Don’t Invest in Rolling Funds

Recently, rolling venture funds have become the next big thing™. (See an overview of them here). Their unique ability to fundraise in public, means that information about (and solicitations to invest in) them have blanketed Twitter and news sites.

They make for flashy pitches. This first wave have generally been started by individuals with large followings, and track records of angel investing into now unicorn-sized startups. The GPs of these funds talk about democratizing being an LP and strong financial returns. To quote Sahil Lavingia, one of the most prominent rolling fund creators, rolling funds prevent “disproportionate benefits for those who already have their foot in the door.”

Rolling funds really do make it easier to be an LP. A first-time LP could find out about the fund on Twitter, click through to the one-page description on AngelList, and set up quarterly ACH payments to commit. The minimum commitments are (compared to traditional venture funds) small, spread out, and shorter.  These new benefits gloss over a more important question, “If you are a marginally-qualifying accredited investor, should you be an LP?”  If you couldn’t tell from this post’s title, the answer is no. Specifically, rolling funds will likely provide poor financial returns, without the side-benefits of direct angel investing.

Venture as a general category has poor returns.  Of 2,254 funds, 68.9% had returned less than 2x, after ten years! VC funds, much like the startups they invest in, are highly outlier-driven. Outside of the poor returns, LPs normally spend over a decade with no access to liquidity on their investment. For comparison, investing into the S&P 500 ten years ago, would have been a 2.85x return, and you can access liquidity at any time. Access to liquidity is greatly underrated. Shit happens, and if you suddenly need access to cash, investments in a rolling fund won’t help.

Angel investing is also a poor financial choice, for basically the same reasons. It is highly outlier driven, and as a non-famous angel, you are unlikely to have access to the (on paper) most desirable companies. All that said, if you can stomach the financial risk, direct angel investments have a lot of side benefits. You get to meet some great founders, learn about a range of industries and companies, and get the occasional warm, fuzzy feeling that you are helping some of those founders succeed. More over, you are building a track record. Putting your money on the line is a very concrete way to prove what companies you believe in, and that, sometimes, those beliefs are ahead of their time and right.

Funnily enough, it is this exact track record that enables rolling fund creators to raise money. Sahil regularly mentions his angel investments into Figma and HelloSign. Cindy Bi lists three unicorn investments in her twitter bio. If you invest in Cindy’s fund, and that fund invests in more unicorns, that doesn’t help your track record. You can’t list those investments, because you did not make them. Angel investing is not a good risk-adjusted way to invest money. Neither are NYC rent, craps, oyster happy hours, or many other things that I spend more than financially optimal. However, I still find angel investing to be worth it. It’s enjoyable and aligns long term with my personal goals and love for early stage companies.

I can’t see close to the same benefits — financially, personally, or professionally — in becoming a rolling fund LP. They combine the poor returns of venture, with the aloofness of investing in public equities. I have no doubt that they can make sense in a larger portfolio full of existing venture investments. However, rolling funds claim to be a great opportunity for a new class of people. That claim is, frankly, bullshit.

Thoughts on API Design and Compliance

I got a chance to do an interview with Pandium about my thoughts on GraphQL, productizing an API, and HIPAA-compliance. They did a good job framing the questions in a way that should still be understandable to non-technical people.

I left GraphQL out of my 2020 predictions, since it is pretty technical, but I am very confident that GraphQL will become the de-facto way to build APIs. The benefits are humungous, and the surrounding community and libraries are quickly catching up to (and surpassing) what exists around REST. We’ve been using it since early 2018, and it’s transformed how we are able to build things at Healthie.

Check out the interview here

Hurdles to Being a Solo-Creator

In all the discussion of of people leaving large media companies to go solo, I have not seen anyone address how creators (and subscribers) are going to handle life events that prevent creators from producing content for prolonged periods of time.

Money Stuff’s Matt Levine just went on paternity leave, and is taking a break from writing the column. I assume this doesn’t present any career or financial issues for him. He’s a Bloomberg employee, and he can take paternity leave knowing that his salary is secure, and that his job is waiting for him. If Matt Levine was a solo-creator on Substack, how would he be able to handle this? His column is based on current events, so he would not be able to write a backlog of them, and schedule their release. Would monthly subscribers be OK with paying the monthly subscription fee for a newsletter they aren’t actually going to get?

