USV Thesis 3.0

At USV, we are thesis-driven investors. We state clearly and concisely what we want to invest in and, implicitly, decide not to invest in anything else. That is a good thing because, as our partner Rebecca wrote in this post:

The commitment to a thesis is part of the fiber of USV–a shared set of ideas creates a framework that allows us to operate with focus and work on what matters most

Today we are publishing, and committing ourselves to, version 3.0 of our thesis:

USV backs trusted brands that broaden access to knowledge, capital, and well-being by leveraging networks, platforms, and protocols.

But this is actually the fourth USV thesis.

When we started USV, Brad and I committed this to our investors (before this was tweetable):

At USV, we will invest in the applications layer of the Internet.

That was an acknowledgment that, in 2003, we had moved beyond what Carlota Perez calls the installation phase and had reached the deployment phase.

But we quickly realized that the applications layer thesis was too big and we needed to narrow our focus. So in 2006/2007, we started focusing exclusively on networks that were quickly amassing large users bases on the web. That was eventually articulated by our partner Brad in this tweet:

By 2013, we knew that the network thesis was long in the tooth and had started adjusting to that fact. We had gotten interested in the blockchain sector in 2011 and had also started building large portfolios in fintech, education, and health care. Our partner Andy finally forced the issue and we went public with what he called “Thesis 2.0” but was actually the third revision.

And now we find ourselves again iterating slightly on what we are focused on. And so we turned to the newest member of our partnership, Rebecca, to explain it to the world.

One of the things I like most about Thesis 3.0 is that is a modest adjustment to where we have been with our prior efforts, but we have tightened things up and added a few nuances. Like the prior iterations of our thesis, we have already been investing behind this thesis and you can see elements of it here and elsewhere in our recent investing activity. Notably, the words blockchain and crypto are nowhere to be found in thesis 3.0 but it leads us inexorably there.

#VC & Technology

Comments (Archived):

  1. Twain Twain

    Where is a “trusted brand”? Seed? Series A? Series B? Thanks!

    1. jason wright

      I’m thinking exactly the same thing. A “trusted brand” can be a young company, but probably only because it has not yet had enough time to behave badly and abuse the assumption of trust. One would like to assume trust from the earliest of times, but experience of web companies suggests that may be a little bit naive.

    2. fredwilson

      There is an aspirational element to this version that I like very much

      1. Jordan Jackson

        Working on Brand thus far in my career, it has been abundantly evident that the brand in large part (especially at early stage) is synonymous with the uncompromising conviction the founder has in her vision of the future. Which, bleeds into the early internal culture and org structure / principles. I am unclear what this means with “trusted” – implicitly I would imagine USV has always looked for this. Or maybe not – Airbnb is the quintessential “brand”.

        1. Jordan Jackson

          I don’t bring up Airbnb in a negative way at all.. simply to illuminate my confusion about the new distinction of looking for brand as a signal of a USV investment. Again, it seems like this is something that has been done implicitly already. I guess my question is – what does this mean explicitly.

          1. Matt A. Myers

            I don’t trust Airbnb – I have had bad experiences as guest and host, and overall their behaviour leads to capturing as much value/revenue for themselves as possible, the traditional VC way.Trust is generally an afterthought for companies/organizations. Trust is a very deep, foundational institution – like justice. Unfortunately society has been essentially forced because of limited options and how economics has driven us, to “trust” blindly – or without any thought at all; related to this comment of mine – https://avc.com/2018/04/usv

      2. Mike Porath

        “Trusted brands” is my favorite part. Some companies have it in their DNA. Some don’t. I think you can tell at an early stage.

        1. karen_e

          Trusted brand, trusted brand. Maybe if I say it ten times it won’t sound like P&G or Target or Marriott.

      3. Girish Mehta

        Understand you meant it aspirationally.The problem in the sentence construction is that you put trusted brand at the start of the sentence (maybe to signify its importance). That sentence gets broken down as: USV backs trusted brands. The trusted brands broaden access to knowledge, capital….Brands get built over years, trusted brands over longer periods of time.Alternatively, you could have the sentence flow as follows:USV backs enterpreneurs/companies that:- Broaden Access to Knowledge, Capital and Well-being- By Leveraging Networks, Platforms and Protocols- Based upon a foundation of (put your aspirational language about trust here).A Brand is not what a company says it is. A Brand is what the customer says it is. A trusted brand is an outcome. It can be aspirational, but not self-defining.p.s. Not commenting on the thesis. Just language to your point about the aspirational element.

