DockerCon 2015: Outside the Echo-chamber

DockerCon tore through SF last week and the feeling is that we are at the apex of the hype cycle. Fear not, we at Redpoint are here to (attempt to) distill signal from noise. Here’s a recap of the top story-lines as we see them along with some thoughts…

You down with OCP…?!

What happened: Docker and CoreOS got on stage, kissed and made up and announced the Open Container Project (‘OCP’). OCP is a non-profit governance structure, formed under the Linux Foundation, for the purpose of creating open industry standards around container formats and runtime. You may remember back in December ’14 CoreOS made headlines by announcing rkt, an implementation of appC, the company’s own container image format, runtime and discovery mechanism, which, in contrast to Docker’s libcontainer, was open, both technologically and in its development methodology. Then in May at CoreOS Fest, CoreOS’s inaugural conference, momentum for appC appeared to be gaining steam and image format fragmentation seemed inevitable. Instead, a mere seven weeks later, it appears Docker and CoreOS are willing to put aside differences to work together (and with the likes of Google, Amazon, Microsoft, Red Hat, and Intel) towards an open container spec.

Our take: The big winner is the broader container ecosystem. There are at least half dozen credible alternatives to Docker’s libcontainer emerging, and while competition is generally a good thing, the introduction of multiple different image formats creates ecosystem fragmentation which constrains customer adoption and broader momentum. Consolidation around the OCP spec will ensure interoperability while enabling vendors to continue innovating at runtime. More importantly, by agreeing on low-level standards, the community can move on to solve higher-order problems around namespaces, security, syscalls, storage and more. Finally, the loser in all this appears to be the media now that there’s, at very least, a ceasefire in the Docker-CoreOS war.

Docker Network and more dashed startup dreams

What happened: In early March of this year Docker acquired Socketplane to bolster its networking chops and the fruits of that acquisition were displayed in a new product release called Docker Network, a native, distributed multi-host networking solution. Developers will now be able to establish the topology of the network and connect discrete Dockerized services into a distributed application. Moreover, Docker has developed set of commands that enable devs to inspect, audit and change topology on the fly – pretty slick.

Our take: The oft-forgotten element to enabling application portability is the network – it doesn’t matter if your code can be executed in any compute substrate if services can’t communicate across disparate network infrastructures. Docker’s “Overlay Driver” brings a software-defined network directly onto the application itself and allows developers to preserve network configurations as containers are ported across and between datacenters. The broader industry implication here is that Docker is continuing to platform by filling in gaps in the container stack. The implication for startups? You will NOT build a large, durable business by simply wrapping the Docker API and plugging holes.

Plug-ins and the UNIX-ification of Docker

What happened: Docker finally capitulated to industry demands and announced a swappable plug-in architecture and SDK which will allow developers to more easily integrate their code and 3rd-party tools with Docker. The two main extension points featured were network plugins (allowing third-party container networking solutions to connect containers to container networks) and volume plug-ins (allowing third-party container data management solutions to provide data volumes for containers which operate on stateful applications) with several more expected soon.

Our take: For a year now there’s been an uneasy tension between Docker and the developer community as Docker became less a modular component for others to build on top of and more a platform for building applications in and of itself. The prevailing fear was that in Docker’s quest to platform, it would cannibalize much of the ecosystem, create lock-in and stifle innovation. Docker’s party line has always been that “batteries are included, but swappable,” implying you can use Docker tooling out of the box or swap in whatever networking overlay, orchestrator, scheduler, etc. that works best for you.  The plug-ins announcement is a step in that direction as it appears Docker is finally not only talking the UNIX philosophy talk, but walking the walk.

Container Management Mania

What happened: Whether it’s called “containers as a service,” “container platform,” “microservices platform” or plain old “PaaS”, it’s clear that this is the noisiest segment of the market. We counted no less than 10 vendors on the conference floor touting their flavor of management platform.