In other freelance fields, like web development, it is possible to take on extra work in compressed period of time (“Feast Period”) in order to bank money for when you are unable to work (“Famine Period”). This doesn’t work for newsletters. If you normally send a daily column of 5,000 words, you can’t just write twice as much, and expect your subscribers to pay double that month. Subscribers are price-sensitive, and doing extra work likely won’t lead to higher average revenue per subscriber. Even if a solo-creator decided to just not charge their subscribers while they were out, they would be at a serious disadvantage when they returned. For topical newsletters, the archives have minimal interest, and would likely not draw in any new subscribers. Meanwhile, regular churn (due to both real reasons, and delinquent cards) would continue to pile up.

If Pete Wells (another of my favorite columnists) took two months off, it would have no bearing on me continuing my New York Times subscription. The Times has a lot of content creators (even within the Food section), and still delivers a lot of value. For creators going it alone, the freedom can become a liability in cases like these. Life events happen, and they need to be accounted for when we talk about financial shifts (like the rise of solo-creators and Substack). One obvious solution is for solo-creators to band together (like what Everything Bundle is doing), but then we eventually end up right where we started.

If you know how people are planning to handle this, let me know on Twitter!

Creating Independent Board Seats. Don’t Do It

My company, Healthie, has a board. It’s made up of myself, my cofounder, and the VC who led our seed round.

We do a board meeting every quarter, and it’s invaluable. It gives my cofounder and I a chance to break out of the day-to-day grind, and do long term strategic thinking. Our third board member brings a wealth of experience and insights, sanity checks us, and pushes us In a way that is motivating and not overbearing. The non-founder board seat comes with the ability to approve or veto certain things, but it has never once been an issue, or forced us to do something that my cofounder and I disagreed with.

As that dry background shows, I have had nothing but positive experience with our board, and boards in general. That said, I vehemently disagree with the recent advice I’ve seen saying that founders of early stage companies should voluntarily create boards, and that those boards should have independent board members. This advice is bad for founders because it introduces new risks outside the founder’s direct control, without giving the founder any additional upside.

At the founding of a company, the founders have 100% of the risk and 100% of the upside. When founders reduce the share of upside they have (by diluting themselves), they do so to either reduce their risk, or to increase the absolute value and/or liquidity of their upside. This is true when you are raising a round, giving employees or advisors equity, or even choosing to sell.

When founders create a board as a condition of fundraising round (like we did), it can make sense. We traded some control of our company, and share of the upside, for money and help that we were confident would lead to a greater overall absolute return for us. The board seat that we created and gave away was part of the control that we sold, and we got something back for it. When founders create board seats, just because they read that it’s a good idea, they gain nothing and lose both control and upside. Common arguments for creating and expanding boards don’t address or explain these negative repercussions on founders.

One major pro-board argument is that independent boards ensure better corporate governance and better company performance. I disagree.

Early-stage founders should be, both in self-confidence and actuality, the best people to make final decisions for their company. The unique mix of insights, passion, and drive that lead founders to start a company, also means that good founders are uniquely positioned to ultimately make the right calls. Founders should seek advice, review data, and listen. However, impeding the founders’ ability to make hard and unpopular decisions jeopardizes what makes early stage companies able to punch above their weight.

In addition, a board preventing bad governance or behavior, means that the founders initiated (or at least condoned) bad behavior. If you’re a founder who does that, why are you reading this? If you aren’t, why get get dragged down by unrelated people’s bad behavior?

Following this advice does not only lead to you immediately giving up control of the board, it also leads to further equity dilution! Fred Wilson’s advice is to to give every independent director 25 basis points per year. If you appoint the recommended two, and have them for four years, that’s two percentage points (more if you raise money in between) that you’ve lost, for the pleasure of having less control over your company. In my experience, the people who give the best advice either do it for free because they like helping, or because they have invested in the company. It is very possible to get more advice then you need, without resorting to further dilution.

The second argument is that large boards, with diverse board nominees, increase diversity. That is fundamentally true, but is a cop out/founder-unfriendly way to solve the very important problem of diversity in tech. The most heralded (and most committed) way to increase diversity, is to hire or wire. If VCs know diverse people who’d be amazing board members, why not invest in them? If you’re a founder, why not go all out trying to hire them? If you don’t know them well enough to do either of the above, how can you be comfortable with them representing your best interest on your board?

When we got our initial term sheet, I (as the CTO) was not included on the board, and we would have had an independent board member instead. That was the area of the term sheet where we put our foot down the most. As a founder, I can’t imagine not being in the room when some of our company’s largest decisions were made. Whether you hate them or respect them, there’s a clear precedent of successful founders, like Zuck, maintaining board control and building large companies. Don’t give in to pressure or online advice to give up board control.