        1. PhilipSugar

          See my other comment. Personally I think, and this is just an opinion (which we know like an a-hole everyone has and they always stink) and this is Fred’s blog, and USV is his company. It would read better if said:We invest in companies that build networks and monetize those networks to the benefit of the members of that network in a way that the member understands the relationship and is rewarded for their benefit to the network.

          1. fredwilson

            Too long. I stopped reading about half way through

          2. PhilipSugar

            Point taken. I like brevity. I do not believe you can get more than five points across. Do not however believe in Twitter 140 Characters (I guess neither do they anymore)How about: Build networks for benefit of company and members.

        2. fredwilson

          That’s how it was originally constructedI pushed for the new construction

        3. sigmaalgebra

          English is one heck of a versatile, possibly ambiguous, difficult, natural language.

          1. Twain Twain

            This is why Natural Language Understanding by the machines is unsolved.An NLP startup tweeted this about me yesterday. Word2Vec is the building block of Google’s search engine, its Knowledge Graph and its NLP framework. https://uploads.disquscdn.c…It is why …Alphabet’s Eric Schmidt: It can be ‘very difficult’ for Google’s search algorithm to understand truth* https://www.cnbc.com/2017/1…Google’s Fighting Hate And Trolls With A Dangerously Mindless AI* https://www.fastcompany.com

      4. Twain Twain

        USV backs brands that broaden access to knowledge, capital and well-being via networks, platforms and protocols, and are trusted by users.That’s 138 characters.

        1. Matt A. Myers

          The interesting part about “trusted by users” — is that blind trust, as in no critical thinking has gone into the decision to become a user, etc? Ignorance is bliss?For USV’s thesis to mean anything, they really need to solidly define what trust means to them, in depth – and to have open discussion about what they write – to allow others, if anyone decides, to challenge their view or wording of trust … so then those who may challenge, the followers that trust them, they can get a sense for how much they should pass that trust on. One of the systems I have envisioned with my life’s work would allow for this.

          1. Twain Twain

            Girish is right. Trust is CONFERRED by users. At the moment, we don’t know WHY they trust a system. We can only measure #signups, attrition rates, clicks etc.There is the thesis that blockchain is a “trust” network, isn’t there.Whether this plays out to be so or whether it will be another 2008 financial collapse or 2018 social media privacy, fake news, trolling etc scenario … who knows?

          2. Matt A. Myers

            “Trust” network being misappropriated to “the next best thing” is another example of currency and stock market related terms being misappropriated; crypto-“currency” etc.I’ve personally calmed down since realizing the market should simply outcompete the increased costs that incentivized blockchain will lead to, so long as there is good governance and chain of command/chain of trust — people like Elon Musk steering the ship, with empathy and holistic thinking – how does this affect everyone, etc. The concern is that a tipping point will be reached where enough “investors” / enough of the masses put their money into any of the incentivized blockchains, and will want them to rebound, and collectively push for them to be adopted by the rest of society — meanwhile unreasonably, unnecessarily, reallocated wealth to themselves … and as always, weighted towards the earlier adopters.

      5. ShanaC

        that’s a healthy approach

    3. PhilipSugar

      I think the real question is what and how do you make your business. I think that many businesses have been built on the fact that “there is no privacy” and have just harvested data, you the consumer are the product. For the vast majority of people, they are considered to “stupid” to know this.That I think and have said will change. You can see this in USV investments and portfolio companies. Do you want your ISP selling your browsing habits?I loved yesterday’s post.

      1. Twain Twain

        There are lots of pre-revenue startups that are reliant on subscription / ads / credit cards of the founders.Cloud flare is lucky and has USV’s pocketbook to back them.Nine years ago, my view was that everyone should have a P-Pod (Privacy Pod) the size of a button. They could record whatever information they wanted on it.Brands would pay THEM every time they wanted to access and use that information.But Open Data has already happened. Twitter is open. Github is open. A lot of Google’s tools are open.