Our take: Everything old is new again. The evolution of container management is analogous to that of cloud management platforms (“CMPs”) when virtualization began invading the datacenter. There were dozens of CMPs founded between 2006 and 2010 the likes of Rightscale, Cloud.com, Makara, Nimbula, etc. Several have since been acquired for good, but far from great, outcomes, and the sea is still awash in CMP vendors competing feature for feature. Correspondingly as the compute abstraction layer moves from the server (hypervisor) to the OS (container engine), a new breed of management platform is emerging to provision, orchestrate and scale systems and applications. Will the exit environment this time around mirror the previous cycle?

*   *   *   *   *

Stepping out of the echo-chamber, the big question remains around adoption. There are some technological gating factors that will inhibit enterprise deployments in the short-term – namely persistence, security and management – but the overwhelming constraint holding back containers appears to be general lack of expertise and established best practices. The good news is that these are “when” not “if” issues that pertain to ecosystem maturity, and the steps taken by Docker last week will only help accelerate that process.

With the groundwork laid, we see an exciting year ahead for the container community.  The inevitability of container adoption only feels more inevitable now.  There are many hard problems to solve, but hopefully (fingers crossed) there is now more alignment within the community.  Start-ups and large enterprise companies alike can begin, in earnest, the real work required to drive broad adoption of this technology in datacenters.  Hopefully we will look back a year from now and feel like this was the year that the technology moved beyond the hype phase to real adoption.


Redpoint Backs BitGo

At Redpoint, we are strong believers in the potential of bitcoin to change the way transactions are conducted, both online and offline.  But this promising system of digital currency is still in its infancy, a reality that has been underscored by a spate of security troubles. In numerous cases, bitcoins have been lost or stolen when wallets have been compromised, when exchanges have been improperly secured, or even when individuals have been careless with their keys.  Bitcoin has relied heavily on people and institutions protecting their own keys, and this has been a point of failure for the currency and technology as a whole.

In other payment and Internet commerce systems, third party security providers have helped resolve these issues.  This includes Verisign for websites as well as the card issuers for credit cards and merchants.  We believe that BitGo has the potential to be that provider for the bitcoin industry, which is why we are excited to announce our investment in the company.

BitGo has built the first-ever, simple to use and enterprise-grade multi-signature technology that works with the existing blockchain.  BitGo requires each account to have three keys – any two of which are required to recreate the private key and participate in a transaction.  Typically, one of these BitGo generated keys is maintained by the owner, one by BitGo, and the third is usually held by a trusted third party or in remote, or offline storage controlled by the owner.  Any transaction requires that two of the three keys be used, just like a bank safety deposit box – no single key is enough to open it. Even if a single key is lost or compromised, the owner’s bitcoin cannot be stolen or accessed.  Indeed, this could only occur if both the user (or his/her wallet) AND BitGo are compromised.  This also allows BitGo to act as a fraud detection and prevention player, helping protect users who are themselves hacked, while still giving the user the ability to access and spend his or her bitcoin without BitGo participating in the transaction.  BitGo is working with many major bitcoin players to implement this technology across the entire bitcoin ecosystem.

BitGo also has an exceptional team behind it. The team has a history of involvement in key technological innovations at some of the most dynamic technology companies, including Google, Facebook, and Big Fish Games. In addition, former VeriSign CEO Stratton Sclavos is a new BitGo investor and board member alongside Redpoint. The involvement of the man whose company made it safe to put credit card numbers on the Internet only adds to BitGo’s reach and credibility.

We are truly excited to join such a pioneering team and company and look forward to the great developments ahead.


Data-Driven Applications: The Next Generation of Big Data by John Walecka, Scott Raney and Chris Child

In a sign of things to come, a few months ago startup RelateIQ announced that it has raised $40 million in venture financing, the largest amount to date for an emerging class of enterprise applications driven by big data.  These applications combine the flexibility of SaaS with intelligence gleaned from big data that help users make quicker and better decisions in their jobs virtually every day.

