Episode Transcript
Speaker 1 00:00:08 Welcome to the breakout growth podcast where Sean Ellis and Ethan Gar interview leaders from the world's fastest growing companies to get to the heart of what's really driving their growth. And now here are your hosts, Sean Ellis and Ethan Gar.
Speaker 2 00:00:26 All right, in this week's episode of the breakout growth podcast, Ethan Gar and I chat with Chris Ramini co-founder and CEO of fireflies.ai with fireflies, your zoom and other video meetings are recorded, transcribed and organized in one place. Essentially. It's like an AI assistant that helped you and your team stay in sync. So if you're working on or interested in product led growth initiatives, this conversation's gonna be really valuable for you. So, so Ethan, what did you get out of this conversation with Chris?
Speaker 3 00:00:59 So PLG is all the rage these days. And I think for good reason, I mean, who wouldn't be excited about this approach where product itself is driving acquisition activation and retention. PLG puts a great, a really nice rapper around that, but it's not right for every business. And even when it is right, it may not always be right. It may not be the thing you do forever or do it the same way forever. So I think by taking us through the journey, Chris was really able to show us why a self-service approach like the one he and his team have employed just really has worked so well for them. And why optimizing for discovery within that customer journey has just been so important to their growth. But now he's looking to hire account executives to help some customers who need a more sales driven interaction to actually succeed. So it seems counterintuitive, but I don't think it is at all. So I think we just have caught his story at a time where there's just a ton to learn from it.
Speaker 2 00:01:48 Yeah, for sure. They are not abandoning PLG. Yeah. I
Speaker 3 00:01:52 Didn't mean that now. They
Speaker 2 00:01:53 Said, yeah, they've simply reached a point where it makes sense to compliment that product led growth approach with some human touch. You know, you get too many people that are either in the touch sales camp or the PLG camp and, and become really religious about it. But, but the fact is for a lot of businesses, it's gonna be a hybrid approach that, that ends up working the best. And we've seen it before. We saw it with Karen Flanigan a few months ago from HubSpot. And so HubSpot is kind of the poster child for leveraging a self-service model. It got them to this hypergrowth stage, but even they have evolved the model to where they're starting to touch some of those, uh, higher, uh, upmarket customers that, that need that additional human touch to, to bring them across the finish line to a sale.
Speaker 3 00:02:41 Yeah. I mean, slack Dropbox, a lot of, a lot of companies who are really PLG companies over time have evolved to add in other mechanisms for reaching their customers where, where their customers live. So, you know, but if you're really, you know, early on in the journey, I think Chris surface is just really helpful ideas as well. It seems like they're really adept at gaining learnings through their users and following this idea that he kept talking about that you need to let customers buy the way they wanna buy the way fireflies has figured this out through that self-serve lens. I, I think it's just really instructive for our audience.
Speaker 2 00:03:14 Absolutely. Yeah. Another theme that emerged in the conversation that I thought was an important one that we, we don't talk enough about is, is this need for finding predictability? So in the beginning, it's this, you know, you're just trying to get product market fit, right? And then you're trying to figure out your growth engine, but eventually you want to be able to have a more predictable growth model, where there are a lot of known factors that are gonna help you reach your numbers each quarter, each, you know, every day, every month, every quarter and, and, you know, and a smaller and smaller piece is, is put on those unknowns. And so I thought that was another very interesting part of our conversation.
Speaker 3 00:03:56 Yeah, it's fascinating. Um,
Speaker 2 00:03:58 Yeah, I'm, I'm really glad that we had a, an opportunity to, to meet with Chris. I think their business is, is gonna become more and more important as remote, you know, is becoming, you know, we're, we're moving to, to doing more in, in person these days, but, but I think remote will forever change, uh, to, to become an important part of, of how business is done going forward. And, and that's, uh, they have a really good product for, for sure for supporting that. So, uh, enough jabbering, um, particularly for me <laugh> um, so, uh, let's, let's get in to the conversation with
Speaker 3 00:04:32 Chris. Yeah. Don't play me for all that jab. Let's do it. <laugh>
Speaker 2 00:04:35 Perfect. Hey, Chris, welcome to the breakout growth podcast.
Speaker 4 00:04:48 Hi Sean. Hi, Ethan. I'm really excited to
Speaker 2 00:04:51 Be here. Yeah. As you, as you mentioned, Ethan is also here, so, uh, Hey Ethan. Good to have you on.
Speaker 3 00:04:56 Yeah. Hey Sean. Hey Chris. Good to, good to be here with both of you.
Speaker 2 00:04:59 Yeah, we haven't, uh, Ethan and I haven't recorded a, uh, a podcast interview for a while, so it's nice. Nice to get back in the, uh, in the swing of things. We had a big backlog there for a while. Um, we,
Speaker 3 00:05:10 We, we don't want people to know that we're slackers
Speaker 2 00:05:12 <laugh>. Yeah, well, we have a lot of 'em, uh, scheduled here going, going forward. So, um, first, first of many here, but, um, so I I'm sure that a lot of our listeners are familiar with fireflies, particularly because they've been using zoom to the point of, uh, of probably, uh, never wanting to look at another zoom meeting, but, um, or at least wanting to occasionally see someone actually face to face. But, um, for anyone who's not familiar, uh, can you give us a, a, a brief introduction to what firefly.ai is all about?
Speaker 4 00:05:43 Yeah. So fireflies is an AI voice assistant for the workplace. So specifically your meetings, uh, it's an assistant that joins your meetings across zoom, WebEx, Google meet, go toeing teams, all the major video conferencing platforms. It follows you around, sits out like a fly on a wall, takes notes, transcribes those meetings makes 'em searchable. And then afterwards it also provides analytics and conversational intelligence on top, uh, for teams. So, yeah, it's an AI note taker for the lack of better words. Uh, but then does a lot more on top of it. And, uh, yeah, we've rolled this out right before the pandemic, so, okay. Part of 2020, and then yeah. Perfect
Speaker 2 00:06:26 Timing there. <laugh> right. So what, um, what led you to the realization that, uh, that zoom needed? Something like this?
