Posts from Uncategorized

A Note On Anonymous, Pseudonymous, Guest, and Regular Commenters

One of the best things about AVC is the engaged and active community that envelopes this blog. It has been for many years a conversation among friends and the occasional stranger. I’ve called it a bar where I get to be the bartender. The people in the community come and go. There are regulars who come every day. There are regulars who come every few days. Some come once every week or two. Some have left never to return. Some return on occasion. That’s all as it should be and quite like what goes on in the real world.

I’ve always chosen to allow people to comment using a guest login. I’ve always allowed people to comment anonymously. And I’ve always allowed people to comment using a pseudonym. I believe that allowing people to comment the way they want makes a community richer. I do not think comment identities should always be mapped to a real name and a real identity. It’s great when it is. But there are many reason why that’s not a good option for some.

We’ve managed the trolling and spam by actively moderating the comments. I did that for many years myself and in recent years I’ve been aided by AVC regulars William and Shana who swing by every day even when I’m not active to make sure a thread isn’t filling up with spam or there isn’t some sort of other bad behavior going on. Our moderation policy has been heavy to clear the spam and light on everything else. We lean in favor of giving everyone a voice even when its a tough call.

There is one thing that has evolved into a community norm that is important and I’d like to highlight today. Regular commenters use Disqus Profiles to comment here at AVC. These profiles can be pseudonyms like Fake Grimlock, abbreviations like JLM, or real names, like fredwilson. That really doesn’t matter and I think its best to have a lively mix of all of that. But the frequency of seeing the avatar next to the name in the comments breeds trust, respect, and in many cases real friendship.

If you are a drop in commenter, none of this matters. But if you want to hang out here on a regular basis, I encourage you to build a Disqus Profile, invest some time and energy into it, and participate as everyone else does. It’s how we do it around here and it is one of the many reasons this community works so well.

Can Mobile Banking Improve The Lives Of The Poor?

It feels to me like mobile banking is arriving. Whether its M-PESA in Kenya, Venmo in the US, or Bitcoin around the world, more and more people are using mobile services connected to the cloud to store and exchange money with other people and businesses.

And one of the big potential impacts of this trend is on the unbanked, those people who traditional banks won’t service.

The Verge’s Ben Popper and Bill Gates wrote a piece in The Verge this past week about the potential of mobile money to improve the lives of the poor. It’s a good read.

Bill Gates wrote this in his annual letter:

By 2030, 2 billion people who don’t have a bank account today will be storing money and making payments with their phones. And by then, mobile money providers will be offering the full range of financial services, from interest-bearing savings accounts to credit to insurance.

That’s 15 years from now, a long time for sure, but the part of that prediction that is the most important is the “don’t have a bank account today” part. That’s a huge number of people who will have the basic infrastructure in place to allow them to consume other financial services. That feels like a massive market opportunity to me. And it feels like a massive life improvement opportunity to me too. Doing well by doing good. There isn’t much better than that.

Some Thoughts On Seed Investing

We (USV) raised a new venture fund at the start of last year and started investing it in the spring of 2014. It is called USV 2014. We have made six investments in it so far and five of them are seed investments. That’s a very high ratio for USV and we do not expect that ratio to continue over the life of the fund. In fact our next investment will be a classic Series A so we are already lowering the ratio. But it is a bit of a return to form for USV as half of the initial investments in our first fund (USV 2004) were made at the seed stage.

In our core early stage funds (as opposed to our Opportunity Funds), we make initial investments at the seed, Series A, and Series B stages. In an ideal world for USV, there would be a normal distribution of these entry points with the highest percentage in the Series A stage. Over the entire history of USV, that is very much true. But on a fund by fund (or year by year basis) it varies a lot. It is mostly us reacting to the market. When the later stage rounds are too expensive on a risk/reward basis, we tend to move earlier. And when we can get good risk/reward opportunities in the Series A and Series B stages, we tend to move later. The downturn of 2008/2009, for example, led us to move a bit later in our 2008 fund because we could invest in more mature (and therefore less risky) opportunities at attractive prices.