        1. sigmaalgebra

          On Twitter, I “follow” just three people; I don’t post very often; and I sign in with a pseudonym. I’ve never yet gotten anything of value from Github, and I’ve certainly never contributed anything. I use Google ONLY for search and YouTube and will NEVER use their tools for my content and certainly won’t let them have my real name. I have a copy of the Chrome Web browser, but I have JavaScript and third party cookies disabled. On Firefox, third party cookies are also disabled.Those precautions don’t make me really, fully anonymous, but they are a good start.

          1. Vasudev Ram

            >but I have JavaScript and third party cookies disabled.So you do not use any sites where you have to log in?AFAIK, to do that, and also, of course for greater interactivity and AJAX, at least, JS has to be enabled. I’ve been seeing some people on HN say that they disable JS at least selectively and it speeds up the browsing experience a lot, which I can imagine would happen, what with today’s JS- (and image-)heavy pages.. Maybe they use a different browser or browser session for sites for which they need to log in.Going to try that out.

        2. PhilipSugar

          That is why my company was called Smart Button. You are young, the pendulum swings.At first computing was central: Main FrameThen went distributed: Client ServerThen went central: Internet CloudAnd might be going distributed: Block ChainSame for privacy. Your Ying/Yang Avatar is a very apt.

          1. Lawrence Brass

            On a rainy day at the bar, at a table by a window, I would like to hear the story of Smart Button from you. 🙂

          2. PhilipSugar

            Oh, that is not my most interesting story. I keep a young looking Avatar. My hair is as gray as yours… 🙂 If I remember right you are from Chile. I love that country. Yes, sometime we will. I actually really want to visit Uruguay. I find it interesting that all capitals of Chile Argentina and Uruguay are in a line West to East.

          3. Vasudev Ram

            I had read that they eat a lot of meat in Uruguay. I found that interesting for some reason, not sure why. Argentina and beef I already knew about.I had checked out Uruguay when I came across this Uruguayan company called Insophia. It was (at least partly) there that the Python web scraping framework Scrapy was created, and it later led to the formation of the company Scrapinghub, which seems to be doing well in the business of web scraping.

          4. PhilipSugar

            Yes, I love red meat. We have a potluck every quarter, not organized by me. Next one is April 10th. I will cook a beef clod in the parking lot. 10 killos of beef shoulder. https://www.youtube.com/wat…Now in no way is this disrespectful of employees that do not eat meat. We celebrate. I am from Texas. Those employees that do not certainly do not eat, but we celebrate our differences.

          5. Vasudev Ram

            Got it. Interesting.

          6. Vasudev Ram

            Just saw that video. Interesting that it takes 20+ hours to cook it. I suppose due to low heat and a thick piece. Been experimenting with low heat cooking myself, after learning that it helps the food not to burn on the outside while being undercooked inside.

          7. Lawrence Brass

            Toss some oyster mushrooms over the grill and you can please the most stubborn vegan. If they mind the greasy fumes you may put them on a frying pan over the grill with some olive oil. Incredible texture, color and flavor.I know this may upset some Texans, but I can say I enjoy grilled or fried oyster mushrooms as much as a good beef.

          8. Lawrence Brass

            Yes, its interesting. And if you extend that line to circle the world you get near Camberra in Australia, and Capetown in South Africa, other capitals at similar latitudes.Haven’t visited Uruguay living so close. I love Buenos Aires and of course Santiago, my town. If you come down south we could organize a meet somewhere under the light of the Southern Cross and some good beef and red wine.I am planning a US trip probably for late summer or autumn this year, we could meet then too. Will try to meet JLM in Savannah if he is there by the time. Just plans.

          9. Vasudev Ram

            That almost sounds like a haiku, Lawrence.Good one.Here is a poem I like, that I first read as a kid, in a kid’s book [1] – it took me a while to get it, at the time 🙂 :Into the drinkjng wellWhich the plumber built herAunt Eliza fellWe must buy a filter.[1] The book was The Golden Book of Fun and Nonsense.Some of those kid’s poems and stories can be ghoulish.IIRC there was a study done about fairy tales, with that as one of its findings, and of course then some conclusions (right or wrong) drawn from that.

          10. Lawrence Brass

            Ha ha, nice. My dad used to like that type of poetry or songs, things they sang during the war I think.I recall my first experiences chatting in the web were in HTML “chats rooms”. In the particular one I was hooked into everyone played a character. I knew very nice people from different places who enjoyed this type of interaction. It was ludic and fun. I don’t really understand what was the motivation for this, in part anonymity, in part play. It thrilled me.

