Generative design

One of the areas in technology that I am most excited by is generative design.

We have a portfolio company in the space called Creative.ai. Their newsletter is very much worth subscribing to. 

While Creative.ai's current focus is helping humans to augment creative work, at a more abstract level generative design allows you to specify the parameters of the design problem you're looking to solve. And then the computer takes those parameters and proposes one or more optimal solutions based on historical data and generated models, cost, physical constraints, etc. 

If you look around the world, the materials we use and the shapes we are familiar with are geometric, because those are easy to think about, easy to manufacture and easy to work with - for humans. But what was true in a mass-produced world doesn't hold in one where generative design and 3D-printing are pervasive. 

The resulting designs and objects are much more organic but also eerily alien. This is where evolution is taking us, and it's incredibly interesting.

The opportunity in Live streaming

We're seeing tons of activity in live online streaming in the US, but less so in Europe. That's a pity.

HQ Trivia, backed by Founders Fund, has brought an interactive gameshow format to live streaming on the phone. ShopShops, backed by USV, is bringing influencer-intermediated boutique retail to the world. NTWRK, backed by Jimmy Iovine (he of the amazing The Defiant Ones documentary), is another take on live shopping. Of course the "rethinking QVC" approach gets a ton of backing. 

All of this isn't that new. I remember looking at TinyChat around nine years ago, who were doing multi-person streaming and trying to build discussion rooms and communities. As with lots of live streaming (remember Chatroulette), these things tend to go camgirl-naughty fairly quickly. LiveJasmin is a phenomental business (if you're OK with that morally). 

And then of course there's Justin.tv, out of which grew Twitch. We've seen more and more people try to compete with Twitch. That's a really tough proposition. In terms of broad one-to-many streaming, the large players have sucked most of the air out of the room. Facebook, YouTube, Twitch, even Twitter, are just too good, the built-in audience is too big, the network effects too strong. 

Hence you need a deep vertical approach, preferably one with a native monetization model (see QVC above). In general I think there are two ways to go about this: one is communication; the other is entertainment.

In communication you need to go full (and free) utility. See if you can build a better FaceTime or Google Meet (or whatever it's called now). That's really tough in consumer. I know Confrere in Oslo, backed by Point Nine, are doing this using WebRTC for B2B. That sounds like the right take. 

The other approach is entertainment. And what's interesting there is that we've had over 90 years of experience in television. These one-to-many formats have had a long time to mature with TV as a platform. So the opportunity is really to take what works in television into niches that could be viable with a global audience. It's the old benefit of the internet: what's impossible given a limited domestic audience and advertising market could be a great business online.

If it's shopping, then it needs to wow by location, presenter, merchandise. ShopShops is onto something big here, but there's a lot of room for it to be much better than what they are doing today. It also felt very focused on Asia when I last looked at it.

I also think there's more room in gameshows, news, sports, reality. HQ Trivia is like the lottery version of that - high dopamine, big prizes, but ultimately a bit lackluster as a long-term format. And this stuff is _all_ about retention. 

We'd love to see more takes on what consumers will be passionate about in live streaming from the phone. 





What we look for in DTC brands

While working through the opportunity and talking to many companies over the last year, we have developed a fairly constrained framework that attempts to maximize the opportunity of a large, i.e. “category-defining” outcome.

In general, we prefer a highly differentiated product in a market with historically low product innovation. We believe pricing power is a desirable goal and prefer high margins. In this context, we believe high barriers to entry can be a significant moat (limited suppliers, complex manufacturing, regulations).

We like products that are easy to ship. We prefer zero-sum markets, where we effectively remove the consumer from the market post-purchase. Usage data should improve the product or service over time. We like consumables, especially ones that embed themselves into daily routines. We like timeless products that are not subject to rapidly changing trends.

In general, a high frequency/repeat purchase rate is good. It not only lends itself to subscription, but can support medium AOVs. Higher AOVs, while good for profitability on first transaction, frequently imply a low repeat rate. We like verticals that exhibit high brand loyalty, which also serves to drive repeat.

We prefer verticals where there are a small number of highly consolidated incumbents that dominate the market. We like when prices are too high. We prefer when the incumbents have limited direct relationships with their end customer and are overly dependent on mass broadcast advertising and offline retail relationships.

Similar considerations apply to team as in typical venture investments, but ideally a DNVB founding team covers several key skills. We believe the core ones to be ecommerce, performance marketing, branding, community/social, design/production, and logistics/supply chain. We strive to back teams that exhibit deep consumer insight and have a true “authentic connection” to their chosen vertical. We look for a higher purpose than mere financial motivation.

From a perspective of brand positioning, successful DNVBs are often “mass aspirational”, meaning they market a fundamentally mass market product as an attainable luxury good. The goal is to make the end customer feel empowered, smart (value for money), cared for, and encourage an almost tribal sense of belonging. An Instagrammable look and feel and seven-star customer service completes the package by encouraging the consumer to share experiences.