The current market environment has pushed us to invest earlier. Some of it is that the Series A and particularly the Series B valuation environment has gotten very expensive relative to the risk as we see it. And some of it is that we are in a period of flux, where it is not entirely obvious to us where the next big things are going to happen. We have some ideas, of course, and I have been exploring them here at AVC and we have been exploring them as a team on usv.com. We think that in times of flux it is attractive to make a bunch of smaller seed investments in areas we think are going to emerge as important in a few years.

So that explains the move to seed as our primary entry point last year. I think it will continue this year but maybe moderate a bit as some of these developing markets mature and become more investable at scale.

Ok. Now that I’ve explained why I’m thinking about seed investing a lot these days, I’d like to talk about how we do seed at USV. Here are the important points:

1) We do not take a shotgun approach. We do not view seed investments as “options”. We only make a seed investment if we have as much conviction on the team and the opportunity as we would at the Series A round. We are as committed to our seed investments, both in terms of the time we spend with them and the willingness to follow-on in them. They are core investments with as much stature in our portfolio and in our firm as any other early stage investment. This is critical to understand. And it is not true of many (most??) VC firms who make seed investments.

2) We like seed investments in teams and opportunities where they have built and launched a product already. We don’t like investing in a concept or participating in a round where the uses of the capital will be to build and launch a product. This means the vast majority of seed rounds are not a fit for us. We pass on a lot of seed stage opportunities because it is “too early” for us. That is a comment on the specific opportunity however, and not seed stage investing as a whole. This confuses a lot of people. They tend not to think of USV as a seed investor when in fact we do make a lot of seeds (over 80% of last year’s investments, for example).

3) We will often lead the Series A (and sometimes Series B) in companies where we did the seed investment. We led both the Series A and Series B in Etsy and we co-led (with Spark) the Series A and Series B in Tumblr. We were seed investors in both companies. We continue to do that where it makes sense for the founders and USV. That is not a requirement or an expectation, but it does happen and I believe it is a very good thing in the right circumstances.

4) We like to participate in syndicates in our seed investments. We don’t focus too much on ownership at the seed stage. We do focus on the investors coming together around a project. We like partnering with smart angels, seed funds, and even other VCs, if the other VCs are aligned with us on how they are thinking about the particular seed investment. Our investment with Spark in the seed round at Tumblr is a good example of two VC firms partnering up at the seed round and doing a good job working together and scaling into the opportunity.

USV will never be confused with a seed fund, but we sometimes act very much like one, except that we can and will invest 20-30x our initial investment over the life of the company. That combination (a committed and active seed investor + deep pockets) is unusual. You can get one of those two a lot. But rarely both. So if you are working in an area that is interesting to USV, and if you have launched something into the market already, and if you are doing a seed round, please do reach out to us. We are in the business of making seed investments and doing a lot of it these days.

A Lens Into The Future Of Enterprise Software

I’ve been working with our portfolio company Work Market for four years now. It’s been a real learning experience for me as enterprise and SAAS has never been my long suit. We were attracted to Work Market because, as their name implies, they use a marketplace model to help enterprises get work done. Specifically, they created and are the leader in the Freelance Management System market. We like software that has a network effect built in because it is harder to commoditize. A marketplace of freelance workers inside an enterprise software application seemed to us to be exactly that. And that has been true. But along the way we’ve learned quite a few other things:

1) Mobile matters, a lot. I mentioned in my What Just Happened post that mobile is starting to really impact the enterprise software business.

At Work Market, the freelancers want to get work, accept work, and close out work on their phones. So mobile app development has become a huge part of what the Work Market engineering team has to work on. At some point, the enterprise will likely want to issue work orders on their phones too.

2) Freemium and transactional business models work in the enterprise just as well as they work in consumer. Work Market has a free tier with a transactional revenue model for enterprises that want to try the system or plan to be an occasional user.

WM pricing

We know that “freemium SAAS” works well for horizontal enterprise applications like Dropbox, Slack, Google Apps, etc and I believe we will start seeing freemium SAAS models applied to vertical applications as well. We already are.

