Invert. Always invert

I am currently making my way through Poor Charlie’s Almanack - the unofficial biography of Charlie Munger.

There is a lot of gold so far, and I’ll probably write more about things from the book in the weeks to come. Today, I wanted to cover one topic which is so simple and fundamental it’s surprising not more people talk about it. Inversion.

According to Munger, it is most often wise to invert when faced with a challenge. As he says, “tell me where I will die so I shall never go there”. Basically, instead of focusing what to do spend your energy on what not to do.

Most people think about how to win, but if you’re playing long-term games (which he does, and which startups definitely are), the key is not to lose. Given enough time, those focused on winning will make mistakes and end up losing. If you focus on not losing for long enough you’ll end up winning.

While this is not 100% applicable to startups (sometimes you have take a risk and focus on upside maximization in order to get to the next level), I see lots of areas where it makes sense to apply this mental model.

One thing that immediately comes to mind is controlling burn rate and runway. Startups die (lose) because founders give up or the company runs out of money. If you’re obsessed with your runway, you’re probably less likely to scale prematurely. The inverted perspective reduces the chances you’ll go bankrupt.

Bringing an inverted perspective along when making decisions can add a lot of value. I will definitely try to do it more going forward.

Out of office

I’m not working today. If you send me an email, you’ll get this autoreply:

Jeg svarer ikke på epost i dag, da jeg klimastreiker.

Vi har ikke mange årene på å gjøre noe med den menneskeskapte klimakrisen. Ting går for sakte, i år brukte vi opp jordens ressurser for i år før utgangen av juli.

Jeg tror vi lenger fremme vil snakke om “de som var på feil og de som var på riktig side av historien”. Og jeg er overbevist om at det i så tilfelle vil være de unge som streiker i dag som vil være på riktig side. Det å ikke gjøre noe er i økende grad å plassere seg på feil side av historien, og derfor streiker jeg også i dag.

Hvis du også bryr deg om denne saken, er noe av det absolutt viktigste du faktisk kan gjøre å bruke stemmen din ved valget søndag 9. september og alle valg deretter.

Jeg kommer tilbake til deg over helgen. Ring meg om det haster.


I’m not replying to emails today, as I’m on strike for climate.

We don’t have many years left to do something about the man-made climate crisis. Things are moving to slow; this year we spent this years resources before end of July.

I believe we’ll talk about those who were “on the right or wrong side of history” in the future. And I’m convinced that if that’s the case, it’ll be those on strike today who’ll be on the right side. Not doing anything is increasingly the same as placing yourself on the wrong side of history, and that’s why I’m on strike today.

If you care about this issue, probably the most important thing you can do is to vote at the upcoming elections this Sunday - and every election thereafter.

I’ll get back to you with a reply over the weekend. Call me if it’s urgent.


Good intentions

Today, I’m thinking about intentions. Related to my post a few weeks back on bounded rationality, as well as this quote from WSJ’s Charlie Munger interview from a few months ago:

I don’t believe people are born evil. Further I don’t believe most people try to do evil. Still, a lot of actions done by people end up doing more bad than good.

I believe this has to do with bounded rationality. People believe they do good - their intentions are good, but the outcome is not good - because their view of the world is imperfect, flawed.

I find this whole concept extremely fascinating. We have people all over the world disagreeing with and reacting to other people’s actions as if others have bad, or at least different intentions.

What we really should spend our energy on is understanding in what ways their worldview is bounded or different - then educate them and ourselves to reach a common understanding of the world. At that point we can start discussing actions. Jumping straight to actions is not constructive in any way.

What got you here, won't take you there

I’ve been thinking a lot about p2p marketplaces lately. The following are merely anecdotal observations and does not necessarily reflect the bigger picture, but putting this out there to see if others agree or not.

My current hypothesis goes like this: “it is only possible to run a p2p marketplace at scale in the short term. Adverse selection will remove the good actors in the long term, and the best way to cope with this is either to own supply or partner with professionals”.