Speaker 4 00:06:35 So previous to starting fireflies, I was a PM at Microsoft, worked across, uh, a few different, uh, product suites at office. The, the role that I was in was actually more in the growth role. Um, as a PM, we were starting out the first AV experimentation platform for office, which was, which was new at the time. And, uh, you know, I just reflected on my role as a PM and the amount of time. And that was my first corporate experience outta school. Uh, you know, I loved the role, loved the job, but just the amount of meetings that like run corporate America, you know, the globe in general and the amount of time we spend in meetings. And then everyone becomes a glorified secretary in terms of taking notes, kicking off conversations, having the repeat meetings again, because someone didn't show up having technical issues on those meetings.
Speaker 4 00:07:26 Uh, and then the person that was supposed to remember stuff and be on point is on leave. So there's, I've just gone through everything, uh, when it came to meetings and I said, look, we have an archive for emails. Uh, we have an archive for files, which is Dropbox. Uh, so, and how can I do that for my meetings so that I have perfect recall over every conversation I've ever had. And there were products that have existed, let's say like Evernote notion that help you create this like notion of like a second brain of all your information. But most of that information is still locked up and trapped in meetings. And I wanted something where automatically anything I said gets created into a repository. And now if you take that repository, uh, and build it in for a team, uh, so that that's one thing led to another, and that's what got us to start fireflies. Um, but yeah, we hadn't started in the voice space. When we, when we first started doing this, we were actually building conversation, NLP chat bots, and other stuff. Uh, the technology was similar use case and domain was different. And then it was later. Then we
Speaker 2 00:08:32 Realized the evolution of the business then
Speaker 4 00:08:34 Evolution was in voice. And that was like the biggest opportunity for us to tackle at the time.
Speaker 2 00:08:38 Cool. And then what, what's the business model?
Speaker 4 00:08:40 So the business model is SAS product led growth. So, uh, per user based, uh, per seat based pricing, uh, we have a freemium plan as well. So very similar to the calendar leads the slacks of the world, uh, the drop boxes of the world, where it's great, uh, in terms of single usage. So I can use it for myself. Uh, but then it works even better when my team is using it and we centralize everything into the team workspace. So that's been the, uh, go to market motion for us for, uh,
Speaker 2 00:09:11 And, and so is it, um, is it that you, like, like with zoom, for example, you know, one person might have a premium zoom account and then there's a lot of other people that are attending that meeting. Um, I, is that how the licensing works or is, is each person needing to have, uh, Firefly to, to essentially make it
Speaker 4 00:09:32 Work? Yeah. Any person that's having the meetings, uh, would have a license and, uh, any person that is added to your workspace would also have a license. So usually if there's 20 people on a team, uh, you know, maybe all 20 of them have meetings. Everyone has meetings, it's just in different departments. So they would all have, uh, license, but you don't need a license per participant on a meeting. Right. So I don't need a, so like a person, let's say we have this meeting and then afterwards, you guys will get a meeting recap. Uh, you can sign in for free and view all that content, just similar
Speaker 2 00:10:03 It's per host then really?
Speaker 4 00:10:05 Yeah. Per host. And then if you want to aggregate your team into a workspace, uh, then it's like, yeah. Perceived, like,
Speaker 2 00:10:12 Got it. So when you, so you think, yeah, you better get in here.
Speaker 3 00:10:16 <laugh> yeah. It's as usual. I'm just like, you know, I get like three, three questions in if I'm no, I'm just kidding. <laugh> no. Um, but, uh, Chris, a couple days ago, when we, when we first chatted, you'd mentioned that you'd really, you really dove into the PLG, uh, approach fully self-served model. I was curious, is that something that was an evolution, like the evolution of the start of the business, or was that intentional from the beginning? Like, we want to be a PL whatever we're gonna do, we want to be a PLG company and we're gonna focus on, on that approach.
Speaker 4 00:10:45 Yeah. I mean, Sean said this, but one of the hardest things, you know, you can build a really good product, but distribution, right at the end of the day, how do you stay on top of people's minds, especially when you're creating a new category of software, uh, you want to eliminate as much friction as possible. Uh, so fireflies, the reason we have the bot join the meetings, the reason it connects to your calendar is so that once you set it up, you don't even have to think about it. And it just works in the background. So I'm a very big fan of workflow software. So that was like the starting point of, or the delivery mechanism from that delivery mechanism, we realized, oh, meetings are a very viral place. Uh, people, multiple people on the meeting can see fireflies and they'll say like, oh, what's this AI note taker.
Speaker 4 00:11:27 And then they'll go search it up on Google. You get direct traffic from Google. Um, you know, and then the whole recaps thing as well, like, so I can have the email recaps automatically sent to all the participants after a meeting. So you can experience fireflies before you even started using it. Um, very similar to let's say like Calendly. Uh, and because of that experience, we said, Hey, the best evangelist, the best marketers for this is going to be other customers that are using fireflies, even the compliance piece, right? So someone can turn on a notification where an hour before the meeting fireflies will send a little notification saying, Hey, I'm an AI note taker. Uh, Ethan sent me to, uh, transcribe and take notes for this meeting. Uh, this is what I do. If you wanna turn it off, click here. So even the compliance opportunity turns into a, uh, conversion point.
Speaker 4 00:12:19 So we've felt that like the self-service model was like the starting point. And in the early days, fireflies was really built for single end users, right? Like we didn't have this concept of teams about a year ago. And then as we started building out the team's use cases, verticalized use cases like there's specific integrations and analytics for sales teams, uh, versus like recruiters and recruiting teams. And then internal teams, then we started realizing, right. PLG is great. We wanna let people buy the way they want to. Right. It could be like first time, like a few people in your org starting to use it to then them adding more teammates and then swiping their credit card. Uh, and then over time, as we realized, like fireflies penetrating into more organizations, uh, we needed to allow people to buy the way that they need to buy it.
Speaker 4 00:13:08 So sometimes people will raise their hand and they need, uh, a demo. Uh, they need to talk to someone or a manager needs to know about how fireflies works for, Hey, I just noticed like 20 people in my company are using fireflies. Um, I see the value for them, but what's in it for me. Right. Or how can I, uh, make sure, you know, we have all the controls, the admin settings, the, uh, you know, enterprise features. So there we said, okay, you can't avoid having a conversation. So we are PLG. And that is like the bulk of what we do, but recent times, and maybe in the last couple months, uh, we realized that the need to have conversations and, uh, use PLG as more of like a way to get product qualified leads, uh, and then, uh, help those that, that need need help.