    4. PhilipSugar

      You know I had to look up legal definition of “trust”: “confidence placed in a person by making that person the nominal owner of property to be held or used for the benefit of one or more others”

      1. Twain Twain

        Wait until you have to look the legal definition of “fairness” and discover it doesn’t accord with the mathematical definition, the layman’s definition and the economics definition.And THAT makes it challenging to work out universal standards for moral, fair and ethical AI.

        1. sigmaalgebra

          As long as we have no AI, we don’t have to worry about moral, fair, or ethical AI.First get some AI, and then think about moral, ….We’re reaching what looks like AI peak hype!! An AI fall of failure and then another AI winter are on the way!!

          1. Twain Twain

            Actually a lot of people are looking into the moral, fair and ethical issues of existing AI, including Joi Ito of MIT MediaLab.@fredwilson:disqus — Joi writes: “We also need humans in the loop to develop the metrics that will fairly assess the costs and benefits of new technology … Already, technology and automation are reinforcing and exacerbating social injustice in the name of accuracy, speed, and economic progress.”@girishmehta:disqus is right: “A Brand is not what a company says it is. A Brand is what the customer says it is. A trusted brand is an outcome. It can be aspirational, but not self-defining.”I branded my system with the word “US” in it for a reason. It’s the crowd that makes a system trustworthy. It’s not network effects, per se. Moreover, the current mathematical structure of networks means that we reinforce and exacerbate the speed of echo chambers.There is a better way to do networks, platforms and protocols.https://www.aclu.org/issues

    5. Matt A. Myers

      And is it trust based on what’s right, or trust based or hinged on loyalty?

  2. Olivier Issaly

    “Like the prior iterations of our thesis, we have already been investing behind this thesis”So it confirms portfolio makes the thesis, not the reverse, as we can often see vc firms with mismatched or inconsistent thesis/portfolio ?

    1. fredwilson

      It’s like chicken and egg. What comes first?

      1. Lawrence Brass

        a trusted egg of course… or is it a trusted chicken?

        1. Matt A. Myers

          And what are you trusting the egg or chicken with? Of course clearly not investing if you don’t align in how one trusts … are you open to investing in/trusting someone who’s view may cause fear/discomfort/turn-off because it’s contrary to a foundation you’ve set yourself to stand on, a foundation you’ve likely set yourself up to need to trust?

          1. Lawrence Brass

            For me trust is half track record and half gut feeling on a case by case basis, so I wouldn’t say I have a foundation. I guess we all need some common point where to start from, some bonding to build trust from there.I tend to trust people more than concepts.

        2. Vasudev Ram

          One thing we can trust in is that both of them will be eaten.

          1. Lawrence Brass

            Eaten in what order? And we are again on the start. 🙂

          2. Vasudev Ram

            Clever 🙂

  3. sdso234

    Serious question: why do you need the second half of the thesis? Doesn’t every business leverage networks, platforms and protocols? Second serious question: can a pre-revenue startup be a trusted brand?

    1. sigmaalgebra

      Ah, maybe substitute “trustworthy” for “trusted”??Besides, as some might conclude from Facebook, big does not always mean “trusted’. Maybe with “trusted” USV is trying to avoid, preempt, head off some crypto currency fraud or the current mess of Facebook, avoid a long walk on the short pier that got Facebook all wet. Maybe USV is trying to avoid a self-driving car that runs over people, a news aggregator that has a political agenda and pumps out fake news, etc. Or maybe should say that sleazy business is bad business?

      1. Pointsandfigures

        Might parse it differently given prior comments. What is crypto? Part of it is “trust network”-which actually might be the most powerful part of crypto if you think about it broadly. Imagine being able to do business and trust someone implicitly that is part of the network and would sustain reputational cost if they broke the trust somehow.

