Synthesia raises Series E: 90% of Fortune 100 deployed, SaaS margins, and Nvidia as investor

Phantom Cash, fund your wallet without exchanges or middlemen and spend with the Phantom Card.

And up next we have Victor. Victor

from Synthesia.

Can bring him in.

Keep him waiting.

Victor, how are you doing?

Good to have you on the show.

Thanks so much for hopping on.

It's great to see you.

Give us the news. How are you doing?

I'm a very proud technology brother today. We just announced our series Z $2 million.

Let's go.

Incredible. Incredible. Billy,

incredible stuff.

Uh, what unlocked this round? What's growth been like? What's adoption been like? Is there a clear turning point? Is it more of the same steady compounding? What's the business look like today?

It's all free, but it's it's very much, you know, steady and and compounding. We've been growing really fast. I think the thing that stands out to, you know, both existing and new investors, I think, is very much the quality of revenue. I think it's very clear with the numbers that we're not a shop where we just have a bunch of people coming with their credit cards like playing around doing some experiments.

This is like real deployments multi-million dollar tracks 90% of 4 to 100 very kind of multi and uh yeah I think that's for us it's the main story.

Yeah

the the sort of second thing around this race right is just the the next big opportunity. The last 5 years have been about making video faster and better with AI. Yeah.

And what we're exploring now and launching some new products around is the intersection of agentic experiences and video. So kind of video you can talk to rather than sort of passive to just

uh watch. Imagine a corporate training video for example. You're learning about how to beat your competitors kind of a video battle card. But instead of just consuming that after you've consumed the video you go into a role play with an agent pretending to be a customer. You have to overcome objections. prove you've actually understood the content. Um, which is just going to be a completely different video experience. And I think, you know, for us that's the hundred million dollar opportunity.

Yeah. What's uh what percentage of the business is internal customer video generation like what you described? Uh that certainly seems like a sweet spot, but is it 100% of the business or are there companies that are starting to vend this externally?

No, we have a lot of external use cases. Uh I think we estimate that around 55% or something like that is is internal and the rest is external. I think what most people get wrong though is that most people think of external content as being marketing, advertising, storytelling, creative things. I think you know that's not what people use synthesis for.

It's much more like PowerPoint style use cases. You know like PowerPoint is used a lot internally but it's also used a lot externally with customers to explain complex parts around the product etc. But you don't put like a PowerPoint as the first thing you see in the website, right? It's not the storytelling part of of the company. So enablement, support, uh product marketing in the sense like product marketing videos that explain the very like fine details of a product, why the product is better than competitors, not like performance marketing and that kind of stuff.

Yeah. What's the sweet spot for one of those use cases? I can imagine at at the very low-end, you know, consu direct to consumer companies, they probably you just want to see like a fact sheet and then you just click buy. Uh at the very high end, like want to have a dinner with the salesperson to really understand what I'm getting myself into. Is it is it bound by some sort of middle ground? Where are people seeing success with that ex with that use case? I I think if you really boil it down,

it's these is used to explain things to people, right? And so the sweet spot is generally like complex products.

If you sell insurance, for example, most people have no clue how insurance product works.

Having a video to explain to people how insurance works rather than a 10-page document is very very effective. Same thing with pharmaceutical, you know, software like complex products that require a lot of explanation both internally and externally. That's where Synthesia really shines. And so I think one way of thinking about is like we're less focused on making content that ends up in a social newsfeed competing for eyeballs and attention or helping people create content for like all of you know big software company has like 100,000 landing pages. Yeah.

Probably 10% of those would perform much better if they had a video explaining whatever that landing page explains. No matter if you like it or not, like people don't want to read. Like the data is so clear that we see with our customer testing. It's absolutely wild how

people will read, but they need subway surfers. They need the video, you know, the car going down the track next to it. I had I had this moment. There's uh John and I are looking at joining this like uh this like members club and they're sending all these different decks that all have like different information and like you kind of want the information from all of them, but you're kind of like switching around PDFs.

And I'm just thinking the whole time like this one should be like video because it's like highly visual. It's like they're still in development or whatever. And then you want to be able to like just chat with the information in real time like you were describing earlier instead of having to email somebody and wait to get kind of information and come back to it. Um and so yeah, I just think it's very obvious this is where uh media is going. Uh what uh how's it been just kind of like managing through the growth? You know, it was reported that you guys turned a multi-billion dollar offer from Adobe down at the end of last year. Uh I imagine some employees are like you know what what are you doing? I think there was some some employee liquidity as part of this round as well but um how's it been kind of scaling the team and and managing through this process? I mean, it's hard, right? Like, I think there's there's no there's not there's nothing wrong that when when a company is scaling this quickly, you're doubling your headcount every year, doubling your revenue. Like, that is just like that creates a lot of um of hard problems to solve, right?