It’s a big bet in a space that is just beginning to show its enormous potential. Up to now, big data has been defined by players like Palintir, a leader among companies focused on addressing shortcomings in the enterprise infrastructure. Specifically, these big data infrastructure companies have enabled complex organizations and businesses—from Homeland Security to Wall Street–to analyze and utilize their treasure troves of information. To date, the market for these innovators tops a hefty $16 billion, according to IDC research.

But here’s the thing:  Much of the value derived from these technologies thus far is primarily accessible only through data scientists and engineers. They spend countless hours culling through and then analyzing information to reap the most meaningful insights they can. Business users, charged with making split-second decisions every day affecting everything from sales to marketing to finance, have been left waiting—and wanting.

Until now, RelateIQ is just the latest in a new breed of so-called Data-Driven Application upstarts poised to fill the gap. They are coupling big data with machine learning to fundamentally change the way business users make decisions. In the case of RelateIQ, which targets customer relations management (CRM), the company hopes to make it simple for users to track and manage all their business relationships by analyzing data from call logs, emails and calendars to recommend follow ups, optimal times to reach out, and make sure that no promising relationship is overlooked.

If everything goes as expected, virtually every category of enterprise applications will be transformed by the insights automatically derived from a multitude of data sources. In just the last 12 months, an estimated $150 million has been spent funding startups focusing on everything from security to ITSM to advertising to human resources.

Their missions follow much of the path already blazed by large, consumer-facing Internet companies, such as Facebook, Amazon, Netflix, and LinkedIn. They have leveraged the massive amounts of data they collect and process to provide friend recommendations, personalized content, suggested products, and to help target potential new employers and employees. These processes have led to smarter, faster and more strategic decision-making.

Business users increasingly want the same level of sophisticated, quick insights powered by smarter software. And they are likely to get them from the dozens of companies sprouting up across the country.

This new generation of big data enterprise startups is a natural evolution from what began in the late ‘80s with the emergence of client-server companies. Peoplesoft and Siebel provided packaged software to help companies manage their employees, finances and customers. Then, about a decade ago, SaaS companies like Salesforce and Workday leveraged the Internet to change the way software is consumed and delivered. As powerful as the move to SaaS has been, it is clear that businesses today expect more than just automated processes. They want help making quicker, better and smarter decisions, every day.

That is where the powerful convergence of big data and SaaS comes in. Though virtually every sector will be moving toward these innovative software tools within the next few years, here’s a breakdown of some of the industries most likely to feel—and embrace—the transformation first:

  • Cybersecurity: Security companies have long focused on new technologies to defeat malware, but recently user behavior has come under scrutiny as well.  A new generation of companies like Fortscale is tracking user access logs to find unusual patterns that may show either a compromised computer or an employee with malicious intent, which they call the Snowden problem.
  • Human Resources: HR organizations have long embraced systematic and technological ways to measure employee performance and potential.  Recently, analytics has become part of this. Gild is one company taking a software developer’s entire online presence (including open source code but also social media profiles) and comparing it with known successful engineers at a hiring company. These profiles let them determine if an engineer is not just a good fit professionally, but also culturally.
  • Sales and Marketing: The art of converting marketing leads into sales leads and then bankable deals has always been a bit of black magic practiced by top sales people.  Infer is a company working to bring data science to the problem by building an enormous model of a company’s past leads to close deal conversion, and tying those deals to all kinds of outside data they can track down. Then, they use that model to predict which incoming leads are most likely to convert, and thus, where to allocate salespeople’s time.

The multi-billion-dollar opportunities for this next generation of big data innovators are as immense as the problems they are trying to address. And if they get it right, they will undoubtedly touch not just every function within a company but every person as well.


John Walecka and Scott Raney are partners at Redpoint Ventures, a backer of RelateIQ and Infer. Chris Child is a senior associate.