Some background: Uber, Airbnb, Getaround and others have grown massively the past 10 years. They’ve done this similar to most startups, by targeting a part of the market that incumbents don’t want to touch (because it’s cumbersome, difficult with old tech etc). Coincidentally, using freelancers was also the most cost-efficient way to start and scale, and thus possible to grow from nothing to a lot in a decade.

My question is, as the title says, will what got them to where they are today get them to where they want to go? What is next? Some observations:

My experience with using Airbnb the past 10 years is that more and more listings are posted by professionals. Hotels, hostels etc are they can get more bookings by putting rooms up on Airbnb, similar to Deliberately or not, Airbnb’s supply has been professionalized.

My experience using car sharing platforms is that I prefer seamless experience where I don’t have to coordinate with another person. I’ve been on the other side (renting out my car) for the past few months, but too much hassle (accidents etc) means I’m probably quitting. On the supply side, I’m (a “peer”) churning. On the demand-side, I prefer to buy from someone owning supply.

Reading about Uber’s difficulties with driver retention (or, how they don’t disclose any information about this) is another data point. I can see them partnering with taxi companies (professionals) or offering more employment-like agreements (guaranteed salary, benefits etc) as a way to address (potential) driver churn issues.

These are only anecdotal observations, as mentioned initially. I could be completely off. But currently I’m more bear than bull on p2p marketplaces’ plans to continue scaling as before.

Parkinson's Law

Last week I learned about Parkinson’s Law for the first time (I first read about it here). Since then I’ve been using it all over the place, as I believe it’s very broadly applicable.

At one instance, I was discussing a product deadline with a startup. Their deadline was 4 months into the future, and my objection was that would probably not be the wisest thing to do. Startups are about learning per dollar spent, and if you’re not able to ship (learn) often enough chances are you’ll fail.

Whatever you’re going to do will take 4 months if you give it four months to complete. It could be that the activity actually takes 4 months to complete regardless of how you divide it up, but it’s an unnecessary risk to take. Plan for shorter periods, which is the essence of Scrum, is a good idea. And Parkinson’s law is part of the explanation why it’s so popular.

Parkinson’s law can also partly explain why large organizations tend to grow fat and lazy. When you have limited resources, you can’t afford to hire for “somewhat necessary” roles. If it’s not super urgent and important, you’re not doing it. As an organization grow, you will have less control of how extensive any job is. If a job needs to be done, you can afford to hire someone to do the job, and so you do it. The risk you run over time is if the “job to be done” could be done in less time. Parkinson’s law states that that won’t happen, rather the person(s) involved will overcomplicate the job so that it fills the time available. This can be super contagious over time, and kill productivity, morale and so one - and can help explain why large organizations are slow.

I love this mental model/universal law, and will continue to look for places to use it going forward.

StartupLab Accelerator

Did some digging in the archives today, and found this post from 4 years ago called “are accelerators really the solution”. While it’s always somewhat embarrassing to read old writings (as it should be), revisiting this is especially so - since we launched our first accelerator program less than a year later.

This month we’re accepting applications to our seventh batch, meaning we see accelerators as part of the solution. What happened? A few things:

  • We realized accelerator programs appealed to certain founders that we had difficulties reaching otherwise. People “stuck” with regular jobs and mortages in need of funding to really take the step and become entrepreneurs. An application deadline creates urgency in a good way.

  • We realized there are certain advantages to funding companies in batches. You can streamline the selection process and also the paperwork, meaning entrepreneurs can receive funding quicker.

  • We realized there are certain advantages to working with startups in batches. Or really - startyps benefit from being part of a batch. Having relationships to a few founders in roughly the same phase - a peer group - can be very helpful when you’re stuck. It can normalize problems and take the edge off. And sometimes someone else has already solved a similar problem.

  • We realized cannibalization wasn’t something to be worried about. If starting an accelerator meant we would do less ad-hoc investments through the year, that wouldn’t be negative. The objective is to fund and support top entrepreneurs, regardless of how it’s done.

We’ve been doing it for a few of years now, and the results are gradually emerging. Becoming a overnight success in startupland takes 10 years, but companies like Imerso, Confrere, Fundingpartner, Oivi, Age Labs, Defigo and Diffia are gradually getting there. Also, the program has improved significantly, and continues to do so.

