Welcome, Descript

Nabeel Hyatt
Spark Capital Publication
11 min readJan 12, 2021

Nabeel: Hey everybody, this is Nabeel and Bijan from Spark. Normally this would be a fundraising blog post, which basically I started to write a couple of days ago to announce our investment in Descript, but it felt not native to the medium. Here we are announcing a new investment in a company that we believe in strongly and then not using that tool felt wrong. So I messaged a Bijan. We gathered together on a Sunday, and we’re here to chat Descript.

What makes Descript magical

Bijan: Do you want to just take a minute and describe what it is? Cause I don’t know if folks know what it actually does.

Nabeel: Yeah. Descript is an entirely new way to edit audio & video. Audio and video editing tools today from the high-end of After Effects that are being used to edit a feature length film all the way down to an internal podcast being recorded for a mid-level marketing manager, they are still basically the same set of tools and user interfaces. And it has been for 25 plus years up until Descript.

Audio and video editing today is a lesson in abstraction. Traditionally, you hit the record button and what you get back is a set of wave forms or key frames in an editing suite like Adobe Audition, or iMovie that are not really directly the thing you’re trying to edit. What you’re really trying to edit are the words that a person said.

That’s a design concept called direct manipulation. The feeling that you’re working directly on the task versus trying to learn another language of engineering and design in order to get what you want out of your creative tool.

What Descript does fundamentally, is it takes that audio or video that you recorded and gives you back a transcript. It gives you back the words that you actually said.

And then when you delete or edit a word inside of that script, it will remove that piece of audio or video from the thing that you’re editing.

Then more recently, they launched the ability using artificial intelligence, akin to deep fake technology to actually be able to insert your words into that script as well. So you can type in a word the way you might type into Google docs and then have it actually show up in your voice.

All of this is really to the central vision of creating an audio video editing suite that decreases the amount of time that a professional editor would use by 10 X, but also increasing by 10 X, the people that can do this kind of editing.

Bijan: Yeah, I don’t know if you would agree with this and if you don’t, you can always nuke it in Descript. The first time, Apple put FireWire and iMovie on their Macs, it was a game changer. It was desktop software and it just gave this level of creativity and power to the user.

And then I feel like all of that just got stuck. Like the way you edit, video and audio has been living in that paradigm all of this time. And Descript, just has magically created something brand new.

Nabeel: Dylan Field at Figma likes to say that at IBM the engineer to designer ratio went from one in 500 twenty years ago to one in 10 designer ratio today. Everybody is a designer and a creator now, everybody’s streaming on Twitch or trying to create an internal video for their team to teach them some new intranet tool. We are all in the mass communications, design and creative business, and I’m not sure our tools have really come along with that.

I was just going through this with my son last night. My son is doing what lots of kids that are 12 years old do, starting a YouTube channel.

He’s streaming his Minecraft. The kids got an hour and 30 minutes of content and the typical pathway would be that he would then spend an hour and 30 minutes trying to edit that content and get it down to something like 20 minutes. And at its root I think of Descript as taking that hour and a half of time and turning it into 10 or 15 minutes and that 10 or 15 minutes being spent more in flow in creative flow of how do I want this to feel and look, and edit versus, wrestling with the menu system.

Ultimately there’s an entirely new UI or UX paradigm that needs to be invented.

Fundamentally the UI of looking at the script of the things that you’ve said, and then building an whole new set of AI and machine learning tools, starting from that new starting point. That’s what feels so fundamentally differentiated.

How we found Descript

Bijan: Yeah, absolutely. We’ve been investing in creative tools for a long time and Descript really got us all super excited, it’d be great, Nabeel, you led this investment for us, kinda of how did you meet Andrew and how did this all happen?

Nabeel: I reached out cold to Andrew. As you know because it’s just a phenomenal product.

I’ve been working in non-linear video and audio editing back to art school and beforehand back when you had to fill out a time sheet in college to get an hour to try and edit your video.

Bijan: Now you are really dating yourself.

Nabeel: Sorry to date myself so badly. The first wave of software was moving all of that digital to the desktop, the Avids of the world going away, and the Adobes being created.