Another Marketplace Ready for Reinvention: Why we invested in Beepi

Every year since Redpoint was founded we’ve made a marketplace investment. Technology offers brilliant ways to lower friction, enhance turnover, discover favorable economics and surface marginal demand in existing marketplaces. Stars in our portfolio like Loopnet, BlueKai and HomeAway.com have set the bar for others to follow, and exciting new investments like Thredup, Homejoy, The Receivables Exchange and Axial are on promising trajectories. They all have a common thread – a team with a determined vision to take on an existing marketplace in need of radical new solutions to simplify and speed up the connection between demand and supply.

One such category that has not yet been truly reinvented since the dawn of the internet is the process of buying and selling a used car. That’s why we are thrilled to announce our investment in Beepi – a company that has made it dead-simple for individuals to buy and sell used cars. Beepi bests the existing consumer alternatives – car dealers (who provoke a well founded mix of fear, uncertainty and doubt), and cumbersome, risky and complicated, peer to peer selling sites (such as Craigslist)– with an elegant solution that enables an individual’s second largest financial purchase to become a simple, high value, risk-free and positive experience.

After a seller’s car has become Beepi-certified by passing Beepi’s comprehensive pre-purchase inspection process, and has accepted an attractive and guaranteed price from Beepi, the Beepi created listing is made available to buyers. Beepi handles all the listing creation, payment, paperwork, pick-up and delivery of the vehicle to make the experience completely frictionless. Buyers can find attractively priced vehicles, without negotiating, or driving all over the state, and rest easy knowing Beepi has not only kicked the tires to find them a great car, but also offers a simple, highly reliable transaction and an unprecedented 10-day no questions asked money back guarantee.

In short, the potential for Beepi is to not only reinvent the car selling and buying experience for the better, but to change how consumers think about when and how to buy and sell their cars. We believe that Beepi can meaningfully penetrate the $300 Billion used car marketplace.

We are very excited to partner with two dedicated co-founding entrepreneurs in Ale Resnik and Omer Savir. Ale and Omer both have had the personal experiences and broad vision to build something really big in this category. In addition, we have assembled great partners and co-investors who bring to bear a great wealth of experience, insight and support in building this market: Fabrice Grinda, founder and Chairman of OLX –the Craigslist of Europe, and also chairman of Beepi; Brian Sharples, cofounder and CEO of HomeAway, the world’s largest vacation home marketplace and former board member of iMotors, Rich Boyle, the former Chairman and CEO of Loopnet, the largest online marketplace for commercial real estate, and Tina Sharkey, founder of iVillage and former CEO of BabyCenter.

It will be an thrilling ride, (pun intended) and we’re looking forward to Beepi’s first chapter here in the Bay Area. Check out their cars now – and maybe get ready to sell yours.


Reflections On YCombinator Demo Day: How The Seed Market Has Changed

Earlier this week, I attended the Spring YCombinator Demo Day. I’ve been attending for six years now. Each time, I’m impressed by the intelligence, ambition and the polish of the founders presenting companies only a few weeks or months old.

As I listened to the pitches, I wondered if the types of startups founders decide to build at YC has changed over time and whether those trends are lagging or leading indicators of the market as a whole. At each Demo Day, the YC team provides investors a list of all the companies pitching and I’ve kept a few. To get a sense of the broader trends in YC companies, I’ve compared the Winter 2012 class and the Spring 2014 class by sector (consumer v. enterprise), segment (ecommerce, education, social, gaming, delivery) and by revenue model (subscription, ads, transactional).

These are the trends I observed in the data:

Mild shift toward enterprise: In 2012, 48% of YC startups were enterprise. In 2014, enterprise startups were 57% of the class.

Within enterprise, there has been a shift toward industry specific Software-as-a-Service (vertical SaaS) at the expense of horizontal SaaS. Vertical SaaS startups comprised 29% of the 2012 enterprise companies and 40% of the 2014 class. To make this idea more concrete, here are two examples. VidPresso provides software to the TV broadcast industry and is an example of vertical SaaS. ZenPayoll, a provider of payroll services, serves many different types of businesses and is a horizontal SaaS company.

