Lenny's Podcast: Product | Career | Growth
00:52:13 · Jason Cohen
[00:52:13] S1: Is the pricing right? And why do we think it is? Because we probably don't have good reasons yet. Okay. So the third one is our existing customer is growing. I think everyone probably knows this, but just to say it. Okay, if cancellations overtake marketing in magnitude, one way to come back, one thing is, okay, make cancellations lower, but they can't be zero. So what else do we have to combat cancellations that would be proportional to our size so that it keeps up, unlike marketing or basic direct marketing? So one would be, all right, 2% left, but of the remaining 98%, some of those upgraded or otherwise paid us more, maybe it's usage based, whatever it is. They're paying us more, and so that covers the gap, and yes, if I tripled the company overnight, that would triple, and so that's the answer. So that is the answer, and maybe that's obvious but it's useful to tie it back to the sort of mental model we've got going. And of course, the metric here is NRR, net revenue retention, and the way that's computed is you say what is the revenue of customers right now? So existing customers, existing whatever, just the whole total, and then one year from now, what if that remains? So not new customers coming in, not talking about them because we're asking about the cohort that exists. What remains? So with cancels, it goes down, with downgrades, it goes down, but with upgrades, it goes up. So when I say remains, it could end higher than we started if upgrades exceed cancellations and downgrades. And now we are talking about MRR and not n, because n doesn't have this. N doesn't have an upgrade. N just only goes down, which again is why I think the n is actually the most important one. Because think about it. A lot of times, people think ... So if you've heard of NRR, you... ... because think about it. A lot of times people think... So if you've heard of NRR, you might think, "Well, that's my golden metric. I'm done." But the issue is if NRR is positive, but N goes down too fast, it doesn't matter because not enough people are left. And so there's not enough people left over to upgrade. And so actually, you're wrong. And so NRR does not include that, and therefore, it actually undercounts what's going on in a bad way, in a way that hurts you. There's yet another way to see why what I'm saying is right.
00:54:31 · Jason Cohen
There's this thing in investments where, let's say I started out at $100 and the stock goes down 5%, so now it's at 95, or no, it goes down 20%, now it's at 80. Then it goes up 20%. Is it back to 100? No, because 20% more than 80 is 96. So if it goes down 20% and up 20%, it does not come back to zero. It's worse. When you have a loss, a percentage loss, you have to have a greater percentage gain just to get back to where you were. In this case, a loss of 20% requires a gain of 25% to get back to where you were. This is why NRR isn't quite right, because NRR is saying that a loss of 20% from cancellations is offset by 20% from upgrades. As we just saw, no, it's not. That only gets us to 96% actually. So this is why, again, I believe in NRR. I'm saying you got to track it. It's good. Just in the back of your mind, realize it's not quite that good, and looking at N keeps you honest about what's really going on with these customer cohorts. So that's why they're both useful, in fact. This is why they're both useful. So NRR, of course, is important. A nice way to see this is, if everything I'm saying is true and there's these limits and stuff because of cancellation, then there should be no way to get a big company like a public SaaS company unless NRR is greater than 100. Otherwise, cancellations should just win. And that is in fact the case. There's over a hundred SaaS public companies, and something like two of them have NRR less than 100%. That's how it goes. And those companies have horrible financials and their valuations are bad. It's not good. It's not a good thing. And in fact, the median for an IPO SaaS company, like at IPO, the median NRR is 119%. So yes, that's what it takes. You can't do this. You're limited in less. Now, your goal may or may not be to get that big, but the point being it's mandatory for growth. Now, if the literal customers are just leaving, you got to plug that hole first, like you said. But okay, if they're okay, that's why this is in order, if that's generally okay, now we turn to NRR to say, "Okay, but the ones who stay, they're hopefully happy, they need to grow. Okay." So that's the full story of NRR. I think people who've heard of NRR don't necessarily think about all those things and realize that. So a good question is, okay, what do I do with NRR? But I think the answers are pretty clear.