If you’re looking for funding and early-stage support for your startup, I would strongly recommend you submit an application for our fall batch. The application deadline is September 1st, and you can read more and apply by clicking here.

Bounded Rationality

Back to blogging after a 4 week hiatus aka summer vacation. I’ve spent the past weeks relaxing but also reading, and one of the books that made the biggest impression on me was Complexity. It was recommended by Bill Gurley in this podcast, and is about the story of the Santa Fe institute and their research.

This is not supposed to be a book review. But the one thing that really stuck with me after reading the book, is how important the concept ofBounded Rationality is, and I’ll try to explain it here.

The Santa Fe institute was unknown to me before I read the book. In short, it’s an interdisciplinary institute founded in the 1980’s focusing on complex adaptive systems. Where a lot of research focuses on order and equilibrium and throw in as many assumptions as necessary to have something orderly to investigate, the Santa Fe Institute focuses on real-life scenarios with emergent structures that never reach any equilibrium (which is how the world operates generally).

Classic economics, the kind I was taught at school, operates under the mantra that “all models are wrong, but some are useful”, and try to simplify the world to the point where their models disconnect from reality. While many such models continue to be useful, it’s astonishing how many things they fail to predict.

Research on complexity takes a different approach to this. Instead of simplifying the world, complexity research tries to understand the full picture.

Early on, the Santa Fe Institute understood that the only way to go at complexity is to look at the individual agents it consists of. Agents can be anything from organizations to molecular cells, but here I’ll talk about agents as people - as that is what I mostly relate to.

If you understand how every agent act all the time, you’ll be able to make very very good predictions.

Most regular economics assumes that all agents are rational, and subsequently tries to make predictions. And all such predictions fail because people are not rational. This is where it get’s interesting. Are people not rational?

If you ask me, I’ll definitely say I’m rational. I am agreeable to reason. Still, a lot of things I do would probably be considered irrational. What is happening at these moments? Am I mentally disconnecting at these moments?

The concept of Bounded Rationality gives an alternative explanation.

If I do something others consider irrational, I don’t believe it’s because I disconnect mentally. It is just that the information I have available when making certain decision is limited (read: bounded).

It could be that I don’t understand all the data available to make the decision. It could be that I don’t understand what my best alternative actually is. Or a combination of the two. I still make decisions - either because I’m not able to gather more information, or the decision is irrelevant enough that I don’t bother collecting more information.

I believe this explains why people act more rational as the stakes are higher. With higher stakes, you spend more resources gathering information, and as a result your bounded rational decision is less bounded.

More generally, the outcome is that many of my decisions will appear rational within my bounded environment, but still be irrational. And every prediction that is made assuming my rationality will be wrong.

This is very interesting. And has implications for how we make preditions about the world.

Build companies, not acquisition targets

In the recent months we’ve seen three Norwegian startups acquired by larger companies. Dragonbox and Poio were acquired by a local company (Kahoot), while Nabobil, was bought by US-based GetAround.

While the Kahoot-acquisitions were met with applause, not surprisingly people react to the Nabobil-acquisition saying it’s a shame ownership leaves the country (likes in this article). I weighed in on the matter on Twitter last night:

Every company is a special case, and it’s difficult to make an outside-assessment of what the right choice for each shareholder is. In many cases founders are hungry to keep building, but facing an incumbent with a “buy or build” decision it’s understandable that they decide to sell instead of taking on bigger competitors in a fight that might lead to bankruptcy. Or the offer might just be that good. Or something completely different. My point is, we shouldn’t go at specific companies for choosing to sell.

That said, I think there’s a proper discussion to be had around how to build Norwegian companies, not acquisition-targets. I think we can all agree it would be beneficial if we had more locally owned large tech companies. The question is how to get there.

Incentives rule the world, and I believe this is the way to go at improving things. And incentives can be both monetary and non-monetary.

Obviously, the more valuable a company is the better for the founders. But at a certain point, when their stock is worth a life-changing amount of money, the mindset can shift from offense to defense. And the most comfortable next step is then to seek an acquisition. It is common to allow founders to sell some shares along the way to mitigate this, but that in itself is not enough.