Speaker 4 00:13:54 And, uh, yeah, so it's a natural evolution. I, I think even slack had this where like, we're never gonna have sales people we're, you know, just gonna sell like with the, the self-service and it's just gonna be great. Um, that works for a lot of customers, even like notion, uh, you know, grew on the backs of that. Like, I I'm assuming like 70% of their, uh, engine is self-service today, but there are some people, no matter how easy the product is to use or sign up, they're like, this is cool. I'm looking into this space. I want to talk to someone first. And so we're like, okay, we gotta let people buy the way they need to.
Speaker 2 00:14:27 So was, was that mostly then a response to just, uh, people wanting to buy that way? Or was it, was it part of like, we're, we're struggling to hit our targets with the purely PLG approach and, um, and, and we should try some of this to, to see if, if it's gonna help us, uh, you know, stay on track with targets.
Speaker 4 00:14:45 What we noticed was it was more of an added bonus, like, like a lot of the PLG metrics that we were trying to hit in terms of meetings like that were being processed through fireflies. Number of people getting, uh, introduced to fireflies people, signing up people swiping their credit card like that self-serve motion was helping us move the needle really effectively. And early days was just a lot of individuals like Grammarly, right? Was the individual product where people were swiping their credit cards. I heard the same thing about one password, which almost actually started as a consumer product and then later into a business product and then later into an enterprise product. So there have been some companies like Grammarly, one password, a few others that made their way from like almost a, B to C to this B2B enterprise world. And fireflies has been similar in that there were a lot of individuals in the first two years of us building fireflies out that were using fireflies.
Speaker 4 00:15:39 And for that motion self-service was great. We were hitting the, like the metrics we need to with that, just with that engine, I think we could grow. But what we started seeing is larger teams though need, like, you cannot, you know, they'll have like sec secur security questionnaires. We have like SOC two compliance. We list all this stuff on our website, but they still need to have that conversation. And so we realized, um, we can really amplify or juice the funnel, um, with, with this. And even there, we can automate a lot of the outreach part, right. Like we don't need SDRs. Right. Because we can look at product signals and then send out
Speaker 2 00:16:13 Email. Yeah. The product's kind of your SDR at that point.
Speaker 3 00:16:15 <laugh>. Yeah. So with all, with all of your success, with the self-serve PLG led approach and understanding that it is evolving over time to meet your, what you're learning from the market is there. And we've definitely seen companies that sort of, they love the, the idea of PLG of freemium of, I mean, it just sounds wonderful if you can make it happen, but from in your experiences, is there anything you would warn companies and other founders about like, if you wanna get into this, what you should stay away from or be careful about, or at least take into consideration?
Speaker 4 00:16:49 Oh yeah. So one is not every product is built for PLG. So you can't force PLG, uh, or true PLG. Like for example, collaboration products are perfect for PLG because it's users and then're collaborating with teammates and then it gets awareness, right? That's how slack, you start inviting other teammates and you have a good experience in slack. Uh, the challenge is like, let's say you're building an API company or some other sort of enterprise security company. It's much harder to do PLG there without the sign off from decision makers. Um, it's not like single player versus multiplayer. It's more of like an enterprise deployment and then you have to have access to APIs. So the feedback I give is not every product is meant for PLG. Um, it re really has to have certain notions of virality, certain notions of, um, that consumer mode. And I think that's true PLG in my book, but some people will have like a product that's very self-service and they'll still be like PLG, right?
Speaker 4 00:17:52 Like I, I think, um, or they'll call themselves PLG. So like, if you look at, uh, CRMs or some of these, like other SMB type CRMs, uh, you can sign up and start using it, uh, on your own and then pay for it. But that, uh, viral mechanism that you get in like a Dropbox, uh, doesn't quite exist or like Calendly doesn't quite exist. So in that regard, it can work. You just need to have a very tight funnel. Uh, the other thing with PLG is pricing. So there's huge discussions around, should I have a freemium tier versus free trial? We're realizing like we're all freemium. And a lot of people hit our paywalls at like certain periods of time works out, but there might be a need for free trials for our larger customers that are coming in where they need access to all of the business features, um, to make that decision faster.
Speaker 4 00:18:39 So our question is like, you know, which, which will help them get to that decision they need faster. Right? So, uh, long story short it's much, I believe much easier to build a product and monetize a product. Uh, if you have a very clear cut function where, okay, you're gonna get seven days, I know how much I have to spend on ads. What might payback period looks like? And I know how to acquire you, right? Like, that's, that's a very straightforward motion. Whereas with PLG and specifically freemium, it takes a very long time because you have to have a really good product out of the gates. Virality takes time, and you have to keep experimenting. You have to have a really robust data, uh, infrastructure to measure all of those changes that you're putting out. And, uh, also the price point, usually when you're doing freemium is may not necessarily be as like cheap, cuz you have to go for a volume.
Speaker 4 00:19:34 So a lot of those things, like if you're your goal is like, I need to get to revenue as quickly as possible. Then the best process that, that has been proven out in the past and the advice we were always given was pick a mid-market customer segment, hire a few sales people, try to close like 20 K uh, you know, ACV deal values. You'll quickly get to your, your targets, uh, having been a product and engineering focus founder, and then same with my co-founder. We said, nah, we don't, we don't wanna just do that. Like the sales motion cuz that that limits of who we can support. So that's why I just took a much longer time to build out. And then I think another added thing is, is your product very niche? Like, does it support like a particular vertical, like someone in, you know, legal is the only person that can use your product or is, does your product have a very horizontal, uh, user base, right? Mm-hmm <affirmative> people in sales, marketing, uh, product managers, engineering can use it then that also helps make, uh, PLG a little bit easier to spread.
