2016 VC Half-Thoughts: Seed Companies Aren’t Being Overfunded, They’re Being Prematurely Funded

In the spirit of clearing out some half-formed thoughts, I wanted to share my response to venture LPs who ask about the effect of “so much money” flowing into the seed space and how we think about its competitive impact upon our fund Homebrew.

[Context regarding our fund’s strategy is important here: we make 8-10 seed stage investments per year, where we’re leading or co-leading the round and stepping up to serve as a committed, proactive investor for the company in the years to come, typically serving on a Board or otherwise spending time ongoing with the founders until, say, Series B. Most of our investments are listed on our website.]

So, we actually don’t think most new funds or seed investors are competitive with Homebrew in the traditional sense of “we’re all looking at the same deals and need to ‘win’ deals away from new funds.” There’s a few reasons for this, namely, there are actually very few funds like ours – seed stage, focused on leading rounds, committing sweat and reputation ongoing alongside capital, started by ex-operators, institutional LPs so we have follow-on capital, in-firm resources to coach on recruiting strategies and tight with the next round of VC partners.

Given this, from a performance standpoint we focus on communicating who were are to talented founders, exceeding their expectations post-investment and continuing to improve our ability to identify the most promising (and best fit for us) opportunities from the talented teams we meet each day, knowing that we’re still going to be wrong a lot of the time. So, yes, we have “competition” but our business seems to be less about competition at the “winning the deal” stage and more at “seeing the deal.” That is, how do we ensure Homebrew is among a small group of firms that talented founders seek to connect with early in, or prior to, their fundraising process.

At the same time, there is a marketplace impact from the swelling of capital looking to invest in seed stage companies, even if most of that capital is in sub-$500k checks and non-lead positions — ie their models aren’t directly competitive with ours. And it’s probably one of the biggest differences in seed investing when you compare 2006 vs 2016. Namely, seed stage companies clear the market faster, even when they’re very early in their development. 

This is especially – perhaps even maddeningly true – when the teams pattern match – ie engineers from top companies or degrees from top schools. These startups are still very often able to raise $1m-$2m on a relatively unproven hypothesis with little accomplished. When there was less money at the middle and bottom of the pyramid for seed, these teams might have raised a smaller friends & family or angel round to work for 3-9 months on their companies. During this time top seed funds would get a chance to see their businesses evolve and build relationships with the team. That’s not to say there ten years ago there weren’t plenty of rounds raised on a demo but those tended to be truly exceptional founders or displays of technology prowess, not “hey, we’re two engineers who worked at XYZ company together and we have an idea.”

So what’s the impact upon Homebrew? Well, when we see a company at this stage we assume their round is going to get done. Maybe not from the quality of investors they were originally seeking or the terms they proposed, but we can’t always get a second bite at the apple. The resultant hastening of the market funding means that, at time, we have to decide whether we want to pay seed prices for pre-seed risk. We tend to pass on these types of companies when the technology is undifferentiated or the value proposition to the potential customer is unclear. Or there’s not a missionary zeal from the founding team as to why they’re pursuing this goal.

Basically the tldr regarding all the new money in seed is that we don’t feel like it impacts our strategy overall but it does pull some percentage of companies out of market prematurely that we otherwise would have gotten the chance to see evolve. The additional cash isn’t leading to overfunding so much as it is premature funding.


2016 VC Half-Thoughts: The Industry Has Shifted Back to Investing in Technology, Not Business Models

2016 VC Half-Thoughts: The Industry Has Shifted Back to Investing in Technology, Not Business Models

In the spirit of clearing out some half-formed thoughts, one takeaway is that we’re again in a period where the most exciting investment opportunities possess real technology risk. When we started Homebrew in 2013 our industry was, in retrospect, probably midway through a cycle where innovations were in business model rather than underlying technology. Observing a handful of companies, such as Uber, Airbnb, Warby Parker, founders were taking those models and trying to apply them to other verticals. Sure there were underlying technology changes driving the success of these companies (mobile app penetration for example), but many of the successor startups relied upon applying a business model innovation to a seemingly new vertical.