    2. fredwilson

      You may be right about the first question. We debated that a lotThe answer to the second question is that we are intentionally aspirational in the use of that phrasing

      1. sdso234

        Makes sense. At the end of the day, I am not an important audience for your thesis.From the outside, though, your thesis makes it hard for me understand what you would NOT invest in. By contrast, your earlier thesis put an unmissable stake in the ground about the future. It was great precisely because it could easily have been wrong.It seems that “trusted brands” does a lot of work here. It seems to encompass “relationships of trust”: how do companies/tech establish trust for transactions, certifications, accuracy/quality of information?In the past, we trusted (and therefore relied on) institutions like banks, media companies, etc. because they were big, and had a lot to lose. But networks, platforms and protocols play a huge role in enabling the conditions for trust to occur, even in the absence of big institutions or “big brands”.

  4. JimHirshfield

    “Like the prior iterations of our thesis, we have already been investing behind this thesis”Tail wagging the dog?Playing devil’s advocate here, ’cause I don’t have a horse in this race…If you change your thesis more frequently than your socks, is it still a thesis? LOLThat said, I like this thesis.

    1. jason wright

      bigger feet, bigger socks, bigger shoes.This new thesis feels a bit fat.

      1. JimHirshfield

        You pullin’ my leg?

        1. jason wright

          with my bigger gloves. this thesis is getting a bit out of hand. I can imagine Fred’s email inbox now becoming even more overwhelmed once the world has read this post.

    2. fredwilson

      Thanks JimI think evolve is a better word than change for what we are trying to do

    3. PhilipSugar

      I hope Fred changes socks more than every four years. I have had a child that doesn’t for four days, and I can assure you you can tell because they stink.

  5. awaldstein

    Thanks for sharing Fred.Visions iterate little and are possible with a group of a couple of people. Brands are built at the intersection of communities and markets.Curious how this plays to your statement.

    1. sigmaalgebra

      For brands Chevy, Toyota, Amazon, Western Digital, NAPA, State Farm, Apple, Levi’s, L. L. Bean’s, Casio, Holiday Inn, FedEx, and Ford, we can see the “markets” but what the heck are the corresponding “communities”???? E.g., each of these brands sells strongly into markets all around the US or in some cases, e.g., Toyota, Casio, and Apple, all around the world.

  6. andyswan

    Feels like an over-fit to me…. does the portfolio make the thesis or vice versa?Does “trusted brand” mean that companies without major scale need not apply? Wouldn’t this have weeded out Twitter in 2007?I do like “knowledge, capital and well-being” on top of the networks and platforms layer…. although “well-being” seems oddly if not intentionally broad… a wild card for loophole opportunities?In any event, awesome that you share your thinking and process. WTF do I know I’m just chucking peanuts from the cheap seats. Always love watching USV succeed*!*Except on net neutrality

    1. Richard

      Your tumblr rocks

      1. andyswan

        Do I really still have one

    2. fredwilson

      I think we may need to explain that we are saying “trusted brands” aspirationally

      1. andyswan

        Love that. Has been one of our core values/aspirations in finance. Unfortunately a major differenciator.

      2. Prathamesh Shanbhag

        aspiring to be a “trusted brand” as they build the company?

  7. William Mougayar

    Going where the puck is headed is easier said than done.Crypto/blockchain are enablers of something else. They are the means to an end, not the purpose itself. It is the technology that leads us to new places, habits, knowledge, and benefits.

    1. sigmaalgebra

      >where the puck is headedLong ago, R. Hamming explained “The goal of computing is insight, not numbers.”There was also ‘Typing is no substitute for thinking.”Well by now, much of what is important and valuable from computing is good, new information. What we do is take in available data, maybe data we have deliberately arranged to be available, manipulate that data, and report the results as the coveted information.Well, this data manipulation is necessarily mathematically something, intuitive or not, heuristic or not, understood or not, powerful or not. For more powerful manipulations for more valuable information, we should proceed with, and the candidates are, intuition, heuristics, current machine learning and/or artificial intelligence, (so far, mostly just crude applications of 50+ year old regression analysis), what we have long understood how, at least in principle, to do manually, or pure/applied math, possibly new and for the particular problem. And may I have the envelope please [drum roll]? And, “RIP”, the winner is — pure/applied math.The big advantage of such math is the methodology of theorems and proofs. So, look at what we know about the real problem and see some features we can use as assumptions in some math. E.g., maybe it is reasonable to regard the data as probabilistically independent and identically distributed. Or maybe we have a production process, e.g., an oil refinery, and have good data on the costs and supplies of our inputs and the values of possible outputs so want to know what inputs to buy and what outputs to produce to maximize earnings. Then with these assumptions, see what theorems apply and what are the consequences of these theorems. See if the consequences are powerful for the real problem. If so, then likely have to write some software to do the data manipulations. Then can check the assumptions, the math, and the software; so, get some unique, world class quality control. That little paradigm has for 70+ years been a key to US national security and now is necessarily much of the future of computing, national security, the economy, and civilization and “where the puck is headed”.