Platforms-as-a-Service, which enable developers to build and scale applications (Heroku), have also seen a decline in numbers. In 2012, there were 5 PaaS companies while in 2014, I counted 2. Shifting to consumer, social apps have fallen from 24% of consumer startups to 15% at the most recent demo day. Unlike the 2012 class, there were no gaming companies in 2014. Food delivery companies, education companies and consumer market places have cropped up in their place.

As for revenue models, subscription remains dominant. 53% of 2012 YC companies chose this revenue model and 56% chose it in 2014.

Also notable is a marked increase in the number of non-profits. The 2014 class graduated 6 of them, up from zero in 2012.

All in all, YC startups do seem to be shifting with the market and/or YC partners are screening for startups that are more reflective of the environment. The shifts toward vertical SaaS and away from social and gaming apparent in this class are consistent with the patterns I’m seeing in the fund raising market. Unfortunately, the data isn’t able to tell us who is setting the trend. In any case, I’ll be tracking these trends in the future and hope to be able to draw more conclusions over time.


ThredUp: Leveraging Data to Understand Customers

ThredUp, the leading online resale marketplace for women’s and children’s fashion, has gathered interesting data points on its customers’ shopping trends. Here’s a quick look.

Ever wonder which clothes in your closet might fetch the most money? Or in which cities people shop the most for “practically new” shoes and dresses? Or which clothing brands hold their value the most? These are the kinds of juicy shopping tidbits you will find in thredUp’s Second Annual Resale Report. Founded in 2009, the fast-growing company has seen it all when it comes to fashion trends—and is willing to dish. Here are just some of the most interesting insights from 2013:

Of the 10,000 U.S. cities embracing the online fashion resale marketplace, residents in San Francisco, Brooklyn, Seattle, Chicago, Houston, Miami and Los Angeles are among the most enthusiastic. Indeed, shoppers from the top 10 cities collectively saved almost $1.8 million buying and selling clothes on thredUp in 2013.

What were these folks buying, you might ask? The fastest-selling brands for women include Burberry, Coach, Gucci, Toms, and True Religion. For kids, Zara, Crocks, Patagonia, Ralph Lauren and Matilda Jane were among the hottest. These brands, part of the more than 20,000 sold on the site, typically sell out within hours of being listed.

Of course, much of what you buy depends on where you live. thredUp checked that out and found some interesting—and surprising—results. For instance, Weston, FL residents bought up the most activewear clothing. The most shoe-obsessed population seems to hale from Lancaster, PA. Preppy clothing is biggest in Fontana, CA while formal wear is more the norm in Pleasant Shade, TN. Who buys the most designer brands? That honor goes to the fashionable people of Brooklyn, NY, who purchased more pieces by the likes of Stuart Weitzman, Missoni and Dolce & Gabbana than anyone else.

With savings of up to 90% off retail, the online resale marketplace is growing fast. In 2013, stripes, florals and plaids were the best-selling patterns. Wonder what they will be in 2014?


The Super Bowl of Startups: Interview with DraftKings CEO Jason Robins


As Super Bowl XLVIII fades into the history books (hopefully never again to be so lop-sided a defeat), another game is quietly gaining steam as the place to be for the most avid sports fanatics. DraftKings, a leader in online fantasy sports games, has seen its customer base, daily engagement, and giveaway prize totals surge almost as quickly as the Seattle Seahawks have transformed themselves into Super Bowl champions.

Launched in 2012, DraftKings paid out $50 million in prizes in 2013 to thousands of players who excelled in weekly fantasy football, daily fantasy baseball, daily fantasy basketball and daily fantasy hockey. In just the last four months, the Boston-based upstart’s user base has grown fourfold, with nearly 50,000 active daily users and as many as one million registered players. Even better, DraftKings users spend an average of more than two hours every day on the site.