Then the last big step function change was moving things into the cloud. In the cloud we get browser based, and newly collaborative, but they’re still fundamentally the same tools. I don’t know how old is Google docs now, maybe 15 years old or something like that. Google Sheets is still basically as hard to use as Excel. It’s more accessible because it’s in the cloud and it’s more collaborative because everybody can, real-time edit, but it’s still fundamentally just as difficult to author in.

And editing audio is still as hard as it’s ever been. I’m still staring at a waveform. And Descript felt like seminal moment along that arc.

Bijan: So it, wasn’t our world famous podcast, Hallway Chat that led you to Descript.

Nabeel: It certainly helps to be using these tools to really feel when there’s a real step function change. [laughter] Sure, it was really my personal frustration with editing our podcast, which led me down this path.

How does AI impact Creative Tools

Bijan: Hey, so one thing that happens with VCs when they announce a new investment, obviously they’re super excited and that’s true for us as well. But I thought if this is, a big infomercial for Descript it may not be as revealing. One thing we often talk about and debate about is, how does AI impact creative tools and is there a slippery slope there and what are the trade offs?

We’ve obviously invested in things like Grammarly where the machine starts getting smarter and assisting the user, but in, in the world of creative tools, you can imagine people either being super excited or concerned about, the machine taking over.

Nabeel: Look, one way to think about it is, is the software trying to be a force multiplier to the number of people that can be truly creative in the world, or is it trying to be creative for them.

We like AI when it’s taking away a hundred hours of tedium and we don’t like AI when it’s taken away a hundred jobs. 1997 was when Big Blue first beat the greatest grand master of chess in the world. And and so we’ve been dealing with AI for a while.

But despite that the best chess player in the world is not an AI. It is an AI and a human working together.

I’m firmly in the world where I don’t want artificial intelligence as the “automagically” button. Michael Cook is working on this thing called games by Angelina, which is an AI that creates games. That’s really not that interesting to me. I want an AI that helps everybody make the game that’s in their brain.

Bijan: I do think the magic of Descript is that it allows the creator to do things they would be doing. And if would just take them 10, 20 times longer, or maybe they wouldn’t do it at all, because it’s so hard.

Nabeel: As we intro introduce artificial intelligence, computational photography, deep fakes, its a new era, and we do need to have a separate conversation about where we draw the line about whether those things are true or not.

But I think the first thing we have to ask is whether it’s in the hands of the creator to even make that decision. And if the NPR audio producer can remove all the Uhs and Ums, because they’re willing to spend, 10 extra hours doing that work that creates a separation between what that NPR producer can do and what an average Joe can do.

And what I love about a product like Descript, is it just collapses that. It’s so easy.

Bijan: Yeah. And, I think the thing you said earlier is exactly right. Google Docs by moving to the cloud and making collaborative was a big change, but this is a whole new way of creating and collaborating. When you’re working in kind of single-player mode on Google Docs you can imagine other desktop software having a similar kind of thing that the change happens when you start collaborating. But Descript, when you’re in kind of single-player mode right away, it’s different. You can’t do those things.

Nabeel: We’re in a world where millions of people every single week are streaming things to Twitch or recording something inside of a corporation and are now having to wrestle with a new world of editing tools. There’s 10 X, more people trying to do that work.

We’ll see over the next five years, but with a tool like this, how many more people suddenly feel like it is accessible. I think my son would have taken his Minecraft YouTube video previously and maybe spent 15 minutes in iMovie, cut a little off at the beginning and the end. But he probably just uploads an hour of relatively boring footage because he’s not going to spend half a Saturday video editing.

Bijan: Right, right, right.

Nabeel: But with this, he was really excited yesterday. He immediately got it in script form, and we literally cut out an hour of UHS and ums in 10 seconds. And, and then he cut the whole thing down to 20 minutes, which is ideally that much more interesting. That means that many more people watch it.

And that’s the creative feedback loop. He creates better content. He gets better feedback. Ideally he keeps doing it, because it’s that much more interesting and fun to do the work. Now, I don’t know that Descript will capture all the net effect of that new creativity out in the world. But a really good platform never captures anywhere near the amount of value it creates in the world.

Bijan: Yeah.

The Power of the Pivot

I wonder if there’s another thing we didn’t mention that we should bring up?