Allocating capital to a venture portfolio over time

Over the last five years, we have developed a more sophisticated capital allocation framework. Most credit here is due to Yacine Ghalim, who runs our Paris office and joined us from Goldman Sachs five years ago. 

Capital allocation over time is a major driver of venture fund return. Sure, ours is a business of power laws and the most important driver of return is initial company selection. But making sure that capital is funneled to the best companies over time, ignoring the false positives of fast external fundraising or traction that doesn't really show product-market fit, is a major driver of absolute return at the end of the fund lifetime.

So here's how we do it. 

Portfolio company developments are continually shared on their own #portfolio-[company-name] Slack channel. Every quarter we run a portfolio review day, where we update each other of each company's latest status and revised outlook.

Following these updates, each member of the investment team estimates an expected enterprise value at exit for each portfolio company. This EV at exit is a multiplier of probabilities and set enterprise value ranges: write-off, under €100 million, €100 - 250 million, €250 - 500 million, over €500 million. 

The input sheet per fund (on Google spreadsheets) looks like this:

The expected enterprise values at exit are aggregated across team members for the entire fund and result in several analyses.

The first is the quarterly change in percentage and value terms - seeing which companies are moving up and down. 

The second is the relative ranking of the companies in our portfolio, which displays as a candlestick chart showing the average EV at exit as well as the range across team members. This variance also shows you the amount of consensus in the team. As the fund matures, the relative ranking displays the well-known power law of venture outcomes. 

The third analysis looks at "buckets" of companies according to several criteria: absolute EV at exit, probability of getting above €250 million, probability of write-off, time in portfolio. 

It also looks at the marginal investment opportunity - what the last valuation was, what a current valuation would be, and what the likely cash-on-cash multiple of additional capital deployed into the company would be. 

This estimated multiple is the money shot: it provides  clear guidance on where capital should be deployed and directly informs the investment manager's approach on whether we should do our pro rata, whether we should try to invest more than pro rata, or whether we should attempt to limit additional cash into the business. 

Like most of venture, capital allocation is not a purely rational endeavor. But having a joint framework and common understanding of what makes a good investment decision over time helps the team more quickly make the right decision.

Our job is to take risk - and the best way to do this is by being permissive in the team. Each initial investment decision does not require unanimity in our partnership. We encourage everyone to follow their gut (though we do keep each other intellectually honest). 

However, our follow-on decisions do require unanimity - and this is where the capital allocation framework is a great help in fully aligning the whole team behind the best companies. 

Two things destroy companies

"Two things destroy companies... mediocrity, and making it all about yourself." - Halt and Catch Fire, ep. 3

If you haven't watched HCF, it's worth it. I think it's on Amazon Prime in many territories.* 

This is a great quote. There's lots of things out there about what kills companies. Running out of money. Being too early. Not making something people want. 

But the core, really, is not being good at what you do and ego. Both can be fixed. But in conjunction, they're 100% fatal. 

Remember: truth, love, and growth. This is exactly what that means. To make it about the thing, and to be excellent at making the thing. 

*How I hate geofencing. It's one of the most pure expressions of the hold the "real" world now has over cyberspace. It's time the pendulum swung the over way again. 

Contemplative thoughts upon returning from Israel

Love never fails. 

But where there are prophecies, they will cease; 
where there are tongues, they will be stilled; 
where there is knowledge, it will pass away. 

For we know in part and we prophesy in part, 
but when completeness comes, what is in part disappears... 

For now we see only a reflection as in a mirror; 
then we shall see face to face. 
Now I know in part;
then I shall know fully,
even as I am fully known.

- Corinthians 13:8-10, 12 (NIV)

This is the passage in my weekly email from Contemplative Outreach (Fr. Thomas Keating, O.C.S.O.), the organization that's bringing back contemplative prayer, a Christian version of meditation that finds its roots in the practices of the Desert Fathers.

Israel was an emotional trip for me, particularly the stunning Basilica of the Annunciation in Nazareth. The cult of Mary is one of the few redeeming features of Catholicism (not the Church) for me: the archetype of the God-mother in all its facets of Earth - fertility, birth, nurture, care, comfort, sacrifice, sorrow, fortitude... I use the pietà as a Twitter background image for that reason. 

In the imagery around Mary contributed by countries all over the world lies an otherworldly quietude and grace. It points towards the yearning for the indwelling spirit, the program for divine transformation we each carry within us. This is the Secret of Secrets: the mother's love for her child.

Below is the image contributed by Japan:


The other overwhelmingly emotional bit was our visit to the Western Wall. Rare are the places where you can sense such deep loss and sorrow. It opens the heart. 

The flip-side of Israel was the elevation of holy sites to a sort of Golden Calf idolatry. It echoes the worst of Christian literalism, the narrative overshadowing an ill-interpreted message. 

But I also sense that my contempt for the Church following the Pennsylvania report and the ongoing gay/trad civil war may blind me to the deep connection that others feel to Christians throughout the ages. 

And so we touched the Stone Handprint in Jerusalem (Via Dolorosa V) and felt a bit of that, too. 

