Harry Stebbings was kind enough to have me on his podcast The Twenty Minute VC. That episode dropped today and you can listen to it here.
Here are some of the topics we talked about:
1) Building startups in the time of COVID
On creating startups during an economic crisis:
Innovation doesn’t stop during a downturn. My own experience was that the two unicorn companies I was involved in creating were primarily built during downturns. PayPal started in 1999, but it was really built in the wake of the dot com crash in 2000–2001. We IPO’d in 2002. Similarly Yammer started in the wake of the 2008–2009 economic crash. So my experience with downturns has been that it’s still very possible to create great companies. And in fact, there are things that get easier during downturns. It’s easier to recruit talent because there’s less competition. There can also be fewer copycats crowding a space. The one thing that gets harder, obviously, is fundraising. But innovation doesn’t stop during a downturn and you can still create great companies.
How the sales pipeline will be affected by the downturn:
There are going to be air pockets ahead. Nobody is going to be immune. What we’re going to see over the next several months is who is truly mission-critical and who’s not. There’s always been this mantra in SaaS investing that you want to invest in painkillers, not vitamins. Well, during good times, people are willing to buy vitamins, but during a downturn, it really does shift to painkillers. And you’re going to see over the next several months which services really are mission-critical.
On whether startups should give discounts:
If the underlying customer is in a very hard hit industry and can make a case for some relief, it can make sense to invest in the goodwill of that relationship and not demand things like increased deal sizes, and it may even make sense to offer discounts or defer payments. But the thing that startups have to keep in mind is that there is increasing transparency on the customer side in terms of these discounts. You have to expect that what you’re doing for some customers, other customers may find out about.
[Since recording this podcast, Craft announced its investment in Vendr, a service that helps companies manage their software spend. It’s a no-brainer for any company that wants to reduce its software bills.]
Do all founders need to be wartime CEOs now?
The short answer is yes. They do need to be wartime CEOs. And what I associate with a wartime CEO is somebody who is not afraid to make tough decisions, who is not afraid to flip on a dime and do the tough things that are necessary. The reality is, if your business just dries up overnight because of COVID and no revenue is coming in the door, obviously that implies that your Burn Multiple has become horrible. You’ve probably got some really tough decisions to make because you go from having an acceptable burn rate to one where the company is going to die in a number of months if you don’t react. And wartime measures are called for.
Reacting to a crisis:
The most important thing is you have to react to the new facts, to the new situation on the ground. Where people go awry is that they’re trapped in some sort of legacy thinking — they’re too anchored on the past. They’re too anchored on what the old plan was or what they thought they were going to be doing at this time. The old sales plan, the old hiring plan. So they’re slow to react. Another version of this is just having rose-colored glasses on. They don’t want to acknowledge the new tough realities, and that causes them to be slow to react.
How do you feel around the “celebration of failure” culture?
Frankly, I hate it. I don’t believe in celebrating failure. I think that being a founder is so hard. Running a startup is so hard, and I’ve never seen a case where a successful startup didn’t almost die multiple times. Certainly when we were doing PayPal, the company came so close to death many times. When I was doing Yammer and it had to compete head-on against Salesforce, we had all these really scary challenges. I think that psychologically speaking, if you allow yourself to think that failure is acceptable, you’re much more likely to succumb to it. Psychologically speaking, you’re better off burning the boats, so to speak, and saying that failure is not an option, it’s not acceptable, and we will do whatever we have to do to survive and persevere.
Now, at the end of the day, we know that in reality, failure is an option. It can happen and we’re not going to hold that against somebody. One of the great things about Silicon Valley is that when there is a failure, everyone just kind of dusts themselves off and they move on to the next one and that’s okay. It’s not like hundreds of years ago where we sent someone to debtor’s prison or something like that. So on the one hand, we do allow for failure, but I think that it’s better for founders to pretend like it can’t actually happen.
Whether to invest in startups that are seeing a boost from Covid:
You need to ask the question whether the growth is a flash in the pan or is part of some larger trend that’s being accelerated. You’re seeing both effects. What COVID is doing in a lot of cases is accelerating an underlying trend that was already there. If that’s the case and the startup is benefiting from it, then you would give them credit for that.
2) SaaS Metrics: Unit Economics, CAC, Churn, and Capital Efficiency
How to think about unit economics:
What you want to avoid is a situation in which the company is basically selling dollar bills for 90 cents. That’s the classic unit economics problem where you’re selling a product for less than what it costs you to make on a variable basis. And the reason why this has taken center stage is because over the last decade, as software has eaten the world, you’re seeing more and more startups have a hybrid software/real-world model. It’s not just a pure software business; there’s some physical world component to it. It could be a restaurant delivery service where it’s not just a mobile app, you actually have a driver delivering the food. And in these cases, you have to make sure that the cost of providing the service does not exceed what people are willing to pay for it.