3) Enterprise applications must also be platforms if they want to scale into the largest enterprises. Salesforce is the poster child for this trend. They have become a very powerful platform and distribution partner for SAAS applications. But every SAAS application should have APIs that allow their users to plug enterprise software together. Work Market can talk to the other large applications that enterprises use for managing talent (HR, VMS, etc) and that is a requirement for the largest deals. It will soon be a requirement for all deals.

I am seeing a bunch of new SAAS companies get started whose entire value proposition is building on the open APIs that most enterprise SAAS products have released in the past few years. If you are in finance, or HR, or marketing, or sales, you are now using a host of SAAS applications to get your job done and a big trend in the market is new applications that tie all of those together (via APIs) so that you can have a single view into your workflows. This is the “platformization” of SAAS and it is upon us.

The big takeaway for me is that all the things we have seen happen in consumer web and mobile software are happening in the enterprise and the impact of that is already being felt. I am seeing it up close at Work Market and fortunately they got started recently enough that they have been able to take advantage of all of these trends (marketplaces, mobile, cloud, freemium, platform) as they go to market and build their business. That is, among many other reasons, why we recently led a growth round to help Work Market’s new CEO, Stephen DeWitt, scale into the freelance management system market opportunity that they created a few years ago.

Video Of The Week: The Berlin Conversation

Last month, my partner Brad and I found ourselves in Berlin at the same time. We decided to do an event for entrepreneurs in Berlin with the help of the folks from Tech Open Air Berlin. We rented a space, brought in some entertainment and food, and then had an hour long conversation about technology, startups, trends, and a little bit about Germany and Berlin.

Brad and I started USV together in 2003 and we’ve worked closely together for over a decade now. We rarely (if ever) do public appearances together so this was a pretty special event for us. I think this video does a good job of showing how each of us comes at the issues from a different place but get to the same answers most of the time. It’s been a very successful partnership and I think this video shows some of why that is.

Feature Friday: The Dashboard

Our portfolio company Duolingo did something interesting this week. They launched a new product by adding a dashboard to the existing product.

For those that don’t know, Duolingo is the most popular language learning app on mobile phones (and the web) around the world.

They saw many students and teachers around the world adopting the Duolingo app as learning tool in the classroom even though Duolingo was built as a horizontal product without any specific use cases in mind. So they thought that building a version specifically designed for schools would be a good idea. And the big new feature that allowed them to do that was a dashboard for teachers. The students use the same app they’ve always used and the teachers get a tool to help them understand, manage, and react to each student’s individual progress.

This is a great example of the consumerization of vertical applications and markets. There are plenty of companies that try to sell technology into schools and school systems top down. They build all the features to support schools day one and then attempt to get schools and school systems to purchase their product.

Duolingo was shipped as a free mobile and web app that anyone can use to learn a language. Teachers and students adopted it. And then Duolingo shipped a single feature, the dashboard, that makes the service way more useful and impactful in the classroom.

Duolingo is entirely free to use, including the new Duolingo for Schools. Duolingo monetizes its business by providing certification services like Duolingo Test Center. I wrote a bit about this freemium model in education last summer. I think its a powerful model when combined with a bottoms up (consumer) go to market approach.

The State Of Bitcoin – January 2015

This is a slideshare from Coindesk:

I think this slidedeck shows that 2014 was a tough year for bitcoin, something I mentioned in look back and look forward posts. This year has started off poorly with problems at Bitstamp and the price breaking below $300, probably in relation to the Bitstamp issues.

It’s tough times for the bitcoin sector but venture capitalists continue to make investments and transaction volumes continue to rise. I believe we are witnessing a transition from one phase (monetary and speculation driven) to another phase (new applications and bitcoin as a platform) and there will be a shakeout as the transition happens. I think we are well into that shakeout.

Finding ROI In Higher Education

The news is full of stories where students paid hundreds of thousands of dollars to go to college (and beyond) only to find themselves stuck in dead end jobs and unable to pay off the cost of student loans. We have a crisis in the US in higher education. The costs have risen and the benefits have declined.

It has gotten to the point where I believe if you have to personally shoulder the cost of your higher education, you should think twice about the traditional model. If you can get scholarships or if your parents are willing to pay the tuition bills, I still think its a valuable experience, but sadly it is not one that makes sense if you have to make the investment personally.