Labor marketplaces for a variety of professions. A ton of private label vertical commerce. On-demand and excess capacity apps. Monthly subscription in a box. And so on.

There were – and are – many valuable companies still growing from this time period but the venture investor wasn’t making a primarily technical bet. They were making an assumption about the team, business model and operating excellence. We found a handful of investments in these areas in our first fund, trying to be early or contrarian rather than just throw money after a bunch of “Uber for X,” but also struggled to find conviction in the most crowded of spaces (such as food delivery).

When I look at our dealflow (and investments) over the past 18 months I see an interesting shift back towards more technology risk than business model emulation. Think of what venture investors have been pursuing lately: AI, autonomy, VR, computer vision, bioscience, agriculture, material sciences. These startups are VERY difficult to start without having one or more key technical members on the founding team. And you can’t diligence many of these companies just using assumptions in a spreadsheet. There’s real possibility that the company’s technology doesn’t pan out – or at least not at the level of sophistication where it can become a durable competitive advantage.

Some investors believe it’s going to be difficult to invest in these areas without an advanced degree in the discipline. Satya and I disagree, and to date, have made several investments in very talented teams (well, not in VR — we’re currently sitting that out). But it’s exciting to see the seed stage impact of “software enabling the world” (our version of “software eating the world”) and how a fund like Homebrew can assist.

Socialism & Autonomy: A Self-Driving Car Shouldn’t Be Allowed To Make Left Turns Across Traffic.

What if autonomous cars weren’t allowed to optimize for the driver but instead were more broadly cognizant of the cost any single action imposed on society at large? Would self-driving cars be selfish or unselfish, optimizing for the efficiency of the driver or the impact upon cars and pedestrians?  While this question doesn’t have quite the same moral imperative of the trolley problem, it’s a consideration that algorithms and regulations should consider.

For example, should a self-driving car be allowed to make a left turn across traffic, an action which typically delays everyone behind the car, versus making a series of right turns in order to cross a busy street? In a future where EVERY car is autonomous, they would merely communicate to one another and be able to switch lanes at high speed to move away from the obstruction, but we’re due for many years of hybrid traffic on the road.

If there are a dozen people waiting to cross the street and only a single car approaching the intersection, should the car need to wait while the pedestrian mass is allowed to move forward? A three minute delay to one driver trumps a 36 cumulative minute delay to the group, right?

Has anyone seen good studies on what an optimized traffic system looks like and what the associated decisions trees look like if you’re maximizing efficiency?


“Regardless of who is in the White House, we have work to do from the bottom up and from the outside in” – Five Questions With Jennifer Pahlka of Code for America

Like many of history’s historic friendships, Jennifer Pahlka and I met first on Twitter. I had heard of Code for America but wasn’t exactly sure what they did or how they did it. So Jen invited me over to their SOMA HQ for coffee and afterwards, well, my head spun because the scope of their work as a nonprofit is so massive. But Jen has been one of the leaders over this past decade of finding ways to marry technology and government. I made sure to donate some money to C4A post November election and Jennifer was kind enough to let me pepper her with some questions. Enjoy! And you can help. Donate, come work at Code for America, work in government, or join your local Code for America Brigade!

Hunter Walk: You’re founder and Exec Director of Code for America, which helps government services embrace the Digital Age. When was C4A founded and how is it structured – a non-profit with both employees and volunteers?

Jennifer Pahlka: I founded CfA in 2009, the year after everyone said that “the Internet had gotten Barack Obama elected President.” I wanted to know if that same Internet could not just make a difference in who gets elected, but actually make government work better for regular people. I’d been working on a conference called Gov 2.0 that explored that idea primarily with federal government when a friend reached out asking for help making apps for cities. Back then, very few people in the consumer technology world thought about working for or in government. I decided to start a service year program to let the people in the tech industry dip their toes in the waters of government.  It turned out that a lot of talented developers, designers, product managers, and entrepreneurs found that really compelling, and now hundreds of them have jumped in and started swimming. They’ve come in through Code for America, through the Presidential Innovation Fellows (which was modeled after Code for America), through 18F and USDS (which I helped start as White House Deputy CTO) and through dozens and dozens of states and cities who have opened up jobs and recruited tech talent in to work in an updated framework, one in which they can have enormous positive impact on the lives of people in our country.