Speaker 2 00:20:34 Yeah. Yeah. I think when I'm sorry, go. I just wanna jump in on the freemium comment real fast because I, I think a lot of times, you know, optimizing that freemium flywheel is a lot more complicated, but the, the benefit, um, as you said, I mean, it is, it is more broad reaching, but it's also way more defensible too, that, um, I mean you're starting to see as something like Salesforce where HubSpot essentially came out with with a free version of everything and, and at least the appreciation and value has been a lot more on HubSpot in, in recent years, I think as a result of that. And I think there's a number of categories where, um, when you have a really good free offering that you do end up capturing market share, but I that's, that's really fueled by not just virality, but I would just even say pure, pure word of mouth. If, if people have a super valuable experience with the product and it's free, they talk about it, they share it with other people. And uh, and so it's yeah, your allowable acquisition cost how you figure that out is really hard where you have touch and, and no touch is really hard. Do you have trials on the premium upsells? Like getting, getting all of those pieces right. Are hard, but, um, but I think the model's super powerful if you can get it to
Speaker 3 00:21:48 Work. Yeah. I think you made, you know, you made an interesting point, Chris, that I maybe a subtle one, but that virality takes time. I think we think of, you know, incredible viral examples, like the Dropbox example. Right. And you think about it just happening, like, you know, just like it's just a roller coaster, but the idea that virality actually takes time and you have to work to, to dial it in, I think is probably an important one. Even though the loops may ultimately be very fast getting to the point where those, where you're optimized to drive those loops might, might be a, a bit complicated. So I thought it'd be interesting to kind of understand in your eyes, uh, how somebody goes from, you know, their first ENCA awareness with, with of fireflies. Maybe I'm in a zoom meeting. Like I, I think I told you, this is exactly what's happened to me as I was in a zoom meeting. I see fireflies, I don't know what it is. And I'm like, what is, you know, and then I'm after it's like, oh, it's a note. It took notes for this. This is pretty interesting. But can you tell us sort of the path from that where a customer just becomes a raving fan of the product and eventually it spreads into their team and maybe to other teams.
Speaker 4 00:22:52 Yeah. And I think just want to, uh, wrap up on one previous point about the whole freemium free trial PLG thing, um, which you two points one is, uh, usually when you have that motion, uh, you can spend more resources and energy on building your org around growth and engineering. Whereas like the other companies where, you know, getting leads is their number one priority because they don't have that motion. They're gonna have to spend a lot more energy resources on marketing and sales and you can have the same type of product that's sold in different ways, right? Like the Dropbox versus box example where Dropbox really embraced the, the freemium and the viral loops, whereas box like realize, okay, we're gonna go top down enterprise, hire a lot of sales people. So at the end of the day, it'll eat into margins either way, the cost of supporting that freemium base, uh, or that cost of hiring a sales force and going after that. And so sometimes you're gonna need both like slack needed both right. Yammer needed both. So it's, um, it, it is like the, how your org structure is also built. It will be heavily, uh, decided based on that. Um,
Speaker 2 00:23:54 Yeah, just, just comment real fast on that. The, you know, Dropbox actually, uh, when I was there totally no touch everything and more and more has shifted to B2B sales. So even, even in that case, like the, the model can evolve.
Speaker 4 00:24:09 Yeah. Even I think Jason Lemkin says like eventually every company that's growing will shift up market or a nudge, uh, forward up market. Uh, but the good thing is some of the companies that build the foundations right at the PLG motion, like that's a loop, it's like an evergreen cycle that can keep going. So it's much harder, I believe to go from like enterprise and then down to building a PLG motion versus PLG moving up. Um, and the other part last part that people have to consider is when we were a small team and we were doing PLG and all of these like free users were coming in, like the cost of supporting them having 24 7 support, uh, you know, these guys aren't necessarily generating revenue for us, but like you have to wonder, like, does it make sense to support all these people, whereas like competitors and the enterprise are maybe dealing with 100th of the number of users, but every user that they're dealing with is generating them revenue so much easier. Like if you're an early stage, uh, company to like pick your battles, right. So this is always
Speaker 2 00:25:12 Trade offs. We, we did a lot of that assessment at log me in and, and, uh, ultimately we, we realized that those support costs were a marketing cost that fueled that flywheel, that, that, um, we, we determined that that was the better way to go. And ultimately ended up acquiring the company that took the alternative premium only approach, which was, you know, the maker of go to my PC and, and, you know, merge the companies together. But I, I, you know, I, I do think that if you can get it working, it's really hard to compete against then, uh, that's, it's a good place
Speaker 4 00:25:44 To be. And then cost is another thing. Uh, for example, voice over IP, like those sort of areas are very expensive. Um, and you know, transcription can be expensive and we continually work on the infrastructure to lower those costs so we can support the premium tier. But if you look at zoom versus Google meet, uh, I mean, Google has a lot of resources, so they can do everything in the browser, but it's actually very expensive, uh, right. To do Google meet, whereas zoom encourages you to download the desktop application. So it can offload the cost to your laptop if a lot of people realize, and also you can have a better experience on, um, on like the desktop. But I had wondered in the early days, like, isn't it so expensive for zoom to be doing this? Um, so that must have done incredible work on their infrastructure.
Speaker 4 00:26:26 Uh, their cloud compute costs all of that in order to support that free and tier as a result, they could take all that marketing budget and then put it back into R and D. So that's also one of the reasons why we decided on it, but yeah, Ethan, so shifting to, uh, kind of thinking about the customer life cycle, that's literally all I've been thinking about these past, uh, couple months now that we have different personas of, uh, customers. So our typical self-service funnel, usually a large portion of our customers come through because they were a participant on a meeting with fireflies, or they got a recap meeting recap afterwards from fireflies, uh, or they heard about it from someone. So that's like where a large source of our traffic comes from. Uh, and people end up maybe seeing their recap, the notes of another person's meetings that they were on.
Speaker 4 00:27:14 And then afterwards, uh, they will sign up themselves and then they will start having fireflies on one of their own meetings. Uh, they'll exhaust their free credits, like transcription credits or hit their storage limit. Uh, there's like certain paywalls, or they wanna turn on like a CRM integration that that's only available on the paid tier. So naturally we allow people to self select themselves into these, uh, these buckets. And then, uh, we will have a forever free plan. I think that was the goal when we had started. Uh, but a lot of the people that do decide to convert will convert within the first seven days, 14 days, first month. Uh, and then you have that long tail, the, the last bit of that, that will convert over time, but we don't treat those users any less because that person could theoretically be getting fireflies or, you know, marketing dollars in a sense, right.
Speaker 4 00:28:07 Like getting it exposed to maybe, you know, 10, 20, 30 people per month. So imagine if I had to pay that same for ads spend. Right. Uh, it would be a very, very, uh, you know, we'd have to perfect our marketing budget and that's sometimes a downhill battle, especially when you're going through a recession or you're about to enter a recession. That's where, like the first thing VC's worn is like, Hey, make sure you control your sales and marketing spend. So today we don't spend any, um, money directly on ads, uh, for say, uh, we've experimented with a little bit with like search and here, but almost like $0 on advertising. It's all been like the PLG motion. But I think recently another thing that has been going on is like building community. Like that's something we've started seeing like a lot of really good com uh, companies do, um, communities that help each other out on how to use products and, uh, understand the, the different use cases and values on those.