The tricky thing is separating the unit costs, which are costs associated with each incremental order, from corporate expenses, which would be, say, the cost of developing the software. It’s the lack of that proper attribution that can get startups in trouble because they don’t even realize that they’ve got a gross margin (unit economics) problem.
On whether I agree with Peter Fenton that the biggest challenge for startups today is that there’s no free and open distribution:
I would agree, but I would divide the problem for B2C versus B2B companies. For B2C, you typically do need some sort of new platform or new distribution channel. Consumer startups typically need to grow virally to get to large user numbers because you can’t afford to spend a lot on CAC (customer acquisition costs). The long-term value of any particular customer is low, so you can’t afford to spend a lot on CAC. B2B is different because B2B companies lend themselves towards more of a sales-driven approach. The customers have higher long-term values (LTV), they tend to have lower churn, they stick around for a long time. As a result of that, you can afford to spend money on CAC. It pencils in a different way. I would agree that in the absence of new consumer distribution platforms, it’s very hard to get a new consumer company to pencil.
Rules of thumb for CAC:
There are a couple common rules of thumb around CAC. One would be that you don’t want to spend more than first year’s revenue acquiring the customer. I think that’s a pretty good rule of thumb. Another good one is that the longterm value of a customer should be at least 3X the CAC. I would agree with that, but say that it’s got to be even higher than that because a 3X LTV-to-CAC ratio implies that the customer is churning after about three years. It’s about a one-third churn rate per year, which is way too high. So I like to see a situation where on a revenue basis, if you’re looking at cohorts, that the revenue cohort is always growing year after year.
On blended CAC:
This is something I’ve warned about. You really want to separate your CAC by channel. What can be misleading about blended CAC (reporting CAC across all channels) is that any startup that has good word of mouth will get a bunch of customers through inbound, where there was no CAC. People just came calling. And if you average the CAC from that organic growth with other types of programs, it will give you an exaggerated sense of the efficiency of outbound channels and you’ll end up over-investing in those channels. It may take you some time to realize that some of those channels don’t actually work. What’s really working is inbound, and you’re spreading that peanut butter of good inbound CAC over a bunch of bad outbound channels.
On the virtue of selling to startups versus enterprises:
In deciding whether to sell to enterprises or SMBs — and startups would be an extreme version of SMB selling — it’s really a trade-off between the value of a customer and the difficulty level of selling to that customer. I don’t dispute that enterprises are the most valuable customers to have. You see that especially during a downturn, where enterprises are much less likely to go out of business, whereas startups see a lot of attrition. But the flip side of that is that startups are so much easier to sell to. Therefore they can provide a great entry point into the market for other startups, which may find that the difficulty level of trying to sell to a Fortune 500 company is just too high.
On logo churn:
Logo churn is unavoidable. All startups are going to have some logo churn. It’s just impossible to avoid. But what you really want to see is that the expansion from the accounts that stay with you exceeds, in dollar terms, the churn from the customers that you’re losing. On a revenue cohort basis, you want to see net negative churn aka net positive expansion.
What’s a good churn rate:
Good (or not bad) churn would start at 100% retention on a revenue basis. Otherwise your bucket has some holes in it. And it’s very hard to build a subscription business for the long-term if on a revenue basis, the revenue is attritting every year. So you really want to start with at least 100% retention on a revenue basis. Certainly if you can get up to something like 130%, that would be industry leading. We’ve seen some startups get all the way up to 200% when they have a land-and-expand strategy. They might start in an enterprise with a smallish deal, but are able over time to grow that onto a lot more seats. That’s a fantastic scenario when it happens.
How to think about capital efficiency:
I’ve defined a metric that I call Burn Multiple, which is the net burn in any given period divided by the net new ARR. It’s thinking about burn as a multiple of the growth that you’re achieving in that month, quarter, or year. It’s a very useful way of thinking about why you’re burning so much money. At the end of the day, the reason for burn in the first place is that you hope to achieve growth. So you want to think about, how much burn is it costing me to achieve each unit of growth?
The burn multiple will vary by stage. In the earliest days of a startup, invariably the burn multiple will be bad because the company has to invest in R&D. Obviously if the startup is pre-revenue, the multiple won’t even compute because the denominator will be zero. So in the earliest days where burn is high and new ARR is low, you’re going to have a pretty bad burn multiple. But the burn multiple should be improving over time. In order for any company to become profitable, which is the goal at the end of the day, its burn must go to zero. So the burn multiple should be approaching zero over time. If you see it going in the wrong direction, that’s a sign that something isn’t working and that you need to be more disciplined.