At CfA, we have a medium-sized full time staff, most of whom are focused on making the $470B taxpayers spend nationally on programs for vulnerable people in our society work a lot better.  These teams are working on projects like streamlining the access to food assistance and job re-training or reducing unnecessary incarceration by making it easier to avoid bench warrants or to comply with probation. A few of whom work on organizing the 75 volunteer groups in communities around the country that work on these issues locally and spreading adoption of Code for America’s practices among government broadly.

We are a non-profit, blessed with an amazing board of directors that includes John Lilly (Greylock Partners and former CEO of Mozilla), Shona Brown (former SVP at Google), Stacy Donohue (Omidyar Network), Brett Goldstein (first person to hold the title Chief Data Officer in a US city) and, the father of the Gov 2.0 movement, Tim O’Reilly (who also happens to be my husband.)

More than a non-profit, though, we are a network.

More than a non-profit, though, we are a network.  We have enormous impact through the projects we take to scale that make our country’s safety net work better.  But our hundreds of alumni — the fellows, employees, and government partners who have worked with us over the past six years — are a truly powerful movement that’s growing incredibly fast. When we started, making government work better through technology and design was barely a thing people talked about. Today, if you go into most major cities, you’ll find people who’ve worked with us, or hired our alumni, or attended our conferences who are spreading the gospel and showing these practices working and getting outcomes. Right here in San Francisco, for example, CfA alumna Ashley Meyers is redesigning how the lottery for affordable housing works so that families in need aren’t doubly burdened by a system that’s nearly impossible to navigate. There are thousands of her around the country applying the best practices of metaphysical Silicon Valley to the services that the American public needs from its government today, and that’s another kind of scale.

HW: Talk a little bit about project selection. Do you work with a government entity at the local, state or national level? Do they need to “sponsor” a project? What’s involved?

JP: When we first started, we worked on pretty much any issue or area where we thought our particular approach could be helpful in local government. If an iterative, user-centered, data-driven approach could help get results, and we could get support for it (financial and otherwise), we’d do it by recruiting a team of fellows to tackle it through their year of service with us. That’s been amazing, but it’s ended up starting a lot of projects with enormous promise, and without the means to carry all of them through.  Seven of them have become startups that have spun out as successful companies. Many of them have been carried on by the cities, counties or states we partnered with.  And a few of them have spread naturally from one city or county to another.

One example comes from our first year of the Code for America fellowship program. The City of Boston had changed the public school selection policy to promote more kids walking to school. The process was now essentially a mapping problem; the primary factor was the distance from your home to various schools, with some other factors such as whether the child had a sibling in another school or special education needs. But to communicate this change to parents, the Department was still sending a 28 page printed brochure with descriptions of each school written in seven point type. The brochure contained a lot of information, but it did not — and could not — tell you which school YOUR kid could attend. Parents were beyond frustrated.


One of our fellows, Joel Mahoney, took on the job of making a website that let you enter your address, the age of your child, and the school(s) any siblings attended, and returned a map showing the schools your child could go to. He had part time help from one or two of the other fellows. It took him about ten weeks to launch a beta version of the site, which worked remarkably well, and was, as another fellow said, “simple, beautiful, and easy to use.”


Our city partners told us that if his project had gone through a regular procurement process, it would have taken at least two years and cost at least $2M.

Our city partners told us that if his project had gone through a regular procurement process, it would have taken at least two years and cost at least $2M.

That became a model for our projects for the next five years, and the project itself is now part of one of Code for America’s spinoff companies, OpenCounter. OpenCounter was founded when one of our government partners, Peter Koht, who worked for the City of Santa Cruz, teamed up with Joel, the creator of DiscoverBPS, to offer yet another CfA project to other cities.