Funds can play a part in this. Bigger funds need bigger outcomes, and generally speaking an IPO (independent company) is a bigger outcome than an acquisition. But the only Norwegian fund (Northzone) that went bigger suddenly shifted their focus away from Norway, so incentives work many ways. We could do something like this on the regulatory side,

But just fixing the monetary incentives probably won’t fix it. There’s only so much money can buy. There are soft aspects that needs to be addressed as well. Founders need to want to build something “enduring”. This “want” can be nurtured, both through having a culture that embrace this attitude and appetite, and through having role models that both show it’s possible and set a benchmark that others might want to beat. Probably a lot of other things I haven’t thought of as well.

We can all agree on the importance of more companies being built and owned locally. But let’s not go at specific companies choosing what they believe is their best alternative - rather let’s discuss the systemic part of things. But first, let’s congratulate the Nabobil team and owners on a great outcome!

Sourcing dealflow

I recently did a talk on dealflow; what it is and how to think about it in early stage investing. Here are the key points from that talk. It is mostly written with startup investors in mind, but hopefully it’s valuable to those on the other side of the table as well.

What and why:

If you’ve made it to my blog you probably already know what dealflow is, so this’ll be short: Dealflow is a term used for the quantity and quality of investment offers an investor receives. The term is mostly used when talking about privately held companies, as public companies are available for everyone. Privately held companies are not, and thus having better dealflow can be a competitive advantage.

You could say there are four stages of startup investing (I first heard this in a podcast with Brian Singerman of Founders Fund), and those are:

  1. Finding companies
  2. Picking companies
  3. Accessing companies (aka “winning deals”)
  4. Helping companies grow into winners

Investing in private companies is just as much about finding and accessing companies as it is about picking and helping. I’d argue it’s impossible to succeed consistently if you don’t master finding and accessing, and I’ll refer to this as dealflow.

There are two extreme ways to think about dealflow. Does the founder pick the investor, or the other way around? These two perspectives should not be considered a dichotomy, rather as end points on a spectrum. We’ve done good investments in companies where nobody else had conviction, and we’ve done good investments in companies where we had to fight to get access. Most investors do both, but having awareness around that there are different is helpful when prioritizing how to work with dealflow.

Inbound deal sourcing

On one end of the spectrum; founders pick their investors. Put differently, the founders who eventually end up building the winners are so obviously good they can choose between every investor out there. If this is your prevailing belief, everything you do should be about making yourself as attractive as possible to make sure you meet these founders - that you have proprietary dealflow.The two questions that should guide your work in this scenario is “why should I be contacted?”, and “how will they find me”. Let’s start with why:

First of all, you have to match with the company. All investors have preferences with regards to investment amount, valuation range, geography and/or sectors, and these preferences exclude some percentage of the total available dealflow. If your filter is too narrow, you end up without any dealflow. Opposite, you become the jack-of-all-trades and nobody comes to you first. Find the balance somewhere in between (and break rules).

Most often, there are many investors with similar preferences, meaning you have to add something beyond capital to be approached. This is normally any of the following:

Competence: experience and advice of benefit to the company, either generic “how to build a company-perspectives” or knowledge related to the specific company’s technology or market.

Network: knowing people (customers, bigger investors, advisors, potential employees etc) that can help the company.

Brand: “we are backed by the founder/investor behind [famous company]” can get a company a long way with regards to recruiting, future fundraising and so on.

Fundraising takes time, and startups can’t afford to search for the perfect investors forever. Put differently, it’s not enough to be the best match for any given company - you also have to be on their radar. The “how will they find me” part:

Doing this means marketing yourself, work that can be done both online and offline. Online it means blogging, newsletters, and being available for interviews and comments when journalists report on topics that you want to match with. Offline it means being at events, participating in panels, hosting meetups and dinners and so on. The whole point is to tell the world you’re looking for startups, and doing it so frequently you’re top of mind when the companies you want to fund start fundraising.