Have comments, or disagree? Let me know on Twitter

TL;DR Watch the below video, and replace “Teen Suicide” with “Creating Independent Board Seats”

 

The Horizon Never Gets Closer

This week, I read Alex West’s post about chasing goals when building a company. Here are some of my thoughts/anecdotes on chasing goals when building a company.

I started Healthie because I was bored.  The season for the sport I was playing ended, and I was suddenly faced with hours a day of new free time. I met an MBA candidate with an interesting idea, and figured “why not?”.

At first, it was all great. My only goal was to build an MVP. Each new piece of functionality added, and the time invested doing so, felt like a huge win. There was no status quo, no real plans, and no targets. I was just replacing my HBO-watching time with something a little more productive.

We launched that MVP, and the first few customers we got felt absolutely amazing. My co-founder and I were still in school, and had no expectations. Every dollar we brought in infinitely exceeded what I had planned.

Things got more serious and stressful when we raised some money, and started Techstars. Suddenly, there were expectations and growth goals. I had taken a leave of absence from Penn. We had employees who we had convinced to leave stable jobs to join our nascent company.

Initially, we consistently hit our monthly goals, and it felt good. Not amazing, but good, like we were doing what what was expected. Then, we didn’t, and it felt awful, bone-crushingly awful.

The pain and stress of missing a goal was ten times worse than the joy of achieving one. I had gotten addicted to the growth, and each new milestone quickly became the status quo. Each best month in company history became my new standard, and any performance below that was untenable.

I started (literally) dreaming about customer cancellation emails and product issues. I would physically vibrate at my desk at work from tensing up so much. On an academic level, I understood there would be swings in our business. Mentally, I couldn’t wrap my head around it.

It’s been five years since we started. I’d love to say that I feel like we’re closer than ever to achieving our end goals. I don’t. The more we grow, the more I see opportunities to take us even further. The horizon never gets closer.

When people ask me about Healthie, I say we’re bigger than I ever thought we’d be, and smaller than I want. That answer is never going to change. What has changed is that I now appreciate the journey.

Our company performance varies. We have good and bad quarters, achieved and missed goals. Throughout all that, I enjoy what we do and who I work with. That’s been a much more helpful, consistent, and happier motivation than a sole focus on chasing ever-moving goal posts.

Learning To Code Is Not A Solution To Your Problems

Fred Wilson published a post today recommending the recently unemployed to “Learn to Code If You’ve Lost Your Job”.

Learning to code is great, but the way Fred presents it as a solution for unemployed people is wrong.

Even pre-Corona, the job market has been absolutely flooded with juniors developers, even in traditionally hard-to-hire markets like NYC. AngelList is full of incredibly raw bootcamp grads, who are still months ahead of people just starting on Codecademy

With Corona, there are also a lot less dev jobs available. Most of the companies still hiring are huge companies like Google and Facebook. These companies have strict requirements, drawn out interview processes, and are less willing to take chances on atypical backgrounds.

More people should learn to code, and it really does help with generally applicable problem solving skills, but please don’t rely on it as a quick solution to financial or career issues.

I Spent my 8th Grade Summer Getting Scammed

Freelance Web Development sounds sexy. You set your own hours and rates, can work from anywhere, and get to do a rotating set of interesting projects.

Breaking in as a freelancer is not easy. I was a 14-year old WordPress “developer”, long on time and naivety, and short on everything else. Armed with these skills, I spent my eighth grade summer getting ripped off.

The Backstory: In middle school, I built websites for my nine year-old brother and a twelve year-old rapper (who would go on to great fame), and tried to figure out how phishing sites worked. With that experience, a logo I got for free on Reddit, and a domain name, I started to try win some freelance gigs.

The First Clients: Every day, I would go through the freelance, forhire, and favors subreddits. If the post involved web design or WordPress, I’d reach out. From the reach out, I got two initial gigs. Both did not pay in cash, but I figured the initial testimonials (and other rewards) would be worth it. 

I built both websites, confirmed that they met their standards, delivered the code, and then politely asked for that testimonial (and in one case, a retro soccer jersey) that had been promised.

And… 

Crickets. 

Both of these “clients” couldn’t even bother to write a couple of nice sentences, and went ghost right after I delivered the code. 

This experience gave me a healthy dose of cynicism, and looking back on it eight and a half years later, some major takeaways.

#1 – People value you at the cost you impose on them. 