Image result for jerusalem cross franciscans


Making board meetings fun

I have two companies in my portfolio that make every board meeting an incredible experience. They shall remain nameless, but they know who they are.

They both raised money from international  investors early. They are both headquartered in European cities that are not London or Paris. 

When I started first attending these boards, I was shocked. They would have an elaborate board dinner at a top restaurant (and I mean Michelin star, internationally renowned chef, Pellegrino list type). They would add a cultural event, like attending the symphony, or touring a city, or visiting an art gallery. We'd meet the conductor, the mayor, the artist.

They would vary locations across the European continent where there was some connection  (perhaps an investor was based there, or a supplier). 

I estimated each board meeting cost tens of thousands of dollars given travel expenses and extracurricular activities. Sure, these companies had raised millions of dollars in venture capital. But it still seemed like a massive expense relative to really frugal operations.

I've come to believe operating boards like this is 100% worth it for a few reasons:

First, no one misses these boards. No VC, no independent director. No one dials in. They are awesome, fun experiences. People travel for a day or more to attend. 

Second, the social cohesion on these boards is the very best in my portfolio. People have gotten to know each other over time and there's trust and empathy and a willingness to go the extra mile. Folks are consistently available for management and the board members for each other. It's invaluable when things get tough. 

Third, attending the board meetings for management other than the founders of the companies is a great experience. They get significant exposure to the board in a social setting and the trips serve as a big incentive. 

Fourth, they make what can be a tough trip for VCs and NEDs (flying in for a few hours, turning around and flying out) into a social event that lasts ~1.5 days. People start to spend real time with the company. They start to grok the business and its challenges at a more detailed level. The depth of discussion increases.

Overall, I wish more of my companies took this approach. Turning boards into a social event beyond a few hours spent in a room together pays off over time. 

The next iteration of co-working spaces

My colleague Deepka has a new post about the next iteration of co-working spaces on Medium right now.

Her thesis is that co-working will verticalize and deepen, with new propositions possible/probable in beauty (more flexible salons), DTC brands (pop-up retail space with warehouse/logistics + office), and gastronomy (shared virtual kitchens using Deliveroo/UberEats as front-end). We've already seen a few of the latter and invested e.g. in Taster, a take on the virtual kitchen. 

The independent contractor/freelance economy is growing very rapidly. A partial cause is lower transaction costs of the internet and hence the smaller theoretically optimal size of the firm (see Coase). Another is people preferring a more flexible work life that isn't tied to a specific location. 

The gig economy isn't just Uber drivers. I've seen projections where freelance work will become a majority of employment in developed economies. That's a major opportunity for many people, especially educated folks around the globe. But it also poses challenges in pensions, benefits, and social cohesion. 

It will be interesting to see what happens when the trade war moves to services. 

Travelperk raises $44 million Series C financing led by Kinnevik

A bit over two years ago I got on a plane to Barcelona to meet with Avi Meir and Javier Suarez, the founders of Travelperk

Travelperk had started building a business travel platform, allowing companies to manage trips from booking to accounting all in one place - using the world's largest global travel inventory to take the pain out of business travel. 

The company was the perfect case for our thesis of consumerization of business services, bringing the benefits of the consumer internet to small and medium sized companies. 

It helped that phenomenal people were surrounding the business from day one: Johannes Reck, the CEO of our portfolio company  GetYourGuide, introduced us to the founders and was considering joining the board. Alex Finkelstein, GP at Spark Capital and our co-investor in GetYourGuide, was leading the Series A. The original seed investment was led by Robin Klein at Localglobe. Phillip Riese had started advising the founders and would become a potential board member, as had Fritz Demopoulos, the founder of Qunar. 

Since we closed the Series A in the summer of 2016, Travelperk has built out its product and added hundreds of customers. The business closed a Series B co-led by our friends at Felix Capital and Target Global in 2017 to support its rapid growth. But good things happen fast when companies are doing well. 

And so today Travelperk is announcing a $44M Series C led by Kinnevik, one of Europe's most prominent growth investors, with co-investment from Yuri Milner, Tom Stafford, and existing investors Sunstone Felix, Target, Spark, Localglobe, and Amplo.

There are lots of VCs who are hesitant about celebrating funding rounds. And it's true that financing is not always a necessary and never a sufficient condition for building a great company. 

But the path to building a phenomenal business is long and it's good to take these moments to congratulate the team on the hard work that has gone into putting the company on a phenomenal trajectory. And so it's a good day for Travelperk. 

We are spending the day visiting Jerusalem with the founders, team members, and board, before sitting down to the board meeting tomorrow. Good times. 

Working on my sleep

I recently got the new Fitbit Charge 3, which has been replacing Patek on my wrist. That’s bad news for the Swiss watch industry. 

Besides the heart rate tracking, which is cool, my main use case is to improve my sleep patterns. I rarely get enough sleep and it’s starting to wear on me. 

Here are my sleep patterns from the past week: 



Notwithstanding the jet lag that comes from the Israel trip, that’s clearly not enough. I’ve set it up so that the tracker reminds me to head to bed at 10pm latest. 

We’ll see how that goes.