Outbound deal sourcing:

The opposite perspective is that it’s only in hindsight we know who the good founders were - good investment opportunities are accessible for everyone as long as you find them. And there are many supporting arguments: the best returns comes when you’re both right and non-consensus, and the most groundbreaking companies start out looking like toys. And there are plenty of examples of this: the story of Airbnb’s first 1000 days is one of my favorites, another is Robinhood’s 75 rejections.

Outbound sourcing can be done in multiple ways, and both online and offline. Offline means being at events, as well as reaching out to your network directly. Online means browsing relevant social news sites (in Norway: FB, Linkedin, Shifter, DN etc), looking through lists of companies that receive soft funding (Innovation Norway, BIA, etc), and so on. And you can subscribe to my newsletter to get the Norwegian tech stories I find most interesting :)

Doing outbound activities has its benefits. It’s a quick and direct way to approach companies you find interesting without spending time broadcasting your preferences first, and you can very quickly assess a high volume of relevant companies. Timing might be off - you might contact companies when they’re not fundraising - but that’s not necessarily a bad thing. The negative side of this strategy is that you’re most likely not seeing any proprietary dealflow. If you believe the best founders pick investors, they probably don’t see a lot of reason to showcase what they’re building broad enough that you discover them soon enough.

To sum up these few paragraphs, there are lots of aspects to think about when it comes to dealflow. Probably the most important thing is to be aware of the four stages of startup investing, and having a hypothesis around which (ones) you’ll work at improving. I’m continuously iterating as I learn more, and I don’t believe there’s one best way to have good dealflow. There are many ways to win.

Co-founders, part 2

Following up last week’s post on co-founders, and I’ll keep it short this week as well. Let’s say you’ve decided it’s best for your business if you bring on a cofounder, how do you go about finding/selecting such a person?

As with everything in startupland, it’s impossible to generalize. I’ve seen many examples of companies succeeding doing the exact opposite of what follows. But I firmly believe the following approach has the highest probability of success.

The obvious cofounder is someone you’ve known for many years, a friend whom you’ve also worked with, whom you share values and ambitions with, who is adaptable while at the same time has abilities that perfectly complement yours. But of course, such a person does not exist - there are trade offs. Given that, here are the some characteristics I would look for - in prioritized order (list is obviously non-exhaustive):

1. Shares values and vision: This is rarely talked about initially, but can really fuck things up further down the road. Unless you agree on what to build, the journey will be unpleasant. More than it usually is.

2. Chemistry: Someone you work well together with. Can you disagree and argue in a productive way?

3. Adaptability: What you start out doing might not be the thing you end up doing. Being an expert at the former without adaptability might stop you from start doing the latter, even though it’s the right thing for your company.

4. Complementary skillset: If you’re a builder, get a seller. And opposite. Allows for specialization, and with those two roles in place you normally should be able to get a working prototype in the hands of some customers. But diversity is more than that, preferably you also find someone who looks/thinks/acts different from you, due to the many benefits this brings.

Preferably, you find someone who ticks all these boxes in some ways. If you team up with people who are too similar to, you end up with a very expensive machine that’s unable to build or sell. That rarely goes well. But what I find is that most people start out looking for someone with a complementary skillset, not someone with shared values and vision. This is okay if you’re just looking for a job to be done, but building a company is so much more than that.

This is why I, and many others, advocate finding someone you already know. Discovering whether you actually work well together and towards the same thing is hard to discover over coffee, and something you want to know before you embark on your journey (and before most will fund you).

If you agree with this perspective, it has implications for how you go about looking for a co-founder. Activities such as “co-founder dating” and showing up at a technical university should be replaced with spending time with your immediate network. Do 50 cups of coffee with people you know and potentially one degree further (people they know), and chances are you’ll find someone who would be a good match.

Sometimes you find people who are right and available, they just haven’t broadcasted their availability to the world (very few do this). But when you meet them you learn they are looking for their next opportunity. Other times you find people who are a good match, but unavailable at the time (eg. they have no savings and a mortgage). But I’d much rather bring on such a person part-time to start, with an agreement this person will go full-time once funding allows for this. And sometimes this search is unsuccessful. But due to the benefits of having a cofounder, I’d always recommend doing the search.