But most projects don’t spin out. Our focus today is on spreading some of the successful projects nationally, especially those that aren’t easily spread by for-profit startups. Our work in food assistance, for example, has incredible national relevance, both in reducing the burden on users and reducing the costs of administering the program for governments and taxpayers. But the sustainability revenue we’ve found for that work is only available to non-profits, and it’s best done by an organization with both a long-term view and a hybrid model that combines venture philanthropy with government funding. We’re the natural ones to take this work first to all of California and then to the rest of the country, so we’re raising the funds we need to do that. It’s the most leveraged philanthropy you can imagine, because it unlocks government funds that help feed hungry families all year long, just by overcoming administrative hurdles.

HW: What’s a memorable C4A project that you never would have expected to come across when you first started?

JP: Imagine all the frustration that you feel at, say, the DMV. Now imagine what’s at stake isn’t your car registration, but your eligibility for a job or housing that you desperately need.

In 2014, California voters passed Prop 47, which allows people with low-level, non-violent felony convictions on their record to remove them or reduce them to misdemeanors. That’s important because if you have a felony on your record you can’t get a job, you can’t get student loans, and you can’t get housing. You are basically stuck in a cycle of poverty. This isn’t good for Californians and it isn’t good for taxpayers.

Prop 47 passed because voters wanted to help people out of the cycle of poverty. But only about seven percent of people who are eligible have even started the process and very few people have gotten through. Why? Because the process is not online. You have to go to a legal clinic between 9-11am on Tuesday; you have to find the right piece of paper, fill it out, and answer a lot of confusing questions. Next, you have to find your rap sheet—which is so hard to read that you have to take a class to learn how— and then, certify that your rap sheet is correct. After that, you file more paperwork and play the waiting game. The process takes about a year.

It doesn’t have to be this way. Our team built Clear My Record Clear, a simple way to submit your application online in just 10 minutes. Then we follow our users via text message, documenting the barriers they face, and working closely with our government partners to fix operational bottlenecks.

When I first started, I believed that if we passed a law that was good for people, the people would benefit. Now I know that that’s just the beginning.  Californians voted for Prop 47, but we can’t just leave it at voting. We have to get involved with the implementation of the law if we’re going to actually see it make a difference.

HW: C4A is a non-partisan organization but the tech industry tends to lean progressive and some would even suggest aspects of GOP platform, such as denying climate change, are “anti-science.” What does “non-partisan” mean in terms of how you approach your work and who is a good “client” for your services?

JP: CfA has always been fiercely non-partisan. It’s true we’ve worked with a lot of democratic mayors; it’s also true that I took a year leave of absence to work in the Obama White House. But we’ve also worked with Republican mayors and have been visited and praised by Eric Cantor, Newt Gingrich, Darrell Issa and other Republicans.  We are grounded in two beliefs that don’t belong to any party.  First, that government is what we do together, and it’s fundamentally possible for government to work effectively towards common goals in the 21st century.  Second, that it isn’t working as well as it should, in part because it’s so grounded in 19th and 20th century modes, and that it needs updating. Who is going to do that updating?  We the people have to do it. That’s built into the history of our nation. It’s not someone else’s job. It’s ours.

I have a sticker on my laptop, from a friend who works for the Defense Digital Service.  It says “No one is coming. It is up to us.”  It’s pretty clear it is in fact up to us.

Our work is a bit like those old ads for Miller Lite, where two people fight about whether the beer is great because it tastes great, or because it’s less filling.  You might love our work because it helps regular people interact with government with dignity, or you might love it because it’s orders of magnitude cheaper to build technology the way we do it. Fight all you want. Pretty much everyone agrees on this one. A good partner for us is someone who cares about one or both of those outcomes.  That’s a lot of people.


HW: So, this whole election thing. How do you anticipate the outcome impacting C4A in the coming years? In terms of government interest in technology adoption? Or funding and staffing of C4A?

JP: Yes, this whole election thing. So many feels. So little clarity.

Fundamentally, I think we have less visibility into the future than ever.  I’ve taken some solace in the commitment of the vast majority of the federal workforce to the ideals of our nation and to working to serve the American people, regardless of who is in the Oval Office (or Trump Tower, as the case may be.)  I wrote about that here and here.