I don’t know if it’s too much inside baseball, but like his last company was a pivot. Groupon was a pivot and this company, was it technically a pivot?

Nabeel: Oh yeah it was a pivot.

Bijan: I find this the founder story interesting, this isn’t something that just fell out of the sky and it was like, boom, overnight success.This is tough.

Nabeel: I think it’s worth touching on.

Bijan: Talk about that because I don’t know if people everybody who listens to this will know the backstory.

Nabeel: The company was started by Andrew Mason as he was working on an audio podcasting travel company, this platform where you could walk around the city of San Francisco and have somebody in your ear, who’s telling you a story about that neighborhood.

But what they found is that the amount of time spent editing that audio was really cost prohibitive. Artificial intelligence at the same time had just finally gotten to the point where you could reliably and cheaply create a non-human intervened transcript.

And probably there are dozens of audio engineers that have been saying for 25 years, “Why am I staring at a waveform, and why can’t I just look at the script?” But the tools weren’t there, the technology wasn’t there until now. So they built it.

Bijan: Well, what I love about these stories, where the founder pivots, whether it’s what Andrew did it at Detour to Descript or what he did it at The Point going to Groupon or what Stewart did at Tiny Co to Slack.

There’s a myth that people think that these founders just there’s this moment of an epiphany. And then they launch and they get a whole bunch of venture capital, and then everything works out. Or this fail fast notion where it doesn’t work out. And, I feel like the entrepreneurs that have this desire to build, and they’re passionate, but at the same time they’re open, it’s a very interesting balance. I think it’s rarefied air, frankly, of people that can have an idea, but are still open-minded about where they want to build and create and innovate and, they can Zig and zag and it’s super hard, super challenging.

But when it works, it’s amazing.

Nabeel: I like the phrase you used, passionate, but open. That strong opinions loosely held view of the world. There are very few people that can hold on to that. And it becomes way, way harder once you have a set of employees or you’ve taken investor money, you spent all the time to go talk to Techcrunch or been on a stage.

All of those CEOs before they pivoted, it’s not just that they were in their dorm room. They had recruited employees on that vision. They had gone and raised money on that vision. This was the next big thing, right. They told themselves their wives, their husbands.

Jason, before he pivoted into Discord, had raised venture capital, he’d already spoken to people at the Game Developers Conference about why this was going to be an important game. It’s not a napkin sketch that then pivoted, there really is having to admit some level of failure personally and get excited about a new thing.

And then more importantly gather the rest of the troops. Hey everybody, I know I told you about the other thing.

Bijan: You may convince yourself with the new thing, but it doesn’t always happen with the rest of the team.

I mean when Odeo pivoted to Twitter, not everybody came along for the ride either from an investor standpoint or an employee standpoint. It’s a really tricky one and I just have so much respect for our founders that have this level of conviction to make the pivot.

It’s super impressive.

Nabeel: Back to Descript, almost all of those pivots that we talked about, started out building something that was fundamentally creative and it ended up that the tool that you needed to build the creative thing, led to the new broader thing. That was Slack and Discord as a communications tool to build this creative product in a game. And with Descript, it’s the same.

It is, “I was trying to make audio and, making audio is just too hard today. Somebody, should fix that. Ah, it looks like that’s going to be that’s somebody who’s going to be me.”

Bijan: Yeah. It’s amazing. Really amazing.

Nabeel: Okay, well, thanks for hopping on today and chatting with me about it.

Bijan: It also feels like we can talk and record in a way that feels more relaxed knowing that we can edit away. I find it’s a different level of comfort when you’re recording a podcast knowing that you’ve got Descript in your back pocket. I really, I believe that I, feel the difference.

Nabeel: Yeah. You feel the difference that you feel like you can, run a little bit more naturally and authentic cause you know, if you mess up, you’re going to edit it later. And in this case, you are going to edit it later because I’m going to send you the link to Descript and you are going to be in charge of editing this for us.

I’ll contribute.

Bijan: Thanks.

Nabeel: Thanks. Have a good rest of your Sunday. All right, bye.

--

--

Nabeel Hyatt
Spark Capital Publication

partner at Spark Capital, former CEO, full-time geek, investing time and $ at the fringes. http://nabeelhyatt.com