Um I think I think one of the big challenges is becoming a multi-product company. Uh I think like kind of scaling an existing product and and uh existing go to market motion is is really really hard but it's it's somewhat easier in some way right when you know eight nine years ago we found the company and we spent the first like four years getting to product market fit on the product that we're now scaling today that process is so messy right there's no like recipe for how you you can't like just put a recipe down like how you scale that and going through that again at a much different scale is just um I I think that's that's really hard something one of the things I spend a lot of of my time on in terms of the round and sort of I think everyone's enthusiast um really sees that this is a huge huge huge opportunity because it's evident in the numbers you know and as as I said before not just the topline growth but the quality of the revenue so everyone is is pumped for the next chapters and I think you know this new product that we're launching very soon is is currently in beta with with a bunch of our customers and it's it's very very promising so I mean lots of headaches right I think there is in every high growth company but um I think you know as long as you're growing quickly a lot of those challenges um become good problems to have right

yeah you know

100%

that makes a ton of sense um yeah I mean in terms of structuring the company when you build a new product are you like how much are you copy pasting the organization creating silos versus trying to embed the multi-product strategy across every aspect of the organization I think you have basically two the sort of two extreme choices, right? One is that you you build a completely like separate company.

Yeah.

And the other one is that you deeply integrate your new product into the product you've already built. And we've made a very conscious decision

to do the second one which causes a lot more headaches, right? There's a lot more dependencies. Yeah,

but I I truly believe that for our products, we want to compound then the fact that we have a visual editor that so many people, you know, use and love. If we can layer the agendic experience on top of that rather than having a completely separate product that people have to learn, that's going to be really powerful. And I, you know, I keep thinking of PowerPoint. PowerPoint is is such a um ingrained way in every company, right, for for communicating externally and internally. And I think that the future of these kind of media formats is going to be this sort of hybrid between it's almost like part video, part video game, part like LLM sort of chatbot. And what I want to build is the interface for everyday PowerPoint style users to be able to create these what we call video agents that can not just communicate but do some live rate process automation. It can screen a candidate for you. It can do a roleplay with your sales team to ensure they've understood the content. And for that to really work, we need to build everything into the editor, which is the core and the heart of this indicesia product, right? And that's hard. Like I think for for you just want to launch something, it's probably much faster to like go the other way and just say, "Okay, you know, separate team, separate infra, just go and build something." But um I think in in in the case of like what we're doing, I think especially for creative tools, I think you see Figma using kind of a similar strategy of a lot of the tech is shared even though they they express it in different products. I think for creative tooling that that's how you kind of like build real competitive mode and and differentiation over time.

Yeah. Speaking of Figma, uh I mean you have 60,000 customers now. How do you think about engaging with all of them? Are you going to do a conference? Uh are you already doing conferences for user groups in different cities? How are you thinking about uh stay keeping all of your customers engaged?

So we do a lot of smaller conferences um locally in uh you know North America and in Europe where we operate. We had our first kind of big customer conference last year.

I think one of the things um you I would say we've done really well is to be very focused on who we're building a product for and who we're selling it to and that is the enterprise. It is some of the use cases we discussed before. And so I think we we put less sort of emphasis in some ways on cool you know Twitter friendly demos and PR stunts. and put a lot more focus on actually being there with our customers, talking to them, holding their hands, onboarding them. And of course, you know, we're building, we're building AI, we're training our own models, we're doing so many cool things,

but I think what what has served us well in the past and I think this funding round is is sort of the the crescendo of is really just focusing on real utility, not just pilots and demos and and cool things. We do a lot of these kind of things already. Um, but we are we're kind of very specific. I'd rather have a small conference with a lot of, you know, very qualified customers or prospects than than do this like, you know, very broad kind of a 10,000 people thing.

Yeah, that makes sense. Uh, last question for me. Uh, Nvidia's in this round. Um,

nice.

Uh, amazing. Uh, I I but I'm curious about inference costs. You mentioned you're training your own models. Uh, there's some, you know, discussion of a bottleneck at TSMC at some point. There's been discussions of how quickly buildouts can happen. maybe H100 prices are moving. Uh how are you thinking about inference cost? Is that something that's keeping you up at night at all or are you able to optimize the models to a point where uh it hasn't been a headache for you?

Th those are definitely important inputs to to the business, but they're not even top five of okay,

you know, things I think about on daily basis. Today we have SAS margins. We have a great business

that has awesome margins, great unit economics.