All this before the company’s new mobile app was launched a few weeks ago and new games, such as golf, were added to the existing stable of sports. CEO Jason Robins says his company is poised to blow past these impressive numbers in 2014, thanks to more aggressive marketing and a slew of product enhancements.

With just 32 employees thus far, accomplishing so much so quickly is no easy feat. In a candid interview, Robins opens up about what he believes has been key to DraftKings high-octane growth and about the unique challenges that lie head.


Q: What has been your formula for success?

A: We have spent a lot of time focusing on the brand and the customer experience. Three things have been key: For the customer, we want everything to be fun, easy to use, and engaging. We look at a lot of metrics for every feature we roll out. We see what’s working and what isn’t by testing everything with a group of customers.


 Q. Can you give an example of how that has worked?

A. In the beginning of the NFL season, we launched a feature so people could send private challenges to friends. The social aspect of fantasy sports is huge and we needed to provide an easier way for people to play against their friends. After launching the feature, we found that a lot of challenges were not getting accepted. We wanted to know why and figured out that a lot of challenges were being lost in people’s email inbox.. At the same time, many of these same people were coming to our website without knowing about the challenges their friends were sending.

So we launched a new system of notifications on our website so people could see an alarm bell that shows a challenge. As soon as we launched this feature, our declines of challenges went down—70% on the first day. We will soon launch a push notification on our mobile app, which should make it even better. If you are out and about, you will know when someone has challenged you to a game. In addition, we just launched a feature where you can reserve a seat for a game without having to pick your team right away. People know they want to play but maybe can’t pick their team at that time. So this lets them accept a friend’s challenge and then later set their lineup. We have seen another 45% drop in cancellations with that.

All this was about tackling one key metric: the cancellation rate of game challenges.


Q. What are some things that have surprised you in getting the business off the ground?

A. Sometimes we are surprised by how our users respond to new features. We launched something called Lineups, which is a way to easily manage your teams. If you create 40 different teams and have players across many games, it can be hard to manage when a key player you had on 30 teams suddenly is out because of an injury. Previously you would have to edit out that play in each game. Now you can get the guy out of your lineups automatically. We knew this would be popular but it has become one of the defining features of our product.

Another lesson we have learned is about marketing. I came from Vistaprint, which had success with digital marketing channels at an incredible level of scale before venturing into offline advertising. It took us awhile to realize that wouldn’t work for us at DraftKings. Broadcast marketing channels have been easier for us to scale than digital.


 Q. Is it difficult to build a virtual business that is so dependent on the real world?

A. The beauty of sports is that you already start with a great product. It’s the world’s greatest reality show. We don’t have to create the content. It’s already there. For us, it’s more about how to add to an already interesting experience. Our app fits in particularly well. For example, if you are at a game, it can be a second-screen experience. It opens up a lot of different ways to add on to the experience.


Q. What have been some of the biggest challenges?

A. We are a product-focused culture and we always have to make sure we continue to pay as much attention to the technology and back-end operations. We have a very complicated product with constant transactions occurring. So the issue of security is always there and we have to have a really strong infrastructure that is error free. We can’t have people lose trust.

When we roll out a new feature, we make a list of all the negative things that could happen from launching the feature. Not just bugs but product design etc. We have to have an answer for each one or we don’t roll out the feature. We want to grow fast but there’s a lot of tricky stuff on the technical side.


Q. What keeps you up at night?

A. Staying ahead of the competition on the innovation front and making sure that everything we do is high quality. I also worry about hiring and retaining the best talent. It’s a little easier because of the business we are in. People feel a lot of passion for sports. It also helps being in Boston.




The Best Times of Year to Raise Capital for Your Startup

Aside from a startup’s internal considerations about the right time to raise money, founders should weigh the seasonality of the fund raising market when planning their raise. There’s a rule of thumb batted around the valley that the worst times to raise capital are in the dog-days of summer and after Thanksgiving. As it turns out, this aphorism is only a half-truth.

