Welcome to episode thirty-two of the SaaS Backwards podcast, where we interview CEOs and CMOs of fast-growing SaaS firms to reveal what they are doing that's working, and lessons learned from things that didn't work as planned.
You can listen to the full episode directly below via Spotify, or visit SaaS Backwards on Buzzsprout or wherever you listen to podcasts.
Soul Searching to Scaling in Short Order
With Krish Ramineni, CEO of Fireflies.ai
Edited for clarity and readability
Host, Ken Lempit:
Welcome to SaaS Backwards, a podcast that helps SaaS CMOs and CEOs to accelerate growth and enhance profitability. Our guest today is Krish Ramineni, co-founder and CEO at Fireflies.ai. It's a SaaS that's on a mission to transform conversations into actions. It's voice assistant and transcription tools are designed to automate your team's workflow. Before we get going, could you just give us a little bit of background on you and make real that company about us?
I'm the co-founder and CEO at Fireflies.ai and for the last three to four years, we've been building and scaling up this notion of an enterprise voice assistant or AI meeting assistant. A concept of digital assistant meeting assistance isn't new, we see them on our phones like Siri, we see them in our homes like Alexa, and our idea was what if we had something in the workplace that helped us stay on top of things, remember all the important conversations we're having. And one thing led to another and that started the origins of Fireflies as it is today, because meetings are not only very time intensive, but they're also costly and we tend to have a lot of them so we build Fireflies. It's an AI note taker. It follows you around in your meetings. It records, transcribes, summarizes, makes all the meetings searchable afterwards.
Fast forward, we're also doing a lot more stuff around insights and intelligence for teams, and we've evolved a little bit in that nature. So, that's a little bit of background. Prior to Fireflies, I was a PM at Microsoft, got to work with Skype and a bunch of other teams inside Office. I was not even supposed to be doing this startup, I was supposed to go to grad school, but got pulled into MIT to visit a friend and then before we knew it, we said, "You know what? Let's just do this. I can go do a business school or MBA later. This is way more interesting." And so that's what led to the origins.
I don't do too many predictions on this podcast, but I'll share with you what my grandmother said to me quite some time ago. She said if you don't go to grad school right away, you may not go, so we'll have to check back in with you and see how you do on that.
I'll have to let my mother know that because she was definitely bummed out when I decided not to go to grad school.
Tell me a little bit about that founding, how you met your co-founder to go into this venture. I think that's always interesting, the selection of a co-founder, even if it's someone you know before. So it'd be great to hear how you decided to go ahead.
Sam Udotong is my co-founder and CTO of Fireflies. I've known him since college. We actually started working on several different projects while in college. The actual name, Fireflies, was used in many of those hackathons and projects that we were part of. The original founding name for Fireflies was actually Working on Drones and we wanted to do drone delivery, and so when you have drones flying around at night, they look like fireflies so that was the name origins of Fireflies. Nothing to do with SaaS, productivity, software at the time, but we were too lazy so we just continued using that name for all the different projects we ended up doing. Sam and I worked remotely for two years. We met through mutual friends. We worked on various projects. We never even met in person for a long period of time.
And so we were just working remotely for two years while he was at MIT, I was at Penn, and then one day when I was telling him about my Microsoft experience and he was suggesting why don't you give a little bit of time--fly out to Boston, I'm graduating and we'll just hang out and work on different things. I flew out to Boston and we got some co-working space.
A VC was fortunate enough to give us some free co-working space. We worked every day, we treated it like a job. And then by the end of the summer, I'm like, "This is actually a lot of fun. Let's just do this full time." We didn't even know what we were going to work on. We were working on NLP related stuff, but not necessarily voice. So we were building chatbots, Chrome extensions, email assistance, all these other things, we were consulting.
The first two years though was really difficult because there was no plan in place and we were these rookies that didn't have a lot of work experience. In my case, I only had a year and a half of corporate experience. Sam didn't have any. And so we're trying to navigate this entire SaaS space.