  8. William Mougayar

    In other words, KCW + NPP = USV.

    1. fredwilson

      What are KCW and NPP ?

      1. William Mougayar

        From the thesis narrative:Knowledge, Capital, Well-being + Networks, Platforms, Protocols 🙂

        1. ShanaC

          down with three letter acronyms?

      2. sigmaalgebra

        They are TLAs, that is, three letter acronyms, one of the main curses of the computing community, the inarticulate and functionally illiterate, “gibberish” prone communications.I was going to ask the same question but have started to quit wasting time playing that game of whack-a-mole.So, sure, some deep ideas are difficult reading and, thus, obscure. So, can turn this around and use undefined terminology to be obscure. Then can pretend to be deep or maybe just be loyal to a favorite, shallow in-group. Mostly people have a lot more drive to be loyal to an in-group or tribe than to be clear on deep ideas.Still, a few obscure TLAs are a trigger for me to quit reading. I’ve carefully studied a lot of quite deep material by just terrific world class writers. In all of that, I never once saw an undefined term: In such material, such terms just do not happen. Uh, a term is a word used with a meaning not the same as in a standard dictionary. Then, in good writing, one never, not once, not ever, uses a term without defining, and hopefully motivating and explaining, and giving relevant examples, first.And for obscure, undefined TLAs, that’s just pulling fuming, reeking, toxic, sticky, black and orange bubbling, corrosive stuff out of a dumpster — cough, cough, throw open the windows and doors, run outside, and wash off with a garden hose.

  9. Tom Labus

    “Trust, but verify”

  10. Richard

    Your 2011 thesis was the superb; meaningful and marketable.2018 sounds like a pitch by an a NY advertisement company to Proctor and Gamble for their adult diapers division.

    1. fredwilson

      I agree with you about the 2011 version

      1. Richard

        Ha! Kind of like “you had me with Hello”

    2. JimHirshfield

      USV is very absorbent and never gives you ass rash. So, yeah, you’re right.

    3. Matt A. Myers

      Feelings, feelings everywhere; though I get what you’re saying – it’s potentially shallow and therefore its meaning and result could be as magical as it feels, though as it’s fairly vague – their decisions and thesis depends solely on how they trust.For example, if they only trust people who see a future using incentivized blockchain – to fit into their network (so their previous bets have seemingly better odds), and they don’t acknowledge that bias in the thesis statement – then well, that’s not as honest or clear as it could be.It could also mean they are truly open to learning and understanding and to what that woo-woo-like feeling the statement gives. If I’m able to continue my life’s work then I’d be open to seeing if we could work together still – as their thesis is a perfect match for my work.

  11. leigh

    It feels much broader than any one I’ve seen from USV. I used to be able to easily understand the thesis and could understand what would fit into it and maybe more importantly what wouldn’t. This kinda feels like you could drive a truck through it and that anything could be made to fit if you needed it to. I’m assuming it doesn’t feel that way to you guys so maybe I’m missing something…

    1. fredwilson

      We are passing on a lot that doesn’t fit in here but it is broader

  12. falicon

    1. I love that USV continues to work on this stuff, be open, and share it.2. With that being said, with each evolution I feel like it becomes harder for us outsiders to be on the same page (not that you need us to be; just an observation).When Brad tweeted his, it was fairly easy to look at a company or hear about a company and say “that’s the sort of thing USV would probably back”.When Andy tweeted his, it started getting a little harder to predict…but still, most “USV invested in this” announcements revealed the obvious “fit” in hindsight.As Rebecca tweets this, I’m left a little unclear on what is an “obvious” fit at the moment…or maybe better put, I’m unclear on what is an obvious “non fit” now.Maybe that’s intentional; but I feel like the comments are all sort of along these same lines… internally, you all prob. know what it means inside and out…and it prob. fits like a glove…but externally it’s obfuscated.