Ideally, this would have been solved by charging them for my time. Since that wasn’t really possible, I should have at least ensured that they had spent sufficient time thinking through these projects on their end. These “clients” spent very little of their own time and gave me  short broad asks, that clearly had not been planned or thought through. I then spent a bunch of my time taking that, and turning it into a functional website. 

At Healthie, one of the best things we did was start charging for the product almost as soon as it launched. In cases where it’s not possible to get paid, whether for your product or your time, ensure the other party is at least investing similar time on their end. For example, if you’re thinking about doing an unpaid internship, make sure you’ll have a mentor at the company who will spend her valuable time with you. 

#2 –  Be mindful of your leverage.

In both these cases, I sent over all the code and access before I had received my end of the bargain. Once I had sent that, I lost all leverage here. These internet strangers had no reason to fulfill their end. I assume these strangers didn’t have malicious intentions, but life gets busy. Since they had all they needed from me, it wasn’t a priority to compensate me.

I never made this mistake again freelancing. At Healthie, where it’s not just my time, but our team member’s as well, I remain very mindful of this.

#3 –  Follow, Follow, Follow Up. 

Digging up these old emails, I was amazed that I didn’t follow up more with these “clients”. I sent one softly-worded request, and left it at that. Now that I’m older and wiser, I decided to change that. 8+ years later.

Should You Learn to Code?

We’re at an interesting crossroads in the Learn to Code movement. On one hand, we have huge endeavors like the CS4All initiative, which wants “all NYC public school students (to) learn computer science”. On the other,  the rise of No Code tools like Webflow and AirTable allow non-programmers to build fully-featured applications, tools, and websites without any coding knowledge.

Is programming the new literacy? Or is all the money, energy, and legislative time spent on this a huge waste? Most pro arguments here boil down to economic data, high developer salaries, and access to opportunity in an quickly changing world, versus the con side’s fears of commoditization and developer over-saturation. At the end of the day, the numbers here don’t matter, and I’ll spare you the specifics, because they miss the point.

You should learn to program, because programming fundamentally changes the way you’ll approach and attack problems, regardless of the domain. Whether it’s a web application, a complicated travel plan, or a full-blown life crisis, programming will help you effectively tackle problems big and small.

Why? At it’s core, programming is about taking problems, breaking them down into smaller pieces, and continuing to break them down until they become solvable. Take the Fitbit integration we built as part of Healthie (the startup I work on). It’s a relatively complex feature that involves regularly authenticating with Fitbit, pulling down a wide range of data, and converting it into the data format we use at Healthie. This feature, when broken down, goes from a wide-ranging description into a couple dozen simple functions, that someone with a few days of programming experience could probably understand.

This is a calming way to go through life. The same way of thinking suddenly makes big, scary problems simpler and more approachable. What used to sound overwhelming and impossible becomes an organized path with individual steps and clear progress points.

The hard (and valuable) part of programming isn’t the art of writing lines of code, but the art of breaking problems down. Writing lines of code is one way to do it, but likely not the only way, and maybe not even the best way. As this becomes clearer to educators and legislators, I expect teaching basic HTML to go away, in favor of a more wholistic focus on problem solving, even if it doesn’t involve a computer.

Thoughts or Comments? Let me know on Twitter

If you’d like to get my future posts, right in your inbox, please subscribe at this link

 

 

 

 

 

 

Finishing Fights

Sports idioms are everywhere, startup world included. We talk about home runs ideas, ball park estimates, and (nonsensically) below par quarters.

These analagoies permeate casual conversation, board meetings, and all-hands, but they are (pun intended) off-base.

Building a start up isn’t a ball game, it’s a fight sport. What’s the difference?

1) There’s more than one way to win

In football, baseball, soccer, basketball, etc, there is exactly one way to win. You score more points than your opponent. When my (un)loveable loser Jets are down 20 points, their only hope of winning is to score 21. Hell, even quidditch, the sport from Harry Potter — a fantasy book full of mythical creatures, flying brooms, and literal magic —  follows this. The Golden Snitch just gives the catching team 150 points. If 150 points isn’t enough to close the deficit? Tough luck, your team loses.

Deficits exist in start ups and fight sports as well, and should not be ignored. Periods of poor, lagging performance can leave founders, boxers, and wrestlers behind their competition. If not corrected, time can run out, those deficits can become final, and the competition can be decided on points, just like all the sports mentioned above. However, unlike those aforementioned sports, those deficits can be short circuited.