At the end of the summer, we had something and so we said, "You know what? Let's fly out to San Francisco," (where I'm originally from) and we decided to go heads down. We were bootstrapped for quite a long time. My co-founder has this whole post on Medium. He wrote a blog about how the first year he lived off Soylent and Domino's Pizza and made ends meet--literally getting free furniture, subletting doing everything possible. It was a grind in the early days.
It's like the post-college rock band turned into software hack. Good for you guys that you got where you are. That's a great founding story. And I think it's rewarding and valuable to talk about the commitment it takes, especially if you're very early. It's your first thing, you have to be willing to make a personal sacrifice in a number of ways in order to start your own business. I think that's good visibility into what it actually takes. You had this initial product, can we talk just a little bit about how you went from this initial germ of an idea to the... I don't know if you called it an MVP (Minimum Viable Product), but your first product and starting to generate revenue, how did you make that transition?
The first version of the product, we were hacking things together with Twilio and a lot of things off the shelf. In the beginning, we used to always pivot on different ideas and then we'd spend too much time writing code before validating the product itself. And so we said for this time, we're not going to write any code, we're going to first validate it with 10 of our people, and we're going to see if people are willing to pay for the product. So the early days we literally had a note taker email that people would email and then we'd have a human in the loop process, and we would tell them, "Hey, a human assistant will join and take notes for you." We did this for 10 people or something like that to validate the need for it. And people were very like, "Yeah, I would totally do this."
And then we had to be realistic because a human secretary is way more expensive than AI secretary, we know there's a business use case here, people will be willing to pay for this, now, the software, the engineering problem is the hard part. Before, in the past, the engineering part was easy, finding the customers was the hard part. Here, it was the reverse. We're taking on a very hard challenge, it was a no brainer that someone wanted it, but can we actually build. So that initial idea turned into, okay, how do we start building out the product? How do we improve the speech recognition? How do we get the accuracy? How do we make it seamless so that it's like another agent on the meeting? All of that led to the initial versions of Fireflies. So 2018 to 2019, we were doing these MVPs, hacking things together.
We went through an accelerator at that time. So it was just me, Sam and then we had a few other engineers that were helping us part-time.
And 2019 rolled in, we were confident in the technology, we were confident in the way we were deploying the technology as well on major video conferencing platforms like Zoom and Google Meet and Webex. And then finally, 2019, end of 2019, we raised our seed round. 2020 January, we rolled out the current version of what Fireflies is today.
And then a month later, the pandemic takes off. Everyone's going remote. Everyone's having a lot of Zoom meetings, Teams meetings and we started to help a lot of remote teams adopt and use Fireflies to bring all their conversations in one central place like a repository for all their meetings. It was a rollercoaster.
If the first couple years were hard, just figuring out the problem, the next couple years were “Crap! People actually need this, they want it, and we can't let them down.”
All of the sudden it was a responsibility to make sure Fireflies always joins on time, that it works, that you're not missing out on the notes. Every minute counts. So that was definitely another challenge of how do you scale this up, we were on 10 people around that time, to now how do you scale this product up to start servicing all these people that need it? And the last two years has been a roller coaster.
Wow. So from soul searching to scaling up in a short order. There are other products in the space, and in fact, in our Zoom call, one of your competitors product's in the call as well. And we had a good conversation before recording our episode today about differentiation and I think that's really important to talk about here. And I think that's, again, product vision, founder vision on the product. So I'd love to learn about where you are today, differentiating the product and what you see as Fireflies differentiation going forward.
This is definitely an exciting space and I think the opportunities opened up for several different folks. Just like how cloud computing or mobile opened up the space for a lot of people that started building mobile first products or cloud first products, before that was on print. There's a lot of voice first products now and voice intelligence products. As a market, space is growing. And I think each one will have a different approach to how they execute. Maybe the core technology, the transcription, will be similar, or the focus on meetings. We see on the enterprise side, there's a lot more enterprise level products, which are focused on sales enablement, sales use case, they sell the one vertical. For us, we took an approach that every knowledge worker should have Fireflies and it should be deeply integrated into their day-to-day workflow.