    1. Lawrence Brass

      I agree. On a first thought I think of it as if markets and products and networks are now more sophisticated. On a second thought I get that part of this sophistication is real and the other part is blurry, hype. Whenever I get a blurry idea about something I try to fight the ego telling me to discard it and then try to go deeper to see if I can get a grasp.It happens often that the mysterious and blurry part is just that, fluff. And when it is not it is a call to action, a pending upgrade you are owing yourself. I guess that you as a developer must have that feeling often. We are challenged.On the other side, from the point of view of someone who is communicating an idea, it is your responsibility to be well understood, not your audience’s.Don’t fear the blur.. or evolution.

  13. Frank W. Miller

    Looks generic enough to me that you can invest in just about anything…

    1. sigmaalgebra

      Of COURSE he “can invest in just about anything”; that is a special case of the fact that it is his VC firm! If there are some constraints, then those are not in the public thesis but in the private terms with the limited partners.An investment outside of the thesis needs just one really good — new, innovative, disruptive, radical, … — investment opportunity!!But such a thesis can be a security blanket for the limited partners, let them sleep better knowing that USV is not going off into crazy land.Such a thesis can be something of an assurance to entrepreneurs looking for funding because it can make USV look less arbitrary, less like USV is throwing darts in the dark, more responsible, with a better, to think of some words, “trusted brand”.But maybe in simplest terms the thesis is to increase desired deal flow and filter out the rest.With the thesis, when an entrepreneur doesn’t get funded, they can look at the thesis and see, with considerable uncertainty and ambiguity but with better than nothing, why they didn’t get funded. Or, “It’s not personal. It’s just business.”!But my working hypothesis is much simpler: The main criterion of information technology VC is to fund traction, hopefully earnings, well, maybe revenue, at least users, significant and growing rapidly. Given such traction that might result in a company worth $1+ billion and not too many warts or open, bleeding sores, there will, from a given VC firm, about 10% chance of a term sheet and check. Indeed, the situation is so simple that an entrepreneur might just take seriously the old Hollywood advice “Don’t call us. We’ll call you.”. Or, the entrepreneurs who easily get a check are the ones who have such a good business mostly they don’t need the check. And like the guy who started the Canadian romantic matchmaking company Plenty of Fish, an entrepreneur who doesn’t really need an equity check might not accept one even if offered. IIRC, Plenty of Fish was long just the one guy, the founder, with two old Dell computers, ads just from Google, and about $10 million a year in revenue. Finally the company head count went up, and, IIRC, the company was sold for about $550 million. So, what entrepreneurs would take the term sheet, the check, the Delaware C corp., the founder vesting schedule, and the BoD? Okay, there are five cofounders, each with credit cards all maxed out, each with a pregnant wife, and with traction significant and growing rapidly.

  14. Salt Shaker

    What’s a “trusted brand,” particularly for a start-up? Is it trust in the founders, trust in proof of concept, etc.? “Trust” is earned and generally achieved over time. The context of “trust” takes on diff connotations depending upon the life stage of product. Ultimately it broadens out to encompass end users (or customers), not at the expense of founders, etc., but in addition to them. If “trust” isn’t part of a brand’s DNA, whether b2c or b2b, then it’s not a healthy brand.

  15. Chimpwithcans

    Hard to argue with a track record of tweets and pre-tweets like this one. Broadening access makes me think there’s a lot of third world / Africa in your investment plans?

  16. raheeln

    chock-full of win. i like it

  17. jason wright

    if USV had been around at the time would it have invested in Google and Facebook, companies some say are infamous for centralising and privatising personal data? Does the thesis extend to a philosophy beyond returns?

    1. Twain Twain

      USV was founded in 2003. Facebook was founded in 2004. Twitter was founded in 2006.

      1. ShanaC

        feeling like the internet is getting old

    2. Matt A. Myers

      It feels like the success from their previous investments is what guides their future thesis. @fredwilson:disqus had the experience of being apart to Geocities happenings, and lead to what it did. This all is now making me realize they perhaps have blindspots from bias that has developed, though that they’re missing at least one crucial piece that I don’t want to try to name yet.