What Burn Multiple gets you excited as an investor:
Certainly, if the burn multiple is less than one, which means an early stage startup where they’re able to burn a million dollars or less to achieve a million dollars or more of net new ARR, I would qualify that as being in the amazing category. We have seen a few deals like that recently that Craft has done.
The flip side of it is, if you’re burning more than three times your net new ARR, I would define that as being suspect or bad. And that’s the case where burn needs to be very closely scrutinized. In other words, you’re burning a lot, but you’re not growing that much in relation to burn. The question then is, why are you investing so much? What is it really getting you? Are you over-investing?
On whether startups should “do things that don’t scale”:
I strongly disagree with that mantra. I think it’s one of the shibboleths that gets startups in trouble. The whole point of technology is to find the elegant, scalable approach. The sooner you do that, the better off you’re going to be. When you start to glorify doing things that don’t scale, you end up with a startup that doesn’t scale, which is kind of the opposite of the whole point.
3) Founder Psychology and Board Dynamics
On whether founders need to be “crazy”:
Founder psychology is a really interesting topic. Founders are often described as “crazy.” When everything is up and to the right, that means something positive. Then when the wheels have come off, it means something negative. So the question is, how can you tell the difference between good crazy and bad crazy? My view is that founders do need to be far more aggressive than what the average person might think. They need to have these traits of being visionary and pushing and being able to run through walls. Just having a “sane” level of these traits may not be enough. But where things go wrong is when founders lose perspective and become immune from seeking advice or listening to good advice. That’s when you’ll see a situation where founders can go off the rails.
We always try to do reference checks on the founder to understand the person. But you can’t always tell in advance what’s good crazy versus bad crazy. And the reality is that founders are the ones who choose their board members more so than the other way around. It’s really up to founders to decide who their board members are going to be. And so when I talk about the founder psychology issue, I’m really hoping to give advice to founders who would see the wisdom of seeking to balance their psychology by having a board member who supports what they want to do, but can also give them some constructive feedback.
On the right board member:
The nice thing about a board is that you’ve got multiple people on it. So you can build a team that is well-rounded. Everyone can bring their strengths. Some board members are fantastic because in their previous career, prior to investing, they were a CFO, so they bring a very rigorous, numbers-based approach. They might scrutinize the numbers and give advice on that dimension. You can have board members who are very good at supporting the founder emotionally and have a high EQ and play more of a cheerleader role. And then, you can have board members who are better at strategy. Everyone should bring their strengths.
On the lack of trust between founders and VCs:
The analogy I’ve made is that over the last decade, a little bit of that Hollywood auteur mentality has crept into the tech world, where founders want total creative freedom like Hollywood directors. They want final cut. That’s as it should be. At the end of the day, founders are the ones running these companies. It’s their company, they’re the ones who should be making these decisions. But if it blinds the founder to seeking advice or seeking to balance their psychology or their perspective, I think that’s dangerous. So I would counsel against that sort of mentality. Founders should realize that everyone’s on the same team. I always had great relationships with my VCs and board when I was running Yammer. The relationship should be constructive.
4) Being a VC
Why found Craft Ventures instead of just angel investing:
I spent about 20 years operating and founding companies and being on the founder side of the table. I got to a point where I wasn’t really interested in running companies anymore, but I still love startups and working with founders. So I decided to move from more of a player to a coach role. Once I decided to invest full-time, I liked the ability of a fund to write bigger checks and lead rounds. It also allows you to have a team. We have a really great team at Craft, and it would be hard to have a team like that if I was just angel investing.
On time allocation within the portfolio — do you focus on the winners or losers?
For me, I don’t really think about it that much because I’m there for my founders whenever they need me. So when somebody calls up with a tough problem, it’s not like I think, “Well, I don’t want to spend time on this because this isn’t one of the putative winners.” You don’t think about questions like that when the founders call and need your advice and need your help, you just react. So I’m always there for founders when they have a question or a problem, and that’s always my highest priority.
You tweeted Silicon Valley is no longer a place it’s a way of doing business. What did you mean by this?
Especially as we all move to remote work, but even before this, we were seeing startups pop up all over the world and start to use the Silicon Valley template. That template consists of a few critical elements: One is that all the employees are owners of the company. They have broad-based stock option pools. The second is the way that VCs and investors participate in the company, which is to provide true risk capital. Nobody is too upset if they end up losing 100% of their investment, which is not the way the rest of the world works. A lot of places in the world, if you go to the investors and tell them you lost all their money, they’re going to say, “No no, where is it?” And send a guy to break your legs. This combination of risk capital combined with broad-based employee ownership and support for the very best ideas is something that started in Silicon Valley, but it’s now spreading to many other places, and it’s a great thing.