So whether that is your sales calls, your internal meetings, your client calls, we wanted it to be everywhere for all of those meetings. And then we had to look at putting some sort of box around what is our target market. And so that's also another thing that leads to a lot more differentiation in how you build a product.
For Fireflies, we're focused on teams. It's a great product for individuals to use, but it's an even better product for teams to use. And based on that, we start looking at the needs of the team.
For example, salespeople want all of their notes automatically logged inside their CRM. They hate doing manual data entry into CRM afterward, so we started building out integrations to eight plus CRMs, all the major guys, HubSpot, Salesforce, Pipedrive, etc. And then we look at project managers and they said, "Well, I want to automatically create tasks in Asana or Trello after a meeting with just my voice command, with just the sound of my voice," And we ended up doing that. So, I can say something like a voice command on this meeting and then in Asana it'll create a to-do for me afterwards.
We’re building Fireflies to be that platform that not only just helps you transcribe and search your meetings and have a repository, but how do we help users and organizations get work done from the meeting itself?
Right after the meeting, Fireflies is focused on the workflows, the integrations and the use cases are really built for business users. We integrate with Slack, we integrate with CRMs, we integrate with all of these project management systems. If I want these notes to be saved into my Dropbox folder or Google Drive folder, I can do that as well afterwards.
So you're making real what you said in the beginning of the podcast about it's your personal assistant, it's like the smart device we have at home, but inside your meeting. And I think that's a real point of departure from my experience with the tools that I've been using.
It's not just about recording and transcription now, that's table stakes. It's about what can you do on top of the basic capture. And so that's really now you're getting into the Alexa for business. So it's a broader vision, definitely a harder vision to fulfill, but I'm very much a fan of products like Zapier, which lets you connect to any different SaaS tool and automate integrations. I'm a huge fan of Segment, which lets you pull data and then push data into different apps. In fact, the CEO of Segment was an investor in Fireflies.
So our idea was how can we be this voice workflow system for enterprises and for organizations. So that's what really leads to that longer term vision of if Alexa were to be doing this for the business space, what would those look like? What would those skills look like? What would those integrations look like? Instead of just saying, "Alexa, play this song on Spotify," maybe Fireflies in the near future, I can be at a meeting and say, "Fireflies pull up the Q2 financial reports and tell me what went right and what went wrong." So there's a lot that we can do with just the sound of our voice.
So, that's actually not just pushing out from the meeting, but bringing stuff into the meeting, you might have revealed a doorway here in the podcast so thank you. That's a very exciting vision. And I think just unpacking something you said about the notion of table stakes and then going further. So if you're entering a marketplace that could become crowded, you need to address the table stakes, do a good job with those things, but it's the more advanced or useful capabilities. I'm going to have to switch to Fireflies, I guess, because I would love, within a meeting, to be able to send a Slack to someone and say, "Hey, please take a look at the notes," and, "Hey, Kelly do X, Y, Z for our client here." So I think that's a really cool vision for how this kind of assistant can make a big difference in the daily work of people who are all knowledge workers at this point.
We have to be uncompromising on product. Like you said, that's definitely a starting point. And then the go-to-market is just as important. How easy is Fireflies to use and turn on? How do you think about pricing? If you're really building a product for every person, so there's enterprise products for the same technology, they're going to be charging $100 per month per user or $200 per month. And so some people ask why is Fireflies only $10 per month, per seat. And so our thought process there is, well, we don't just want your three sales people to use Fireflies, we want every person in your org to have access to this. So the notion of democratizing the AI assistant was in our minds for a very long time. If you go to large companies, only the execs have personal assistance or business admins. So we also think about pricing as a direct way to this gets better if everyone has access to it, so those are all also important on top of the table stakes.
Pricing, we talked about that, and I latched onto one thing, which was the big difference in your price points between a month-to-month commitment and annual commitment, so you're really trying to encourage people to get on board and stick with you. Have you found that big disparity in pricing doing that job of having people go with the annual deal?
