploratory Data AnalysisTukey’s lemma (a version of the axiom of choice)Convergence and Uniformity in Topology and much more, long at Bell Labs and Princeton. Long at IBM Research was physicist R. Garwin.Ah, once the power spectra book, that I read while eating broiled flounder, French fries, coleslaw, and little glasses of beer for dinner at a Silver Spring seafood bar, and some software I wrote in a hurry, let me get the company I worked for sole source on what had been a competitive bid!At a US Presidential Science Advisors meeting, Garwin, who had been struggling with long computer time for Fourier transforms, happened to be sitting next to Tukey who was taking meeting notes with one hand and doing Fourier derivations with the other. Garwin asked Tukey if there was something he (Tukey) knew about Fourier calculations that he, Garwin, didn’t?Yup. For n data points, Garwin was spending computer time proportional to n^2. Tukey got that down to proportional to n*log(n). So, ballpark for 10,000 points, we’re talking faster by10,000**2 / ( 10,000 * 4 )= 2500How ’bout that! Am I getting a whiff of disruption?Ah, once there was a guy who wired up a hardware box for the FFT. The US Navy paid for that — right, somehow the US Navy really liked the FFT (extra credit for guessing why!). Well, the FFT is done in stages. So, he had separate hardware logic for each stage. As he was doing final tests of his box, we talked, and I explained that there are several versions of the FFT including one that has the same logic at each stage. So, knowing that, he could have saved about 10 blocks of hardware for such logic! Poor guy! He hadn’t spent enough really fun evenings reading FFT literature!Fourier? Well, remember first vector analysis where have a vector and an orthogonal coordinate system. So, to get the coordinates of the vector, just project the vector on each of the axes.Well, that generalizes, a lot, first to countably infinitely many dimensions and also to uncountably infinitely many dimensions.The orthogonal axes are from just the high school sine and cosine functions. The projection is a generalization of the usual vector analysis inner product.The countable case is for Fourier series, and the uncountable, Fourier integral.Yup, with all this orthogonality, have to smell that there should be a version of the Pythagorean theorem, and, yup, there is — Plancherel’s theorem!W. Rudin is big in Fourier theory. Can get the countable case nicely in his Principles of Mathematical Analysis, a.k.a., Baby Rudin, and the uncountable case in his Real and Complex Analysis.There is more on Fourier theory in applications in physics, especially quantum mechanics, astronomy, holography, molecular spectroscopy, etc.If pass a signal through a linear system, then all it does is tweak the values on the coordinate axes. Well, passing a sonar signal through rock with oil shale is close to such a linear system.Now, we are on the way to saving$686,857,000,000a year! Thank you Fourier, Rudin, Tukey, etc.That should help get the US economy going!Is the following a picture of the person you want in charge of US energy policy?

  31. dpgj

    I really doubt Blockchain will be the future. The technology cannot scale up to enormous scale because it is inherently inefficient. It relies on the prohibitive processing power to ensure security. However with wider deployment and more processing power dedicated to the computation, there is no guarantee the transaction cannot be faked or tempered. Even not, prccessing huge amount of transaction will use up ridiculous amount of processing power.

  32. p-air

    Fred, given your thinking here, would you see USV start to allocate a part of its investments towards ICOs or actual token purchases, or would the firm stay true to its model of investing in the companies themselves (I’m treating equity investment as mutually exclusive of token purchases)?

  33. andrewparker

    This post outlines a framework for protocols to generate money as they become adopted. Whether for-profit or non-profit, it would allow a protocol to pay it’s own bills (and possibly some extra beyond the bills).While that’s a super interesting thought experiment, I still admire the completely decentralized protocols we depend upon every day (TCP/IP, HTTP, DNS, Ethernet, etc) that generate no direct monetary benefit to those who create and maintain them and yet have created trillions of dollars of economic opportunity.So, it begs the question for me, why does a protocol need to generate money? If the answer is: because a for-profit business created it, I find that underwhelming. Instead, just open source the protocol and let it blossom (or fail) unencumbered by the commercial burden. If the answer is: because there’s a material overhead to operating the protocol, then I think there’s an opportunity for decentralization as opposed to commercialization. In a parallel universe Twitter could have allowed anyone to host and serve tweets, decentralized, and in doing so, removed all the economic burden of hosting the API. I’m not saying Twitter should have gone down this path, I’m just showing a case for how a protocolized version of Twitter could have emerged without the need for a crypto token system to pay the bills.