No matter the score, no matter the time remaining, fighters and founders are never more than one punch, pinning combination, or well executed idea from breaking through and short circuiting, not closing, the deficit. It doesn’t matter how bad things have been, and how far behind you are. Until the final bell is rung, or the company is closed, there’s a fighting chance.

2) There are no substitutions

When football games become routs, both sides, winning and losing, pull their main players and ease up. The stars sit on the bench, the clock runs down, and the coaches prepare to shake hands. When a basketball player gets injured, he comes out, a sub goes in, and the game continues. In start ups and fight sports, this concept doesn’t exist. If a fighter gets injured and can’t continue, the fight is over. If the founders burn out and quit, it’s (effectively) a death sentence for the company.

Just like a UFC fighter needs to fight smart and not gas out too early, founders need to be smart about their physical and mental energy as well. To repeat a cliche I used to vehemently not believe, it’s a marathon, not a sprint. Finish the fight.

Agree or disagree? Let me know via twitter or email or elsewhere.

Want to get automatic updates? You can join my newsletter here

Predictions for 2020

Since I love being wrong, here are my predictions for 2020

Bitcoin continues to flounder

Over a decade since creation, Bitcoin has failed to establish itself as anything else besides an instrument of speculation. Much like the year of Linux on desktop, the Bitcoin as currency movement isn’t happening, and the HODLers, most scarily, appear to have accepted that. Without the idea of Bitcoin as currency,  the price of Bitcoin will continue to be volatile, but overall deflate, as HODLers move on to newer and shinier altcoins. Markets can stay irrational longer than the participants can stay solvent, and if the participants are irrational as well, that can’t end well. I’m bullish about the blockchain as a technology, but Bitcoin will end up as the Sega Dreamcast of crypto.

Prediction: Bitcoin slowly declines, spending the back-half of the year under $5,000

Trump impeachment trial goes along party lines, Trump remains president

As the country gets more and more tired of impeachment discussion, Pelosi will have to deliver impeachment articles prior to the main stretch of the general election. McConnell and the Senate Republicans will run a show trial that introduces no new info, and will call none of the relevant witnesses. The vote will go along party lines, and Trump will remain as president (and tweet about being vindicated shortly after)

Prediction: Trump remains president, partisan polarization continues to increase.

Shitty companies will continue to stink, as market regains sense of smell

You can go up on a steep hill and look West, and with the right kind of eyes you can almost see the high-water mark—that place where the wave of insane Softbank fueled valuations finally broke and rolled back. The collective group delusion is done, and 2020 will be the real year of reckoning, for WeWork, Uber, Lyft, Peleton, and other money burners. With no one else to pass the bag too, we’ll see these companies scale back from their grand ambitions and turn into more traditional, smaller, less AI gobbledy-gook versions of themselves. On the other hand, at-scale companies with strong unit economcs (e.g the Zooms of the world), will continue to see their stock price rise.

Prediction: WeWork continues to shed everything besides office space in already established cities, Zoom gets back up to $90/share, Uber ends under $20/share, Lyft ends under $35/share.

Snap continues to show up it’s haters.

It’s been a banner year for the Snap (and it’s stock price), and there’s no reason why it should stop. The worst has happened, Facebook did it’s best to copy all of Snap’s offerings, and Snap stays bloody but unbowed. What’s left is a company with a sticky userbase that has realized it’s not going to be bigger than Facebook, and has set out to build a really solid (dare I say profitable) company, that is finding it’s own (very large) niche.

Prediction: Snap ends the year at over $22/share.

Highly Contested Dem Primary leads to Biden nomination, Trump victory

An overly crowded Democratic field —  with low policy differentiation between candidates, refusals to withdraw, and Michael Bloomberg’s Super Tuesday strategy, leads to a drawn out primary slate, ending up with a battered moderate Dem nominee in Joe Biden. Sanders and Warren will get into each others way until it’s too late, Buttigeg will flame out nationally after good showings in the early states, Bloomberg will be too tough a pill to swallow, and the rest of the cast (Yang, Steyer, Gabbard, etc) won’t ever be serious contenders, thus leaving Joe Biden as the parties nominee. In the general election, Biden’s lack of true fans and unexciting campaigning won’t do enough to affect voter turnout, and Trump’s rabid base will give Trump a victory in 2020, likely while losing the popular vote again.

Prediction:  Trump beats Biden, loses popular vote.

Everyone on Twitter agrees that there are multiple valid ways to grow a company and that VC is neither the solution nor the devil

(just kidding)