Yeah. In the early days there is no heuristic for pricing. You're just trying to figure out what number makes sense, you do some surveys, you figure out what's a reasonable amount, and then if you look at this landscape, people price all over the place, people price $50 per month, $100 per month, $150 per month for the same technology. It's very interesting because I do stay on top of what the market is pricing. And usually, people try to price based on value or price based on cost and this stuff isn't cheap, transcription, the voice engine, all that, so a lot of people focus on cost as a way to drive price.
For us, we try to do as much work as possible so that any benefit that we get in terms of scale, in terms of compute, any benefit, we want to bring that back to the users, keep it more affordable for them.
Our yearly versus monthly was purely just a way to see an experiment with different pricing. And also with that pricing, what we know is you'll have immediate value from Fireflies within the first two, three meetings that you have, but we have a free tier where a person will have four or five meetings where they’ll say “yeah, I'm sold on this, I need this,” and they'll sign up.
So a lot of our customers don't even need to wait that long in order to upgrade to a paid plan. What we do know is that Fireflies, if you think it's valuable in those first 10 days or first 14 days, it's incredibly valuable after a few months and then after a year, because now you have this entire knowledge base or repository, I can go back to a conversation I had three years ago in 2019 with a vendor and recall that conversation with perfect clarity and then go into that next meeting well prepared.
We had a little feature we're experimenting with where if I had a meeting with someone that Fireflies transcribed, captured and summarized, and I'm meeting with that person after several months, an hour before the meeting it'll tell me, "Hey, Krish, you spoke with Ken, this is what you guys talked about last time. This is your meeting prep before your next thing."
And I go into that meeting and I'm pulling up details that we had talked about, numbers and stats, and he's like, "How do you know this?" And the secret is Fireflies. So the reason for pricing like that is we want to encourage people because the longer you're on the platform, the more value you get from it. And the value increases. It's like a treasure trove of all this knowledge. It's your second brain. So that's one of the reasons for pricing aggressively on the yearly so that people can see that value over time.
Awesome. And so you have the free tier, so you're using a PLG motion, right?
Yeah. Very PLG motion. We don't spend any money on advertising today. Pretty much no social ads, Facebook ads. As far as I can tell, we've experimented in the past, but we've fully doubled down on the freemium model because Fireflies is naturally viral in meetings and the ability to get it to different people.
I was just at a summit conference recently, and then the number of people that come up and then they see my name badge and say, "Oh, Fireflies, I know this. I've used this or I've seen it on a meeting." And that's the start of a conversation. There's no introduction needed.
Also, I'm somewhat of an introvert, so having to introduce myself and pitch my product 100 times, instead it's people already know it, that's great, that's a good icebreaker. We fully believe in the PLG motion. There's been amazing companies built around this, Calendly, Dropbox, Slack. So for as long as we can, we build with that PLG freemium mindset because we want every person to use Fireflies.
There's another little bit of insight here. So PLG is the motion for introverted CEOs, right?
Yeah. Well, people don't think I'm introverted. I'm introverted extrovert so if you tell me to get up on stage and present to 1,000 people, no problem, but you tell me to go to a networking session and shake hands with 200 people and introduce myself, I need to warm up and it gets a little exhausting.
I understand. Can you talk a little bit about the changes in the organization as you got funding, so you went from bootstrapped to a funded entity, and what those steps looked like and how you had to grow the business to meet the expectations of investors?
Yeah. We've definitely scaled as in necessity because I think there's other SaaS products where I'm amazed at how lean the company is, where with 10 people they can build an incredible product and maintain it. We've just picked a very difficult space that is very resource intensive, very engineering intensive, and the quality bar must be so, so high in terms of execution, voice is one of the hardest spaces, streaming.
If you look at companies like Twitch, YouTube, think about the amount of infrastructure work that's needed to stream audio, stream video. And in our case, it's real time, it's live meetings. If you miss that meeting, you don't get a redo. If your CRM is down for 30 minutes, that's okay, you can log back in the evening, fill out your stuff.