Speaker 4 00:28:59 So those are two areas we're exploring, but yeah, to your, uh, to question, usually they'll experience it, uh, on someone else's meeting, they'll have it in their own meetings. And then the activation and aha moment is really like, you see it on two, three meetings, you search back, you review it. It's, you know, like it's, it's an instant thing. And then just, it's just a matter of some people having the budget to upgrade soon. Some people taking more time to upgrade. And then in the past year, the magic has been they're like, we tell them, Hey, there's other teammates from your company using fireflies. Let's get them all to start using it in a central workspace. And, uh, and then try to explain the team benefits of using fireflies or aggregating everyone into a team. And that's
Speaker 2 00:29:40 A premium set of functionality for that
Speaker 4 00:29:42 Free users can build teams, but, uh, they'll use up their storage quickly just like slack, where you'll use up your search quickly, the larger the team is right. So we wanna always tie to utility in that sense. So, um, and then there's a, uh, utility differentiation between the free and the paid tiers. There's also a functional differentiation between the free and the paid tiers. So like our business tier comes with a full suite of conversation, intelligence and, uh, uh, voice analytics, all these things that some of the folks that come from some of the enterprise products that are out there in this market, typically for sales, where they're very, very expensive, they'll be charging 150 to $200 per seat. Uh, they'll have upfront license commits where you have to pay a minimum of $15,000. Uh, whereas like with fireflies, that business here is probably like $20 per month, right?
Speaker 4 00:30:31 So where a lot of SMBs, a lot of small teams that don't want to go and just drop 20 grand right away, they wanna like be able to start getting off the ground for a couple hundred dollars. Right. So, and pay monthly. So that's another thing we realized is like, allow people to pay the way that they want to pay monthly or annual. Um, a lot of people do choose annual and we discount for annual as well. But yeah, in general, that's been the philosophy, let people buy the way they want to. And if they like it, they will keep coming back. Uh, and, uh, yeah, it's, it's, that's been the progression of just following the customer journey.
Speaker 2 00:31:05 You mentioned, uh, that you've thought about trials. Have you have actually done any testing with trials?
Speaker 4 00:31:10 We haven't yet, but when, uh, people request a demo or want to talk to, uh, our customer onboarding person. So we have something similar to superhuman, uh, where, you know, we have, uh, when someone has a question or raises their hand and it, it requires something beyond just support. Uh, they'll hop on, they'll be like a customer success customer onboarding manager that will kind of educate them. These are usually very qualified leads, right? Inbound leads, product leads that, uh, are either have 70% of the knowledge about the product or tinkered around with it, or they've read about everything enough, cuz everything is transparent on our, uh, website from pricing to all, all those things where when those guys come in, it's less a question about negotiating pricing and it's more about, uh, uh, how do I use this? Like how do I set up? Like how do I get my like, you know, teammates on? So in those sort of situations, uh, they will usually like request, Hey, can I get a, you know, a seven day free trial of the full business tier so that I can test it out with my team, right. Individuals don't need it, but teams definitely do, uh, that are using
Speaker 2 00:32:15 So that additional functionality doesn't help the individual really,
Speaker 4 00:32:18 Uh, the, it, it can help, like in terms of they need more credits or they want analytics or they want to turn on their Salesforce integration, but it's, uh, it becomes more valuable.
Speaker 2 00:32:29 It's more of a team thing.
Speaker 4 00:32:30 Yeah. Especially teams that are trying to solve specific use cases like, Hey, I have this sales thing that I'm trying to solve your business here. Does all this stuff, this, this enterprise product does. So, um, I just wanna validate that really quickly. Right. And so that's where the free trial. So we do do the free trials for people that are talking to us. We might roll it out more broadly where, you know, people that reach a certain stage in the product, it's like, Hey, you've unlocked a free trial, like seven day free trial. So we might do that like a freemium, uh, plus like a free trial offering. So there's experiments to be run. Um, you know, there's probably things we gotta think about like, cuz today we don't ask for a credit card up front. There have been a lot of studies where like, Hey, you know, we, we switched our whole free trial motion to asking for a credit card up front and opt out within like the first 30 days. So I think that needs as much studying as like freemium.
Speaker 2 00:33:19 See, I, I, I think with freemium, with the credit card, my what's worked best for me is, um, not to have a credit card with, with the freemium, but to, uh, and this is just, you know, a couple of experiments, but worked really well where there was a free trial cuz there's that incremental functionality and um, and giving people, uh, when they come in to sign up for the free product, you say, congratulations for the first 30 days, you're gonna get full premium functionality. If you don't want this, you can go into preferences and turn it off. But after those 30 days, you'll go back to the free version and won't have to do anything. And I think what's, what's good with that is that it drives discovery of some potential premium features that they didn't even know they needed until, until they went in there.
Speaker 4 00:34:06 Yeah. That's actually something I noticed with slack. I set up a new slack workspace, uh, like the other day and the onboarding has completely evolved since when I had previously set up a few years ago. Mm-hmm <affirmative> whereas I signed up, I invited a few people and the moment I invited a few people and like started getting engaged, I get an, uh, automatic email from slack saying, Hey, you've been upgraded to the premium for the next 30 days. So where they're opting you into. Oh, interesting. Yeah. Uh, or the, the premium plan, uh, for immediately. Um, so you'll have the free, free paywalls, but in the first 30 days they wanna really get you to use it. So it made sense that, uh, they don't wanna limit you or PA gate you early on. So
Speaker 2 00:34:43 First time I tried that for what it's worth, I had a, uh, an attorney reach out and say, uh, that's bait and switch where, you know, and so that's why we added the messaging that you can go to preferences and disable it if you don't want to.
Speaker 4 00:34:56 Interesting. Yeah. That's cleared up.
Speaker 2 00:34:58 That's crazy.
Speaker 4 00:34:58 Cause you have a free tier and you're offering more with the free trials. Yeah.
Speaker 3 00:35:02 Yeah.
Speaker 2 00:35:02 It's funny. But you're like forcing them onto the trial.