But I won’t pretend I’m not scared by our President Elect’s rhetoric and his cabinet nominations.

But I won’t pretend I’m not scared by our President Elect’s rhetoric and his cabinet nominations. Both show a disregard for portions of our society that were already vulnerable and for the role of government itself as a force for good. It’s possible, however, that with folks like Peter Thiel working on the transition, that the administration could accelerate changes to the government technology ecosystem, especially the regulations that govern it. There’s no doubt that more change is necessary but there are a lot of questions about what direction that change will take and who it will benefit.

In terms of funding and staffing, we’re growing! Regardless of who is in the White House, we have work to do from the bottom up and from the outside in. That’s what we’ve been doing for the last six years, and that strategy is more important than ever now, in lots of contexts. We can keep helping government work better in cities, counties and states around the country. And we’d better.

Thanks Jen!

You can help. Donate, come work at Code for America, work in government, or join your local Code for America Brigade!


What’s Alcohol Good For? Staying Married.

Winding down towards year end and catching up on the “Study of Studies” section from Atlantic Magazine. It’s a column each issue where they summarize recent academic research on a particular topic. This summer it was about alcohol, called “The Science of Beer Goggles.” Here are some of my favorite blurbs:

  • A version of the classic “trolley problem”—would you push a man in front of a train, killing him in order to save five track workers?—they found that the drunker people got, the more likely they were to say they’d push the man.
  • In one study, researchers offered people 20 euros and gave them a chance to donate some or all of the money to Doctors Without Borders.  Compared with sober subjects, those who’d downed an alcoholic peach drink were significantly less likely to donate.
  • Our appreciation for others also increases after a drink or two. Participants in one study who imbibed a fruity vodka drink found minimally to moderately attractive faces significantly more beautiful than did those who’d consumed an alcohol-free drink. This might be because  we’re less able to distinguish symmetrical faces from asymmetrical ones when we’ve been drinking, and symmetry is known to be an important component of attractiveness.
  • One forthcoming study found that  unhappy couples got along better and were more able to solve conflicts after a few vodka-cranberry drinks.

“Life Isn’t Fair, So Why Do We Behave As If It Is?”

Raising a kid in the Bay Area you hear a lot about “socioemotional development,” which focuses on helping a child understand and convey their emotions. In my household right now this means a lot of discussion about “fairness,” usually motivated by my daughter perceiving that she, or someone else, is getting the short end of the stick.

The Atlantic, which every issue does these great thematic round-ups of academic research, recently published “Life Isn’t Fair.” Here are two of my favorite blurbs from the post:

  • People who frequently patronize a business believe they are more likely than other customers to win a given prize drawing by that business—a phenomenon the researchers called the “lucky loyalty” effect
  • Researchers found that when people felt powerless, they were more likely to say that race, class, and gender disparities were justified [hw note: oh hello Election 2016!]

2016’s Best Startup Podcast Was….

While allowing that the next 16 days might produce a podcast which blows my mind, I’m ready to pick my favorite startup podcast of the year. “What’s Love Got To Do With It” brought three longtime friends together – Jerry Colonna, Fred Wilson and Brad Feld. What’s great about the interplay between the trio is they’ve all worked together in different contexts over a 20+ year period, so there’s no bullshit.  Also, while feeling they are very familiar with one another, they don’t hold static views of who they are. Jerry, Fred and Brad know they have evolved as individuals over their time together and will continue growing and changing.

The discussion itself revolves around being a supportive investor and venture partnership, especially guiding principles and board dynamics. You really should listen to the hour but here are some choice quotes:

“When you are in a group dynamic and you approach things selfishly, it is very hard to be effective.” – Fred Wilson

“There are a lot of investors who walk into the room, take their seat as an investor, and they are scared shitless.” – Jerry Colonna

“If you are going to lose the money that’s fine, but don’t fuck up the company in the process.” – Fred Wilson

“Brutal honesty delivered kindly.” – Brad Feld

…and one exchange I especially loved ->


Thanks Jerry, Brad and Fred for giving back to the tech community and being so generous with me and Satya ongoing.