So what are we going to do about that? We need to find new models. And one new model that is working in NYC is The Flatiron School. The Flatiron School started two years ago and teaches students, both high school grads and college grads, how to become software engineers in a twelve week course that costs $15,000. Scholarships are available for students who cannot afford that investment.

Today The Flatiron School has published an audited report that validates the notion that their model produces graduates who can find high paying jobs. Here is a summary of the report and this is the “money slide” from it:

cost and benefit of flatiron

So for a high school graduate, the tuition at Flatiron can be paid back with six months of after tax income. For a college graduate, you can increase your pay by ~$30k by spending $15k. You get that payback in one year of after tax income.

For the average college grad, it takes roughly three years of all of your after tax income to pay off your college costs. If you go on and do Flatiron, you can pay off everything with two years of after tax income.

Anyway you cut the numbers, The Flatiron School is a great investment. Part of it is that the students learn a valuable skill – software development. Part of it is that the cost of delivering that education are very reasonable. And it isn’t that they do this on the cheap. Here is the work required from a student at The Flatiron School:

educational activity at Flatiron

There’s been a lot of talk that online education is the answer to lowering the costs of higher education. The huge investment in MOOCS that happened a few years ago was based on that notion. The reality is that online education is a part of the answer but not the silver bullet that some thought it would be. I gave a talk at Wharton a couple years ago about this.

The answer to lowering the cost and increasing the benefits of higher education requires a multitude of changes to the current model. And one of them is teaching students skills that are directly related to job requirements. Doing that makes students more employable and more valuable.

This is not a criticism of the liberal arts model, per se. As Steve Jobs said in this interview, learning to code is a liberal art. This is a criticism of administrations and faculties that are rigid in their interpretation of what liberal arts and education should mean. This is a criticism of not evolving and changing with the times. This is a criticism of thinking what worked yesterday will work tomorrow.

And mostly this is a criticism of not making hard choices. Schools that are happy to add courses, faculty, and buildings are not willing to eliminate courses, faculty, and buildings. When you always add and never subtract, you get cost structures that are not sustainable.

The Flatiron School is an example of what can be done with a blank slate. They have figured out how to give students highly relevant and valuable skills at a cost that is both affordable and recoupable very quickly. Adam, Avi, Sara and the entire team has created a model that should be an inspiration for others.

Video Of The Week: A History Lesson On Why We Need Neutral Networks

My partner Brad went down to Chattanooga where they have a gigabit fiber network around the city and attended an event about connectivity and what it does for society.

In this short (~10mins) talk he gives a history lesson on how we got permissionless innovation on the Internet and why we could lose it.

Hashtags As Social Networks

Our portfolio company Kik launched hashtags yesterday. Kik is a mobile messenger so in Kik’s model hashtags are private or public group chats.

If I send a hashtag to a friend in Kik that says let’s chat about tonight’s knicks game at #knickskik, then that becomes a private group between me and that friend (and any others who we invite). I’ve done that so the #knickskik hashtag is now private on Kik.

But hashtags can also be public. If you have the latest version of Kik on your phone (came out yesterday), type #avckikgroup into a chat and then click on that link. Up to 50 of us can be in that group.

The cool thing about Kik is that it doesn’t use phone numbers like other messengers. It uses usernames and is not tied to your phone number or Facebook username. And so Kik, unlike other messengers, is used for both chatting with people you know (like other messengers) and people you don’t know.

That makes Kik an ideal platform for these public (and searchable) group chats. You can meet people in these public chatrooms and then take your conversations private in a one to one chat in Kik.

Ted Livingston, Kik’s founder and CEO, called this “hashtags as social networks” in a blog post yesterday.  I agree with Ted that Facebook’s model of the one network to rule them all has not really worked and that many of us are using messengers as defacto social networks. My friend Kirk told me that his wife’s family uses a group in WhatsApp like their personal family facebook feed. I think that’s the phase of social networking we are now into and so Kik’s hashtag as social network model makes a ton of sense to me.