Lenny's Podcast: Product | Career | Growth
01:23:45 · Lenny Rachitsky, Sean Ellis
[01:23:45] S0: Okay. Any others? Any others real quick? [01:23:47] S1: Yeah. I mean, I didn't really think about North Star Metrics when I was at Dropbox in Eventbrite, you know, like the the term itself. But I was thinking about what is what is a valuable experience with Dropbox and how do I get people to have that more time? But, I don't even know what they go with today, but maybe files in Dropbox, files access might be better than just files hosted. And then probably for Eventbrite again, I would say weekly tickets or something like, you could say weekly events, but then you have events that don't sell any tickets where weekly tickets would be more likely to reflect events are gonna be happy if they're selling tickets and, yeah. [01:24:30] S0: Okay. Sean, we've gone through so much stuff. I I have... I'm trying to limit how many more questions we get through just so that we don't... [01:24:36] S1: We're going long.
01:24:36 · Sean Ellis, Lenny Rachitsky
[01:24:36] S1: We're going long. [01:24:38] S0: We're going long, which is amazing. I think there's so much value here that we're collecting for folks. So let me just ask maybe a couple more quick questions. One is actually from Andrew Chen who is currently a partner at a sixteen z. He was... He wrote about growth for the longest time. He think he helped popularize growth hacking for better or worse with his article on it being the future of VP... Oh, what is it? Growth hacking is the new VP of marketing. Right? Is that the title? So he actually had a question for you that he shared with me. His question is, growth strategies have changed a lot over the past decade. What is the biggest difference now versus when you first started working on growth?
01:25:14 · Sean Ellis, Lenny Rachitsky
[01:25:14] S1: When I first started, just being data driven on customer acquisition was enough to win. And being test and data driven on customer acquisition, all the other companies were CPM focused. So we could do really well just with lots of testing and some creativity in how it all worked. But over time, as now I would say most online marketers are very data and test driven. They know they needed lots of testing. So to be competitive today, you actually have to be able to be super efficient at all parts of the business. So, again, how you convert, retain, monetize, and that's when it gets hard. Getting a marketing team to be data and test driven is pretty easy. Once you start getting into activation and referral and engagement and retention, now you're talking about the overlap between marketing, product, if it's B2B, bringing sales in there, customer success, And those teams are not used to working together. And so it's really hard to drive the collaboration that's needed to have an effective testing program across the entire growth engine. That's pretty much any business that's been successful with it, implemented it super early in the business. And so very, very few later stage companies have been able to make much progress in replicating that type of approach. [01:26:55] S0: It's just gotten harder, basically. Things are just getting harder.
01:26:55 · Lenny Rachitsky, Sean Ellis
[01:26:55] S0: It's just gotten harder, basically. Things are just getting harder. [01:26:57] S1: It's gotten harder, but I but I but I I think it's, I think it's possible. So it's what I obsess about all the time is how how how do you how do you get cross functional teams working together on growth now, and it's still a huge advantage when you can pull it off. [01:27:13] S0: Okay. Totally unrelated question going in completely different direction as we close out our chat. So you came up with ICE, the very popular way of prioritizing work, which is crazy. I did not know that until I started prepping for this conversation. What's your thoughts on Rice, the intercom version of ICE, where r stands for reach, I believe. Thoughts?
01:27:35 · Sean Ellis
[01:27:35] S1: So I think I think it's an unnecessary addition, but maybe it's I'm just being protective of my original idea that the I in ICE is impact. And it's essentially saying, best case scenario, how much impact could we get from this? Reach is a super important part of impact. And so like I think it's already factored in the I in ice. And so I think if there's anything that I would be accused of, it would be being oversimplifying things and there's, I'm not saying them, but there's a lot of people who approach things with, there's got to be a more complex way to approach this and that's just not me. And so, yeah, more testing is better. That's, oh, it doesn't just work like that. I mean, better tests are better than bad tests, but just if you had to hold yourself accountable to anything, more testing would be better. And so I think just one quick note on ICE is that in order to be able to effectively run a high velocity testing program, you need to be able to source ideas from across the company. And that's why I came up with ICE. That if you're having people submit ideas and you can't tell them why their idea was not chosen, they're just going to get upset and you're going to waste a lot of time. But if you have a systematic way of being able to compare ideas, it's it's more likely that people would be able to get it and they'll be able to come up with better ideas.
01:29:06 · Lenny Rachitsky, Sean Ellis
[01:29:06] S0: I love the way you think, Sean. I have a post on prioritization where I basically just make the same argument that there's all these fancy, complicated ways to prioritize. In the end, it's impact confidence and effort, and it really works and rarely is more work necessary. On the other hand, I do also have a guest post called Drice by these two guys called Detailed Rice, which actually I think is a really good point where sometimes it's worth spending, like, thirty minutes per idea to just really estimate how long will it take to to avoid doing things that are just gonna not work and very unlikely to work. Basically, doing this reach piece, and spending the time to do it right. And I think there's a lot of good value there. [01:29:45] S1: Yeah. And that's, what what I think is gonna be really interesting is that, over time, I think AI is going to is going to actually change our ability to model out potential outcomes on experiments and start to whether it's a more informed way of doing ICE or replaces ICE that ultimately probability of outcomes is something that AI will be pretty good at. [01:30:11] S0: Well, amazing segue to the final question. Actually, final question is I wanted to ask you about any ways you've been using AI or ways you think AI will impact the work you're doing or other folks are doing, and maybe you just answered it, but tell me.