  34. Adrian Hope-Bailie

    Lot of confusion in this and Albert’s post about the difference between a platform and a protocol. A token is useless in a protocol because a protocol itself has no state, it is a set of rules. The platform that implements that protocol has state and can therefor have tokens which can be traded and ultimately may even have value.The Twitter analogy is just plain wrong I’m afraid. Twitter is a platform. If you want to call its APIs a protocol that’s a stretch, but let’s go with it. Then that would imply other Twitter clones could host the same APIs, and software that is written for Twitter would work with the clones and you’d have a whole bunch of interoperable micro-blogging (or whatever you want to call them) services based on an open protocol.This is pretty much the opposite of what you have described.I think your point is a good one, that decentralized and open networks with a native token of value have created an environment where anyone can create a platform and monetize it by simply selling the native platform token. Sales of the token should be driven by the perceived utility of the platform so it’s a great way to fund yourself if you can deliver a great platform.But, it’s a platform, not a protocol.

  35. Tom Ding

    Good thoughts quite aligned with our own experiences (as an open protocol labs). Adding some perspectives:1) Advantages beyond just open participation / permission-less innovation, biz protocols (on-chain) are naturally “ex juris” – so they’re global business by default, rather than typically country-based silo’ed FinTech company (many due to regulations / banking system). Instead of trying to reconcile 196 diff jurisdictions’s financial laws, there cud be unified open protocol living on chain, while on the edge local entry/exit points deal with regulations.2) Further, public token distribution (more commonly “crowdsale”) is a potent method to solve the typical startup First 1000 user bootstrap problem. So instead of use your VC money which come from diluted equity to acquire user, you in opposite raise crowdfunding and acquire user at the same time and from the same group, who become your best evangelists. (note: this is diff from kickstarter, as giving out tokens ties your investor/user’s incentive with the protocol).3) Now – to reduce the trust issue of funding with early on crowd-sale, there’re two possible good models: a) manage the funding/dev using software foundation rather than for-profit entity b) custodian the funding completely on-chain based on governance/ pay-out rules (e.g. 5 head counts’s dev salary each month)More thoughts here if interested:

    1. ZekeV

      Presales of vaporware tokens should be disfavored. Heavily pre-selling and pre-mining a protocol makes for a strong incentive to fork it, if it appears to be popular early on. There is also the problem that presales of tokens that promise some financial return (e.g., DAO contracts) may be illegal securities offerings.

  36. Ryan Walker

    The potential of this is exciting but so far most evidence points towards these applications/organizations leveraging more secure/established blockchains rather than incurring the cost/risk of launching a new one. LTBCoin is one such example and probably the longest running experiment around this. A lot of excitement being generated by Steemit but the YoursNetwork will likely prove a much more sensible model as it will leverage an existing token of value and utility (bitcoin).I’ve gone back and forth since getting into bitcoin/blockchains 3 years ago as to whether there will be a few major blockchains or thousands/millions of small ones. The more experiments I see, the most lasting, secure, and successful all seem to be pegged to the most secure blockchain. Next 2 years will likely convince me even more as more crowd funds and platforms launch built upon the bitcoin protocol. We’ve already seen a lot of this with Counterparty but will see even more as Rootstock, YoursNetwork, OpenBazaar, continue to grow while not incurring the redundant risk and cost of establishing their own blockchains.Crowdfunds and speculation will probably distract us from this for a while until people realize that legitimate businesses that need security (have value) will being with the most secure blockchain around.The one thing that would make me change my mind on this would be if future proof of stake experiments (Ethereum for example) are successful and render proof of work irrelevant. Viable POS would likely make the establishment of secure distributed blockchains easier.