Speaker 3 00:35:04 It's it's funny cuz I, I remember, uh, use, I I'm, I'm pretty sure you're also one of those experiments you're talking about Sean is when you're at log me in. And I remember using log me in pro and having exactly that experience where I, I, I, I wanna say it's, it was like 15 years ago, some long time ago, but I wanna say like remote printing was available in the, in the, that trial period. And as soon as that went away and I I'd used, as long as you have U had used that feature once or twice, the next time you go to you're like, well it's a must have I, I can't live without it. And you're like, got
Speaker 2 00:35:38 Yeah, exactly where you would've never thought that was important until you actually used it. So yeah, that can be a good,
Speaker 3 00:35:43 But I think it's so much of this and I, I think probably Chris, you know, as you, you keep referencing experiments and I think it comes down to so much as you're trying to learn about the customer and what their experience is, you have to like really like test these, the, the limits of these things different, you know, differently. You have to really think about different use cases. You have to really, so I think it's interest, you know, going through your, the, the sort of point a to point B, what, what the, the original, like what, how a user goes from awareness to becoming a fan. It's interesting because you start to think about all the experiment points along the way to get to that perfect customer loop where they're, you know, a huge fan and they're, you know, they're telling friends about it or, you know, telling others about it or at least, you know, when someone is in a meeting with them they're and says, what is that? You know, they're evangelizing it for you. So it's an interesting,
Speaker 2 00:36:34 Well, just, yeah, just to one, one thing on that though, is sometimes it's not always just about experiments. It can be also about observation and, uh, and so it's not even an experiment, but you just get these surprising use cases that you didn't know. And so have you, have you seen anything like that, Chris?
Speaker 4 00:36:48 Yeah. I think there's a lot of experimentation on pricing that we've done. Uh, one of the things was way back in the early days we thought, Hey, there's a, uh, operating cost to running transcription and VoIP. So why don't we make a utility based pricing where, you know, instead of paying upfront $10 or something, they're like literally paying cents per minute. Right. And it seems good on in theory, like it's the AWS model and that completely changes the way you build the product. I think there's a lot of money that can be squeezed outta that approach. But the problem is you'll start seeing immediately people, behavior change, people
Speaker 2 00:37:23 Don't like surprises. People
Speaker 4 00:37:25 Don't wanna like, or the amount of that they use it because what people love is like, I want like a buffet, I pay one price. I use product for how much ever I need. Um, and sometimes it's hard, right? Cause like, even with like storage or something like that, you have to put some sort of pay wallets, but at least they want something simple. Um, whereas when you do this, like AWS style utility pricing, and I've seen some competitors and other folks try that as well, and I've learned my lesson from it, but it's just really interesting where, um, their behavioral will change and then their engagement will drop. Right. And they will be much more conservative in terms of how they use it. This is like the same thing with, um, phone plans back in the day, right? Like pay per minute, pay for text. Uh, and then everyone started moving towards unlimited, right? Like you get unlimited this unlimited that, um, and then they try to make money in other ways, right. When they sell you the iPhone, the, and like, uh, the service plan, all, all of those things.
Speaker 2 00:38:17 So, and, and you look at like generationally, how that's affected use of the phone once they move to unlimited, like the preferred way to communicate for younger generations is through text where older generations still call. Cuz that was, that's kind of how they learned how to use a phone. Yeah.
Speaker 4 00:38:31 I remember when I got first got my phones like, oh, you know, you get unlimited after like 6:00 PM or 7:00 PM. Or if you were trying to call the east coast, you gotta do this. And now it's like, oh, let me just make a WhatsApp call or let me just make a, a call through, uh, you know, so like one of the digital or like cloud based services where it's like free, cuz like you have data or you have wifi. So it's just very interesting. I think cost plays a big role as well. Um, I've said this, uh, one time in the past where, when Gmail first came out, uh, it changed the game by giving people like one gigabyte of storage before that people were getting like storage in the megabytes. Right. MB. And so they'd every week they'd have to basically delete or clean their folders.
Speaker 4 00:39:14 But now that you have a gigabyte of storage, I, or more, I can search back to a conversation I had three years ago, four years ago. So the utility and the value of the product also increases. And we saw the same thing with fireflies, like as we increase the storage limit for users, um, and as more teams get in, so one of the additive advantages is you get more storage. If you add more teammates. And now like now that you have more information, like fireflies is great. Like the first week you start using it, but it becomes more valuable, you know, after the first three months or four months, cuz now you have all of this like important information that you can search back
Speaker 2 00:39:50 Through. I think that's even what drives upgrades on slack a lot of times is that once you start using a lot and then you want to go back and, and search old conversations like, oh you're on the free plan. You, you only have so much old conversations you
Speaker 4 00:40:03 Can search. That's what got us to upgrade. Yeah. <laugh> from like we can theoretically forever use slack for free to crap. I gotta search that message in that channel. I need it right now. Yeah. Um, it was like a link. I can't find it on my email. Like if I don't have that link, I can't go to my next meeting. I'm like, all right, let's just upgrade.
Speaker 2 00:40:18 Yeah. And then that behavior kind of reflects well on your product that if, if people want an archive of slack conversations, why would they not want an archive of audio conversations?
Speaker 4 00:40:29 Yeah, absolutely. Yeah. That's uh, that's one of the, the levers, uh, with storage as well. Uh, we took out this, uh, the minutes part as well, like of how much transcription you get per month. Uh, just so that people are gonna be like understanding of like what it is that they need to use. And we've made a lot, we've done a lot of work on our speech engine, uh, on, uh, you know, like all of our, like, like scaling our cloud infrastructure. Uh, I remember a Dropbox, um, right before the IPO. I think that's when they moved off AWS to building their own data warehouses. And because cost eventually becomes like a, like a big factor in these spaces. Um, you don't wanna keep like, growing so fast that you like burn so much on your like cloud spend bills. Right,
Speaker 2 00:41:12 Right, right. So Ethan, did you have another question? I've got one here, if you that want you go, okay. I didn't wanna be jumping in front of you. So, um, I, I'm curious, uh, over time sort of, you know, in the beginning, you, you obviously have no clue how quickly the business is gonna grow. It's all just based on a set of assumptions, but I'm curious over time, how much you've seen, uh, kind of predictability come into the business where you, uh, like, like, are you able to, are you able to now be pretty clear on, you know, what, what is next month's sales gonna be? What is next quarter sales gonna be? And you know, and if so, like how, how are you doing that? Are you, do you have a model that, that has mostly knowns now and the, the slice of unknowns become smaller and smaller? Um, but just how do you think about, about sort of that predictability? This is,
Speaker 4 00:42:05 This is a constant work in progress cuz uh, you think you have all the variables under control until new variables get introduced. Um, even like variables such as, uh, with like the whole market slowing down or freezing. And how does that impact, uh, people's appetite, right? So there's many, many external factors that you can't control for. Um, you know, maybe you have a PR piece that goes out that that leads to a lot more, uh, attention. Or we roll out a feature where we think not many people are gonna use it, but let's just roll it out because it's something that's important. We believe that's inherently important. And then you get a few customers saying like, this is so important to me, this is the reason why I convert it or this is like, I need this to do like, you know, go to the next stage.
