Lenny's Podcast: Product | Career | Growth
00:57:39 · Sean Ellis
[00:57:39] S1: One of my favorite quotes is a quote from a guy Kettering. It was early, like a hundred years ago at GM running innovation. And he says, A problem well stated is a problem half solved. And so I think a lot of it comes down to not the things you try, but how you deeply understand the problem that's preventing someone from using your product effectively. I'll just give you one example. We had one channel after we made a lot of these changes and had already driven a ton of improvement in the LogMeIn onboarding, we found a demand generation channel that was really cheap and the economics looked great. But at just the download step, had a 90% drop off rate. And so we AB tested a bunch of different things there to try to improve that conversion rate. And then finally, you know, we, 10 plus tests not able to improve it. Finally, someone said, well, these people are registering. Why don't we just ask them why they signed up and didn't download the software? And so we don't want to do it in too kind of a creepy way. So we just said, made it look like a note coming from customer service. Channel was sending 200,000 people a day. So 20,000 people were converting to registering. So we had essentially 20,000 people we could email and then 18,000 of them who didn't download. And so we just asked, Hey, notice you haven't had a chance to use the product yet. It looked like it was coming from customer support. What happened? And the answer we got back in not a formal survey was, Oh, this seemed too good to be true. I didn't believe this was free. I mentioned to you we were one of the first freemium SaaS products out there. And so people were skeptical when, especially in a demand gen channel where they hadn't sort of seen a radio or a TV advertisement from our competitor who was a premium only product. These are people who are discovering the category for the first time, they were getting there and so our next test, once we articulated what the problem was, our next test gave us a 300% improvement in the download rate, which was a... We gave them a choice, download a trial of the paid version or download the free version. Put the graphical check mark next to the free version. But when they saw we had a business model and a trial of a paid version, the free version was credible. And so that essentially made that channel work for us.
01:00:07 · Sean Ellis
So I think, again, it's a combination of qualitative research, looking at how others did it. We had this theory. Our previous company had been a game company that didn't require a download. So initially we had this theory that maybe just downloadable software can't be in the millions of new customers a month. And so we're being unrealistic here. But then we're like, are there any counter examples to that? Like, no, the instant messengers are downloadable and they have hundreds of millions of customers. So let's study their download and install process and see if we have any ideas that we could borrow from that. So again, some inspiration, tried some of those things. It a combination of just trying a bunch of different stuff that ultimately led to... I wouldn't say there was one big gain. It was it was a bunch of small gains.
01:01:00 · Lenny Rachitsky
[01:01:00] S0: Awesome. Okay. So a few things for people to try if they're like, hey. How do I improve my activation rate? How do I improve my conversion rate? Is just, like, drill further into what is stopping people from progressing. Like, ask them, why did you bounce here? Why did... What did you think this was gonna be? Why didn't you end up using this? Look for inspiration from other products. I think people probably already know that. You talked about earlier this idea of the positioning having a big impact of just, like, figuring out they want antivirus software. Let's make that very clear. Hey. We're... We've got the best antivirus software. That's what we're here for. So there's probably just, like, messaging that you find works a lot of times. Right?
01:01:36 · Sean Ellis, Lenny Rachitsky
[01:01:36] S1: I mean, your two big levers on on driving a conversion are increased desire, reduced friction. And so, yeah, you definitely want to increase the right desire. And then sometimes it can also just be reminding people along the way of what benefits you're going to get. In the case of LogMeIn, it was probably the most complicated funnel I've ever seen because you couldn't even get to the moment while you're sitting in front of the computer. You actually had go to a different computer to use the service to remote control the computer you're in front of. So it's not surprising that there was so many steps where we could lose people, but we just weren't that intentional about designing each of those steps initially. And it wasn't until we kind of thought through why would we lose someone at this step and studying the data, which steps were we losing the most people at, then deeply trying to contextualize why are we losing them there, coming up with a set of tests that we wanna run, and then having a good way of deciding which one to test first and ultimately focusing the tests on the areas where we're losing the most people. [01:02:45] S0: The other element of this is coming up with an activation metric and aligning on, like, here's what we consider someone activated. I know this is very dependent on the product, but any advice or heuristic for how to help people decide this is our activated user?
01:02:58 · Sean Ellis
[01:02:58] S1: I tend to start qualitatively. So just like when do I think they've had a good enough experience with the product to really know it? So like in the case of LogMeIn, was pretty easy. If they didn't do a remote control session, they didn't use the product. There was no value along the way there. And then at least try to see if there's a correlation to long term retention of doing that. Causation is you need to do some experimentation to prove causation. At the very least, I want to see that correlation. But if I start with two or three ideas of what it might be and then go and study the data, that can help you focus. But again, don't think there's necessarily one exact right answer of what is that moment. There might be two or three different things. I think it's that intentionality about picking something that's experience based and saying, What is a likely experience that someone's going to get a good enough taste of this product? And then I do see some companies that are like, Well, the activation moment should be... They've used it a 100 times. That's going to correlate to long term retention, but it's just not very actionable. It's so far down the user experience. Ideally, there's a way that I could get them there in the first session, in the first day, that's great. And so it's sort of something that's value that can be experienced super early. But I actually give you an example from the first company I worked on was a game company where I actually flipped it and basically instead of making a traditional funnel where they could play our games after they signed up, I made our games the advertisements. So basically, we syndicated our games to 40,000 websites. They started gameplay experience on the other website. Then they would get a message that they now have a qualifying score. And if they register, they'll be in the drawing for the weekly cash prize. And then we could pull them into multiplayer games on the site. And so it's it's kind of the the strategy that YouTube used to grow, but it was, like, two years before YouTube introduced the the approach.
01:05:14 · Lenny Rachitsky
[01:05:14] S0: Yeah. It feels like you basically created Zynga is what I'm hearing there. Okay. So let's move further down the funnel. So we've talked about activation onboarding. The next phase that you focus on is basically... Some people call this growth loops, growth engines, flywheels. Basically, it's like the thing that helps your business grow. And something I... I'm curious if this resonates. I found there's basically four ways to grow, and usually one of these engines is responsible for almost all of your growth. So what I've seen is basically it's... You're gonna go through sales, you're gonna go through SEO, you're gonna go through virality word-of-mouth or paid growth. Does that resonate? Does that feel right?