So, that real time nature is very, very intensive. The other thing is people are not just having meetings in a nine to five window in the PST time zone or EST time zone, Fireflies is used across 100 different countries. So we'll have crazy peak volume from 06:00 AM PST until 04:00 PM, 05:00 PM PST, that covers most of the US time zones. And then other countries like India will start peaking, Europe, so there's just tons of places where it's being used so we need to have a 24/7 operation. So people will be working around the clock from different regions. We have folks distributed across, I think, 11 or 12 countries now.
We scaled up since the launch from 10 to now almost 100 or close to that amount. A lot of it is on engineering, a lot of it is on uptime, making sure the system scales, and then also on R&D, how do we build out the other process.
We have a marketing team, we have a customer support team, but the product and engineering orgs take a bulk of that resources and energy today.
And we were stressed out, to say the least, during the pandemic, it's how are we going to keep up with this volume? You keep building and then breaking and then rebuilding and then breaking because of the scale, the stress that's added. And sometimes you have to rip everything apart and then rebuild it. Do it the right way. If you take any shortcuts, you think it's not going to be a concern until six months later.
When you're scaling fast, it's going to be an issue in two months. So solving technical debt, building great architecture, all these things as a naive college student or someone that came out after a year in corporate would not have understood until I've gone through the pain myself.
You mentioned that you're not doing any advertising, but you do have marketing, so what is marketing doing to help the PLG motion?
For us, marketing is focused on driving value back to the community of users, meaning better education material, better tutorials, better guides, being able to help people understand this space. The biggest challenge, I guess, for Fireflies is really about how there's so many different types of personas that come in. There'll be people in ops, people in sales, people in recruiting, people that are really, really young and very tech savvy. There are students using Fireflies in their university lectures. We can't cover every use case, so we have to, one, build a product to be easily usable so that even my grandmother would be able to sign in and start using it ideally. You must take that into consideration.
A lot of the time, the marketing team is focused on delivering value, educational content, talking to our customers, understanding, and gathering that feedback, surveys, putting up really engaging content that people can learn from and understand the market. A lot of the marketing efforts is product marketing rather than performance marketing.
Is there a role for building the brand, or you and your co-founder as thought leaders?
Yeah. Product marketing is a starting point for that where we try to make Fireflies become essentially a well-known name in terms of, okay, this is what it does, it's synonymous for don't worry, I'll have Fireflies to take care of it or Fireflies will send the notes. We look at it as that's something that happens over time.
Calmly has become a very well-known brand, but for the first five to six years, people were talking about other companies, but everyone was using Calmly. That was the thing. So for us, we like to be under the radar, do our thing, make sure our customers are the ones that are talking about Fireflies and sharing it, so that word of mouth is the ultimate... It's hard to quantify. We start trying to quantify all the things like that, but that is our ultimate north star in terms of how can we delight someone enough to start using it.
The great part here is that someone can experience Fireflies before they've even signed up, just like how you can see the note taker, or afterwards if I forward you the recap of the summary from Fireflies, all of those things give someone a really good initial impression. If we can give them that amazing impression, they're more likely to sign up, they're more likely to share. Even when it shows up on a meeting and you're like, "Hey, what is this note taker thing? How does it work?" And from a compliance point of view, people say, "Hey, it's going to record and take notes. I'm happy to share it with you afterwards, or if you don't want, I can remove it." Even that compliance piece helps become a very good marketing opportunity.
It is. Another question I have here, is there an enterprise play and how does that get activated?
Yeah. So we are just starting to think about this. It's great that you brought this up because I was just in a meeting related to this thing because when we look at where and how Fireflies spreads inside an organization and how fast it's spreading, we get to see Fireflies has attended meetings across various domains, and including it's been at least in a meeting, one meeting or more, across 70% of the Fortune 500 companies, and then we can then dive into how many people there are using it, what departments are using it.
There's a lot of interesting insights for us to tap into where if a lot of people inside an org are using it, how can we then potentially help them consolidate into a team workspace, into the business plan? So, that's an area.
We are actively hiring for our very first AEs sales-type of motion because even a company that's fully product-led like ours, there are certain times where for an enterprise or larger organizations, a conversation needs to happen. Maybe the end user loves it, but their manager needs more information on what are our security protocols, how do we have controls in place, how can it help them as from a management point of view? So we are starting to think around that process as well.