Below is a chart of the dollars VCs have invested by month of year. I’m using Crunchbase data since 2005 for tech companies in the US. There are a few notable trends in the data.

First, the impact of the summer is evident. The slowest month for investments during the year September. I’d estimate there are a few weeks latency in the data between when the investment commitment is made and the investment is disclosed. The legal diligence process of about 3-4 weeks that typically follows signing a term sheet introduces this lag.

Second, setting aside the slowdown in summer, VCs invest more and more as the year progresses. It’s not crazy to draw a parallel between this trend and the patterns quota attainment for salespeople in which the majority of sales arrive in the last two weeks of a quarter. VCs aim to invest a certain amount of dollars and/or in a certain number of companies each year. The later in the year, the greater the time pressure, the better the motivation.

Third, the difference from the best month, December, to the worst month, January, is substantial: a 75% delta in the total dollars invested and 50% increase in the number of investments. These differences are statistically significant with greater than 95% confidence, so seasonality quite clearly exists.

This analysis raises another question: is there an optimal time of year to maximize the size of an investment and presumably pre-money valuation?

The data shows the average investment size spikes in August and December, both of which have over the past nine years, generated 15% larger rounds. The difference is statistically significant with greater than 90% confidence.

Assuming a three month fundraising process, the best times of year to start a financing process is in September, targeting a December close, when investment sizes are 15% larger than average and investment volumes are 50% larger. Starting a raise in May is the second best choice. While the average investment sizes are identical to December, investment pace is much slower, hence a bit riskier.

*Originally posted on Tomasz’s personal blog, here.


San Francisco is the new center of start up activity

tl;dr Start up fund raising is growing across the board and increasing in geographic diversity.

The new year presents a great opportunity to look back at trends in start up activity and fund raising. Pitchbook has created a great summarization of start up funding and valuations from 2013.



We can see that aggregate dollars raised in 2013 is up slightly from 2012 but both down for the peak in 2011. Let’s examine geographic dispersion to better understand and visualize the broader trend. Thanks to Crunchbase’sopen platform and Google’s geocoding, there is raw data to analyze.

The Maps


Venture capital raised in 2013, by US Region.


As also reported by CB Insights, California dominates start up funding with New York and Boston closely following. This static view of the distribution of fund raising activity doesn’t present the full picture. Inspired by a beautiful Foursquare visualization, where check ins represent the heart beat of a city, I’ve taken a pass at representing the heart beat of the US economy through start up activity.

Some GIFs


Venture capital raised, by US Region. Each frame is one year.


In the animated gif above, each frame is the funding data for one year starting in 2005 through 2013. 2005, investment is mainly centered in NY, SF and Boston with little start up activity in more diverse areas. These central tech hubs see large gains over the 8 years covered. As the decade progresses, the start up community begin to also grow in Dallas, Austin, Raleigh NC, Seattle, DC, Chicago and beyond. Some of these secondary cities have now grown beyond where NY and Boston started. In a network effect business, which I would argue includes technology start ups, solving the chicken and egg problem is the most difficult. Now that technology start ups have passed a reasonable scale in the top 10 US cities, we can hope that the pace of innovation will continue to accelerate. There are benefits to starting a company in Silicon Valley but the success of some large unicorns outside of the bay area proves it is not a requirement (this is a big enough topic to warrant an independent post).


Venture capital raised, by CA City. Each frame is one year.


In the more mature tech market of California, the center of gravity has moved north from silicon valley to San Francisco proper. In the mid aughts, Silicon Valley and San Jose represented the majority of the Bay Area investment. As the teens turn, San Francisco became the focal point for start up activity. We also see that LA and San Diego have continued to grow in terms of raw investment dollars. I expect with the recent sale of a company like Gravity1 that the LA area will continue to see new investment.


Venture capital raised, in SF Bay Area. Each frame is one year.