00:56:51 · Jason Cohen
You add features, you have different tiers, you change the pricing in some way with usage or seats or something that kind of goes up automatically as they get more value out of it, so I don't think that's terribly interesting to double click into. It's sort of obvious. I would say, "Oh look, it's tied into pricing because their behavior," but the main thing is you want it to where the customer themselves would agree when they pay more, that they are getting more value. Hopefully they even think they're getting far more value than the price going up. I'm going to say this as if it's precise, which it's not, but they need to feel like if the price doubles, "Yeah, but I'm getting five times the value, so that's fine." That should be the feeling, whether they can measure it or not. A good way to do that is to say, "Well, then you should be measuring whether they're getting value out of it." Often we measure usage metrics and other kinds of metrics within our product because we can. But actually, what's really important is to measure how does the customer value this? And we need to measure that so that we make that go up because if we make that go up, they'll be willing to pay in whatever structure. And if that isn't going up, they won't be willing to pay. So even if we start making them, they'll leave. And we all know, we've all probably done it ourselves, we've all had products we love, but then as we scale, the price goes up faster than we feel the value is, and then we start looking for other products. We've all experienced that. So that's what I'm saying. To do that, you want some sort of measure of the value the customer's getting. If you're really lucky, that can be a number. That'd be wonderful. Then go do that and maybe that's your North Star. But admittedly, it's not always possible. So then the question is, the usual questions and metrics, are there proxy metrics that we understand are not the full picture, but they're helpful, they're part of it? And I'm a big believer in saying not all important things are numbers. I mean, even things like how differentiated are we in the market? Not a number, but it's very important. So this might be one of those things that's important, but not a number. So okay, can we get some proxy metrics, even of behavior and other things that's something better than some metric that's just operational?
00:59:12 · Jason Cohen, Lenny Rachitsky
And even if it's qualitative, okay, can we do that? Can we talk to customers and ask them qualitative questions to try to see? I would just say do your best here, because only when you generate more value for the customer, you can then decide how to split that with the customer in terms of things like price. But that's in fact how I think of it, that very phrase. How do we create more value for the customer and then split that with them? And when you do that, you're keeping the customer forefront in mind. You are taking some. Splitting means you get some. Let's not forget. It's not a charity. And on the other hand, first we should think, how do we generate value for customers, and then we've now earned the ability to take a little piece of that. So to me, this is the right way to think about NRR, not just, "We'll add a feature and make them pay." True, but let's actually take it from this different... Let's get there from this different perspective. [01:00:10] S0: Amazing. In the recent Madhavan episode where we go into pricing, you actually have some really good tactical advice for measuring the value that you're giving to a company to quantify that, which feeds into this idea of how do I create more value for you and then how do we split it? [01:00:25] S1: Yeah. [01:00:25] S0: The other element of this that's top of mind is just this land and expand strategy. There's a lot of companies that are just like, "Okay, cool. We'll get in with some price. We'll expand. That'll be amazing," which is essentially expanding as NRR going above 100%. [01:00:37] S1: Yes.
01:00:37 · Jason Cohen, Lenny Rachitsky
[01:00:37] S1: Yes. [01:00:37] S0: Something Jen actually shared in her chat that was really important is that you can't expand that much, at least for a while, because if you get in for 10K, if you go up to, okay, now it's 100K, someone's going to be like, "What is this? This can't go up 10X. Are we getting 10X value from this?" You can't just raise prices later. You're kind of stuck at that reference point, so you have to be really careful there.
01:00:59 · Jason Cohen
[01:00:59] S1: Yeah, I think that's right. And maybe you don't deserve it. In other words, especially when there's investors or other sort of forces saying like, "Hey, we need to X, Y, and Z," forces that are not the customer saying that, yeah, you can be coerced into making pricing or other kinds of policies that in fact are not good for the customer. So one tool I use is when anyone claims anything really, is, is that really true or is that really actually good for the customer? Because look, we are going to do things that are selfishly good for us. We have to. We can't just do things that are bad for us, but is this in fact good for the customer? Because often even in an internal proposal, we say it as if it is. "Oh, this pricing will be good. It'll raise prices on everyone, but it's better because of this reason," that we're sort of justifying. It's like, well, will the customer say this is better? If the answer is no, it's like, okay, it's better for us, but sorry, we have an equation where it has to be better for us and better for the customer. Sorry, it's an and. And of course, not all companies do that. And we experience that, all of us as consumers on the other end of that, and we don't like it. And that's not a good long-term strategy, even though it might work in the short-term, as many bad long-term strategies are.