Well, I certainly know that's the cohort you mentioned before, they're all looking at, the enterprise play in much the same way, not only on the compliance and management side, but also what are the opportunities. And I think you made a good point earlier on about if I have this searchable archive of my conversations across an organization, there's a lot that can come out of that, so that's really awesome.
I think this was actually a vision of Evernote way back when, and it was a fantastic product back in the mobile-first era, but the idea was Evernote was a great individual user product, but it could have had this amazing opportunity with the enterprise where if everyone's notes were consolidated, not just your brain, but your entire org's brain was consolidated in one central place, and then you have the SharePoint's of the world and other products like that. Notion does that as well--amazing individual product and then they have a great PLG motion.
The nice thing here is with Fireflies you can organize all that knowledge without ever having to type anything, because it's just for meetings, all my knowledge, everything I'm saying in a meeting is getting archived and then everything my teammates are saying in their meetings are archived. And then all of that is available in one central workplace. So imagine having the brain of your organization, but without having to write a Wiki, without having to write a document, without having to take notes during meetings. So I think there's a lot of immediate value that can be driven. For us as the enterprise, that's why we want to capture all the knowledge inside your organization. Not just the sales use case. I think that's the lowest hanging fruit.
But I love the idea of the knowledge base, imagine you're recording the calls on customer service and you transcribe and edit those conversations to become knowledge-based entries. I could also see for copywriting, the people doing the writing in a meeting like this sharing their thoughts on things and taking those thoughts and moving them into more edited pros, so lots of room to grow.
Those are fantastic use cases. I was actually talking to a customer the other day and one of the things they say is they'll go interview someone and then they'll send that Fireflies recap to the copywriter and they can write now a blog post within an hour, before it used to take them almost couple days, because everything is there. They don't have to do a bunch of follow ups. It's right there. They can see the notes. They can search through different topics and then write a blog post.
And then I was like, "We never intended Fireflies to be used by bloggers in that way or your marketing team but that's something that ends up happening."
We have the thing, you mentioned, around customer service, fantastic use case where one of our customer onboarding folks will have a call with a customer.
They'll give a feature request, or they'll say, "Hey, can you improve this?" And instead of having to go convince the PM and engineer, they can literally go into that recap, highlight that 22nd excerpt, create a sound bite and then share the sound bite to the rest of the team on Slack, and then that team can listen, read, and then really empathize with what the customer is saying. And so you don't have to take my word for it, here's what the customer's saying. And for the salesperson or the [inaudible 00:31:33], the customer onboarding person, life is very, very easy, but you can now start collecting all these nuggets like testimonials, like feedback anytime there's a disagreement inside an organization and then on a decision that's being made, they can create that soundbite and then they start a conversation around it. So there's a lot of that sharing that can happen from within a meeting as well.
How do people learn more about Fireflies and sign up?
So fireflies.ai, it's in the name, it's in the note taker that you see, so you can simply go sign up for the free tier, try it out. You'll get free transcription credits. And then the pro and business tier have more integrations, more storage, all the other offerings. You can find us on LinkedIn. We're pretty active now as well around the community. So many, many different places. The best place to go is fireflies.ai.
Great. And if other SaaS founders want to pick your brain, how can they best reach you?
LinkedIn is probably the best place, message me, connect with me and happy to chat. I love everything and anything to do with SaaS.
Well, thanks so much for joining us today on the podcast, Krish, that was excellent episode. I don't think we're going to have to edit a word. Really appreciate it. And if you're listening today and not subscribed, please subscribe to the podcast wherever you get your podcasts. And if you want to reach out to me, I'm on LinkedIn. Thanks for listening and we'll see you soon on the podcast.
Thanks for listening to the SaaS Backwards podcast brought to you by Austin Lawrence Group. We are a growth marketing agency that helps SaaS firms reduce churn, accelerate sales, and generate demand. Learn more about us at www.austinlawrence.com. You can email Ken Lempit at firstname.lastname@example.org about any SaaS marketing or customer retention subject. We hope you'll subscribe, and thanks again for listening.