Zooming into the Bay Area, we see that aggregate dollars raised is increasing and spreading out. The center of gravity has become San Francisco. The bay area, like the US and CA, sees increased geographic distribution. There has been increased investment beyond Palo Alto distributed across communities like Berkeley, Mountain View, Fremont and San Jose.


Start up fund raising is growing across the board and increasing in geographic diversity. In new start up empire, all roads lead to San Francisco.

Code and Data

Here is a R script for generating graphs and see Crunchbase for the raw data.

  1. Redpoint portfolio company. 


How much to raise using Crunchbase data

tl;dr Raise enough to hit the next milestones to raise the next round plus a healthy buffer.

After starting PrimaTable in 2011, raising a seed round and selling to Hotel Tonight and now sitting on the other side of the table, I’ve been asked: How much venture funding to raise? Fund raising can be more art than science with a competitive process that dictates terms. In this post, let’s take a quick look at the data.

It’s been written about before:

  1. Tom Tunguz “Raise enough money to achieve a set of milestones that will attract a subsequent round of investment from new investors.”
  2. Mark Andreessen1 “Raise as much as you can without giving away control of your company, and without being insane. … In a normal scenario, raising more money rather than less usually makes sense, since you are buying yourself insurance against both internal and external potential bad events.”

Thanks to Crunchbase, there is data to examine.

The Numbers

Cumulative Distribution of First Round Amount.


Looking at the distribution of the first round amount (restricting to only recent companies first funded between 2010 and 2013), the median is $1.2M with few extreme outliers. While this itself is informative, further funding history is a layer deeper. Raising a subsequent round can be viewed as a signal for some success, as other people have done (Benn @Mode).2

Follow On by First Funding Amount.


There seems to a positive correlation between the initial fund raise amount and the likelihood of a follow on round. Note, this is clearly an instance of correlation and not necessarily causal. The companies that raise more money may be more likely to succeed or better able to raise a subsequent round of funding. There are likely exogenous factors that influence follow on rounds. However, the relationship makes sense. In order for it to also reflect the collective wisdom on fund raising, the relationship should demonstrate a very specific activity based on individual company dynamics.

Follow On by Category.


Decomposing fund raising impact further by category, the positive correlation in general persists but varies dramatically by category. In many categories, Biotech, Cleantech, and Software, there is no explicit relationship. Companies that have raised larger initial rounds haven’t converted to a second round at a greater rate. In the case of Mobile, the relationship is the inverse. Mobile companies that have raised more are in general less likely to raise a follow on round.

Median First Raised Amount


In addition to differing relationships with additional funds, the aggregate amounts across category varies. The median amount raised when conditioning on follow on rounds varies wildly. In ecommerce, the median first round for companies that raised a further round is $1.3M vs just $.5M for companies that haven’t.


This data can help substantiate the thesis that you should raise enough to hit the next milestones and mitigate risk with a sufficient buffer to weather the inevitable bumps. Valuation for startups only increases on mitigation of risk. For example, in some seed stage companies, this can mean mitigating technical risk and building a working product. For a series A, this might mean finding product market fit and showing user engagement. For later stages, this can be demonstrating a scalable distribution channel and business model. Regardless of the stage, future rounds will necessitate mitigation of risk.

Mark argues to mitigate financing market risk by raising at the max of the range that is possible. This make sense to protect against macro changes that change fund raising dynamics as well as gives as much fuel to hold on as long as possible. This can be detrimental because this ups the level of risk that must be mitigated prior to a subsequent round.

In practice, for most early stage startups, the increments of funding look to be 12 – 18 months of runway. The round size varies with the team but you can use a rule of thumb of $125K / developer / year ($100K salary + 25% expenses).

In summary, raise enough to hit the next milestones plus a healthy buffer.

Code and Data

Here is a R script for generating graphs and see Crunchbase for the raw data.

  1. As quoted on Venture Hacks from an old post subject to link rot
  2. This neglects bootstrapped companies, companies that reach profitability or companies that sell, not requiring subsequent financing.