  18. Putting Fred out to pasture

    USV realizes that the Wild Wild West days of the internet are over. USV has a quixotic notion that LPs will be foolish enough to believe that by cooperating with the new robber barons, USV-backed ventures will obtain a few meager crumbs.Uh. Yeah. Hmmmm. So why exactly should the robber barons share any of their profits with any of these new USV-backed ventures when it is cheaper and easier to crush them?It’s time to for Fred to retire… or go back to shilling for BitCon, etc. He ain’t got no game no mo’.

    1. sigmaalgebra

      Apparently Fred’s been making a lot of money, i.e., high ROI, for his limited partners (LPs) for a long time. I’m sure Fred has some very happy LPs thrilled to continue investing in USV.But for the future? Computing and the Internet and their consequences for individuals, the economy, our society, and civilization are still growing very quickly, as fast as people can have ideas, build computers, write software, go live, and be successful. Saying that the rapid growth of computing and the Internet are over would be like saying, like some did, in 1900 that physics was over or when Carnegie had blast furnaces that iron and steel and metal working more generally were done. Heck, even growing corn, soy beans, wheat, chickens, eggs, cattle, milk, butter, cheese, and beef are not nearly done and, instead, are seeing continued efficiencies, in part from computing.I don’t expect Fred to be investing in milk, butter, and cheese, but likely he gets a wide, deep, rapidly flowing ocean of deals from the US, Europe, and more. Then, many are culled and few are frozen. He knows well, on average, how to pick winners, and I have to suspect that there are more good looking winners in his in-box than he or USV have time to invest.My approach to business and entrepreneurship are quite different from Fred’s approach to business and investing. Still at least for now and for a long time, there will be many more of what Fred wants than what I would want. Net, don’t sell Fred or USV short yet!

  19. Ciaran

    Your early versions were succinct, clear and actionable. This new version feels like it was written by committee, or pitched by an agency. It’s wooly.I appreciate that I’m not providing an alternative, but I don’t know your business as well as you do, particularly after reading your new thesis.

  20. Peng Jin

    If “trusted brands” is intended to be aspirational, not literal, then I suppose what it means is that 1) there is a more meaningful purpose beyond just seeking commercial gains; and 2) it has a good “taste”. This does sound more vague than the previous versions. Adding in a catch-all term like “well-being”, it does seem a lot less clear what the boundaries are. Should it be understood as USV reacting to a market environment where “structural opportunities” are diminishing and investment opportunities will be evaluated more on intrinsic merits?By the way, absolutely love the idea of sharing investment themes. This feedback loop is so powerful.

  21. Matt A. Myers

    Trust, I recently realized is the final layer of competition — where previously governance here has been discussed as such; you could argue trust comes through governance, however if the final importance of government is trust, I’d argue trust should be the named final layer of competition.Who do you trust more – who do you trust to be thorough, to be kind? And as the observer – how do you trust, are you a patient person who waits for a proper response – are you aware of your biases and are you strong enough to counter those — or are you reactionary, do you feed into hype and reaction, do you control to maintain the image you want to present – which generally perpetuates shallow depths of information?Do you trust Elon Musk or Mark Zuckerberg more? Are you properly connected to feelings, emotions – and therefore intuition – to help guide you, or have you learned to depend on ego mind and logic without the influence of proper feeling influencing the logic you’ve developed? How long do you spend feeling things out, investigating them – gathering deeper levels of knowledge around a topic, before coming to a conclusion? Are you open to forgiveness, are you able to detect when someone has truly changed or if they simply control with ego mind to change how their facade appears? Elon wears his heart on his sleeve.

  22. karen_e

    Nice use of the word “inexorably.”

  23. thomasknoll

    When people ask me what USV is investing in these days I’m just going to say “Companies designed to build trust while broadening access to knowledge, capital, and well-being.”

  24. Moe Abbas

    – Who can “become” trusted brands.

  25. Michael Elling

    “But this is actually the fourth USV thesis.” Easy fix. Call the actual, first thesis the origin or genesis thesis; so Thesis 0.

  26. David Gibbons

    Congratulations! Love “broaden access” – that’s the secret ingredient IMO … INCLUSIVITY is a one-word thesis that I subscribe to – anything that grants market access to new entrants will, by definition, also supplant the legacy solutions. Focusing on disintermediation is therefore a distraction – it’s just a byproduct of effectively improving inclusivity.