Speaker 4 00:42:46 There are people that, uh, you know, like need to take like similar to like the remote printing, uh, thing, like people need to export data and do interesting things with, uh, with it. So, uh, yeah, we we've understood like it's really about following that customer journey, getting people to be engaged to a certain level. Uh, and then from that engagement sharing, uh, usage, like those are kind of our core pillars, like activation sharing, uh, repeat usage, uh, some things that we've started to learn a little bit more about and improve on is, uh, how do we get people to get more value out of the platform? Cause I've seen so many people where I have conversations with them or like fireside chats with our customers where they're like, I love the product. I use it every day. I do this. And I'm like, oh, have you looked into this feature on our platform?
Speaker 4 00:43:37 Uh, that might help you do this better cuz you seem to have these type of meetings. And they're like, oh I did not even know. And uh, let me start using that. And then like, oh, let me tell my teammates to start using that. Or let me tell my support team to start using this. And so I had this conversation yesterday and I'm like I realized there is people are coming in for that AI note taker. Uh, but there is like so much more on top of the platform, uh, for them to discover. So the discoverability is something that we think about. I mean, Facebook had this challenge, like a lot of other companies had this challenge in the early days. It's like you could do so much more, um, on top. And if you, and we realize like the value of showing people more, when they're ready also naturally leads them to spend more and uh, upgrade faster. And so,
Speaker 2 00:44:21 So it's early. Sometimes you can overwhelm 'em with complexity. So it's getting that timing. Right's important.
Speaker 4 00:44:26 That's something we still have to figure out. But yeah, when, when you talk about variables and the core usage core use case and you know, uh, levers that drive that, uh mm-hmm <affirmative>, it becomes systematic like, uh, in terms of we know that some conversion rate's gonna be X mm-hmm <affirmative>, uh, and if we hit this sort of traffic, we can almost go down to predicting down to how many people will become paid users. Right. Mm-hmm <affirmative> so it's, it's, it's a funnel.
Speaker 2 00:44:50 And unfortunately like from, from a hitting revenue targets kind of month to month in SAS over time, you start to see what that churn rate looks like. You start to see enough pieces where, where the model does have some predictability acquisition side might have, have a little bit less, but yeah, it can, it can be, uh, it, it can be a challenge there, but you're are you seeing that predictability now or
Speaker 4 00:45:14 Yeah, for sure. You, you see it during the holiday seasons, uh, compared to like the previous year, you see it versus, uh, previous quarters, uh, certain months. Uh, and so I'm always a believer that I look at data, uh, into the past and say, okay, I see where the trends are, but I never project forward because you don't know what variable could change, what, what things can happen, but you know what to expect and you start building a baseline or benchmark, okay, this metric is down. That means these other metrics are gonna be impacted. These are lagging indicators. So yeah, like that can be scaled all the way from sign up to like revenue. So there, there is a thing, that's what all, that's the holy grail of every growth team. Um, and it takes a very long time, right? Like it just takes such a long time where sometimes, like I tell people, advise people really, you want to do that or just go like monetize with a fixed thing.
Speaker 4 00:46:05 Some people don't even have a free tier. They'll say just hop on a demo with me. So what I will say is if you have really good sales people, you can almost bandaid and just leapfrog all those other things you have to do. If you like are not a fan of sales or you don't wanna deal with sales. So I've seen like a lot of amazing YC companies, uh, you know, O other companies where they're like, talk to us, like just hop on a call. Uh, and then they just get them to become a paid customer right away. And they just skipped all of this. Right. So, yeah. <laugh>
Speaker 3 00:46:36 So when you say you're never projecting forward, um, you're not saying I'm assuming, you're not saying that you're not setting goals or, or specific growth objectives. What you're more saying is that, uh, what, what you're not gonna do is try to predict what the, how the unknowns are gonna drive you there. You're saying, this is what we want to accomplish. Let's do these things and see if we accomplish that. Is that more of the approach?
Speaker 4 00:47:01 Yeah. Cause when you look at, uh, hockey stick curves to the right, like, and when you see those VC decks, like that's what people say, but realize, I think it's more of like step functions where you'll have some plateaus and then all of a sudden you make some change or experiment. I was like, wow, that increased it more than I anticipated. So I, we have goals in terms of what we're trying to hit. Uh, but it's the predictability of okay. Cause we're making so many changes, so many experiments, we don't know which experiment is gonna cause that next step function bump. Right. Or which sometimes experiments cause negative bumps we've had that as, or like drops as well. So I think of it as more of a step function and there are certain things as you experiment, uh, and get more rigorous about it and have a really disciplined culture around it where it won't even be like the hockey sticker it'll be a straight spike up and then you'll figure out like, okay, we've kind of maximized that experiment, so it's gonna plateau.
Speaker 4 00:47:54 And then you find the next experiment. So I kind of look at it that way. Um, and so are we in like that, uh, step function, jump phase or are we kind of like in a stabilized phase? Right. So as long as you're step function is moving up into the right, you're fine. Uh, but that's why I'm like, it's, I think companies that are much like, you know, bigger, been around longer, like zoom could have predicted like kind of predict their volume on their like freemium on their paid on their sales funnel. Yeah. Yeah. So, yeah.