Um, thank you.

It's amazing.

Um, and I think the reason for that is that we're both because we train our own models, which means we can bring that cost down.

Sure.

But also, I think one of the misconceptions about Cynthia is that we're kind of an AI model company. We're very much a workflow and application layer company, right? Right. Um the the AI model is a very important input to our platform,

but the but the product we're selling kind of goes goes far beyond just generating AI uh clips.

Yeah. Yeah. Yeah. It's about the management, the distribution, everything that happens within the organization to make sure it's used effectively, not just the generation. And

and I think if you look at a lot of the other video companies more tuned towards consumers, etc. where the product literally is, you know, a prompt box where you type something in, you get an eight 10 second clips out from whatever like the latest hot model is.

Yep. I mean you you see crazy growth in a lot of those companies and I think you know a lot of those products will probably be durable but that's really like a race to the bottom in some sense right like we can offer like the most um credits for the lowest price as a growth tactic and I think it'll be interesting to see how all this stuff kind of pans out in in a couple of years right but you have to build you have to build more differentiation than just being able to raise quickly enough to um to serve you know I've I've heard some of these companies are operating at like minus 10% margins right like That's That's expensive growth.

Yeah, for sure. For sure. Well,

I know some of them have been on the show.

You know, it's good. It's good to figure it out.

Happy that uh happy that you have, you know, annual contracts and SAS margins. Good business to be.

Warms my heart. Warms my heart.

Well, congratulations on all the progress.

Great to get the update.

Thanks so much for hopping on.

Fantastic work.

We'll talk to you soon. Cheers.

Have a good rest of your week. Uh, we have just five minutes.

Okay.

Uh, but we can jump in here. Kenneth Castle says, "Hate to report it, but having a TV dashboard clearly visible to everyone with a number that needs to go up makes it more likely that the number goes up." There was an entire company that just did this. Do you remember it was called Geck? Was it called the Gecko board?

Gecko board or Yeah, I know. No, the number go up.

Yep. The number go up company.

Yeah, it's real time. Real time data. Yeah.

Uh, KPI dashboards to put live metrics front and center, helping teams react faster. Not affiliated, but uh, very, uh, very cool. They should just rebrand to number go up company. The number go up company of San Francisco.

Yeah. I, uh, yeah, I'm trying to think. The the the the problem with that is that there are numbers that you might want to go make that you might want to make go up that you don't necessarily want just out there in your company if you're having maybe a customer onsite or a investor in or something. There's like a whole bunch of things and then if that number doesn't go up and everyone gets depressed it's like that old story of Enron with the stock price in the in the elevator. You know, people are really happy and energized on the day that the stock price goes up. The stock price goes down in the chat and put the stock

stock price in the elevator. Yeah. You know, I talk about this all the time. Um but but no, seriously, like I I've done a tour like an office tour once and the team had like their revenue, their gross margins, like truly everything just public as soon as you walked in. And uh you know, some of those numbers are important. If a customer is like, "Wait, can you give me a better price because I saw that you drove your gross margins to 99%?" Certainly, you can give me 50% off and that could maybe not be good for your business. So, you want to be careful about where those TVs are placed, but certainly in the right in the right format. Uh, I think it could make a lot of sense. Marco Jusk says, "Hilarious how Soft Bank raised more money than God to invest in AI in the 2010s, then put it all into Uber, Wei Work, Door Dash, CLA, FTX, but not OpenAI or Tesla. They invested 4 billion in Nvidia, then sold in 2019 at a loss.

Would be worth over 220 billion today.

But he didn't get into the game. Got into Open AI."

Yeah.

30 at 330 something. Wait, I thought it was even lower. I thought he was in like the 76 round or the 100 round. Maybe he was, but the big

the big one was the still maybe five acts. Who knows where where it gets out, but you know, uh it could it could happen. Could happen. Anyway, thank you so much for tuning in. Happy Monday. We will see you tomorrow at 11:00 a.m. Pacific. Leave us five stars on Apple Podcasts and Spotify. Subscribe TVPN newsletter at TBNN.

The show felt very short.

It did feel short. Uh, I think it I I think when we have a guest join at 11:45, even if it's just for 15 minutes and then we go back into timeline, it sort of feels like the show's accelerating and then you're pulling off. You're you're switching from the gas to the brakes and then the the gas and you're going all over the place.

We hope you have a wonderful evening.

Yeah. And we'll see you tomorrow.

See you tomorrow.

Goodbye.

Thank you for being here.

Nice work, brothers. I'll see you on the next one.