Speaker 2 00:48:21 So yeah, it definitely feels like there's an evolution where the early days is just a bunch of guess work and then yeah. Eventually you, when, when you're a public company, there's just, there's just so much historic data that, that, that the business becomes a lot more predictable and that's why they can, they can kind of give, give guidance to, uh, analysts and investors, you know, months in advance. But, um, in that middle stage, I think a lot of it is trying to just get a handle on, on, you know, collecting up those knowns and, and being able to build sort of a predictable growth curve on the known things. And then as you said, the experimentation, there's such a, such a, a swing that, that a great experiment can give you. Um, but, uh, but yeah, it's,
Speaker 4 00:49:03 Uh, I always wondered though, like, cuz like for growth, uh, there could be some things that like go off and then that, that change the complete like, you know, scale of projections that you go. Whereas when you hear a lot of the analyst calls and for these larger, larger companies, it's usually like what their predicted revenue is. And there's a lot of factors that go into it. But usually when they're very sales led, you know that if you have X head count, you have a Y close rate and then you, it's a very predictable of like how many people do I add? How much money do I spend? How many leads do I get? And can I repeat that process? And so when it's very fascinating to me when like VPs of sales are able to like talk about, yeah, we're expected to close this and I'm like, like based on what heuristics <laugh>, uh, versus uh, you know, with growth, it's much harder because, uh, there's just so many signals, so many different factors. Uh, but again, I always wondered can, like, it's amazing how some sales companies can grow exponentially versus linearly. Uh, and they're also optimizing their pitch. Like it becomes easier to sell. So now think about a PLG company that has to manage that freemium, the growth, uh, levers, the sales levers, the expansion levers, uh, you have to measure a lot of different things versus if you're like, uh, you know, one type of, uh, approach mm-hmm <affirmative>, uh, where you're just like a sales led motion. So I believe,
Speaker 2 00:50:21 Yeah, I feel like sales led motion actually can be scarier in the sense that you've got this great salesperson that you lose. And then <laugh> where at least, at least, you know, there, there's almost less variability in, in, um, product driven growth, where if you've been converting at this rate for the last three months, there's a good chance that next month in that channel, in that flow, you're gonna convert at the same rate as well. And, uh, but yeah, obviously, but even, even on, you know, sales forecasting is, uh, something that, that weighted pipelines and that, you know, there's whole formulas that have yeah. That I've gone into for a long time.
Speaker 4 00:50:58 <laugh> yeah. More about, uh, the other thing is even for growth, you can find channel saturation. Um, so for example, the Airbnb class example where they used, uh, Craigslist as a way to reverse, uh, list on their site, uh, you'll have certain things. So when people use the word growth hackers, like people think, oh, it's like these one off hacks, but actually the real growth hackers, the real growth engineers are building these long term sustainable loops. But for the startups that are kind of listening to this, like, uh, early stage startups, sometimes it is really like one of those hacks that gets the ball rolling
Speaker 2 00:51:32 For you. Right. Right. There's nothing, there's nothing to optimize if you don't have any traffic. And so, you know, a lot of times it does take that big idea that that gets you the, the flood of people coming in to where you can start doing a more rigorous experimentation approach.
Speaker 3 00:51:46 So, Chris, we always wrap up with, uh, one final question, but I actually wanna ask you, um, it sounds like fireflies has grown really quickly and, and seems really like just a fascinating business with, uh, you know, I think it's much more than just this AI note taker that some of us know it as, uh, are there, uh, key roles that you're trying to fill at this point, um, that our audience might be interested in?
Speaker 4 00:52:07 Yeah. So surprisingly we're hiring our first, uh, sales people. <laugh>
Speaker 3 00:52:12 Okay.
Speaker 4 00:52:12 Segue into this. Uh, so we're hiring our first AEs and then requirement there is we have all these qualified leads, uh, and there's people inside large organizations, fortune 500 fortune, 1000 that are using fireflies. Um, and we just wanna start having conversations and seeing how, uh, we can turn those excited individuals in an org and then turn that into, uh, a full-fledged, um, you know, maybe the sales motion learning that, and that'll be an experiment in itself, right. For very similar to our growth experiment. So for everything I've said about PLG <laugh>, we are looking into that
Speaker 2 00:52:47 As well. That's where the, where the invests going <laugh>
Speaker 3 00:52:50 <laugh> but no, I think that's, I think that's been one of the key key points here is that, uh, there is the companies evolve and they, as you grow the, even if, if PLG and self serve is, is a super important part of how you approach the business, it may have to evolve over time to meet, you know, to keep making those experiences for your users. What they're looking for. You've said a couple times here, you want to let customers buy the way they wanna buy. And I think that's
Speaker 2 00:53:14 That's. Yeah. Well, and the product qualified lead piece, I think is really the, the important part there. Yeah. Is that, yeah, there's a touch model, but it's not, it's not lead gen and cold calling it's it's it's
Speaker 4 00:53:28 Ultimately, or I have emails and yeah. So it's, it's because I've had a few conversations so far with people I'm like, wait, so you don't have this SDR function. You don't need to do this. I'm like, no, we have these leads. We just don't have someone like going after them. And they're like, wow, this is like, this is a dream come true for any sale. I'm like, really, like I thought every company would do that, but <laugh> yeah. Cause it's amazing that there are other companies where they're able to create, you know, sales through cold email, cold outreach, cold phone calls getting, uh, time schedule on the AEs calendar. So yeah. I'm hoping that it's a fun role. It's our like kind of a new way of building out our GTM sales motion. Yeah.
Speaker 2 00:54:03 Um, as a, as a former cold call salesperson myself, that sounds way more fun. <laugh>
Speaker 4 00:54:08 <laugh>
Speaker 3 00:54:09 So Chris, we always ask, what do you feel like you understand about growth now that maybe you didn't understand as well, a few years ago?
Speaker 4 00:54:16 So I think growth is a repeatable process. It has to be a sequence of experiments. Uh, a lot of times, uh, when I read about growth hacking was about like you finding this like silver bullet, there is no silver bullet, uh, it's consistent iteration. And, uh, also you move what you measure. I'm one of the biggest believers in that you can have hundreds of signals, hundreds of metrics, but really pick one metric. And then that metric will usually have a lot of lagging indicators that'll determine the rest of your business. So, but if you're not measuring that, uh, you're not gonna think and make, uh, take action, uh, in that direction without measuring for that. So you move what you measure and then it's all about repeatability and experimentation. Yeah.
Speaker 2 00:55:00 I love it. That's that's good stuff. So good. We, we won't, uh, dilute that piece of advice as we, as we wrap up here, we'll leave it on that. And so thanks, Chris. We really appreciate you sharing the Firefly story and we're excited to see where you go from here.
Speaker 4 00:55:13 I had a lot of fun. Thanks.
Speaker 3 00:55:15 Thank you.
Speaker 5 00:55:21 Thanks for listening to the breakout growth podcast. Please take a moment to leave us a review on your favorite podcast platform and while you're at it subscribe. So you never miss a show until next week.
