Matt Mireles – SpeakerText 1 of 2

"Nontechnical founders should be able to hack social systems." SpeakerText provides online transcription, making video more discoverable and engaging.

Matthew Wise: Hi, this is Matthew Wise, founder of We empower entrepreneurs to have a voice and share their story with the world, enabling others to learn about building products and starting companies. I am here today with Matthew Mireles, founder and CEO of Speaker Text.

Speaker Text is an on-demand video transcription company that enables video publishers of all sizes to boost their SEO, improve the viewer experience by offering high quality transcriptions of their video, which enhances the overall viewer experience and enables closed caption text to a wider audience. With that said, I’d like to thank you, Matt, for being a guest on the show. We’re hoping that you can give us a brief bio of yourself so that our audience can know who you are.

Matt Mireles: Yeah, totally. Thanks for having me. My name is Matt Mireles. I’m the founder and CEO of We are, as you said, an on-demand transcription company. Think of it as transcription in the cloud, powered by a very novel hybrid system that combines Cloud source labor and intelligent machines.

Matthew Wise: Why are you so passionate about it?

Matt Mireles: One of the things we care about is helping publishers get paid. I was a journalist and that’s a big process for me to move, how do they make money on the Internet? One of the beautiful things about Speaker Text is that it’s actually good for both consumers and for publishers. That’s actually pretty unusual in technology on the web. It’s good for consumers in that having text synchronized with video allows you to have a much more efficient and engaging experience with the video.

There’s the obvious use case of just reading the transcript. Some people have time to listen to a half hour long interview. Other people, they just want to read the transcript and find the interesting parts. There’s a big problem with accessibility with online video for the deaf. Close captioning in transcripts helps on that, but also what we’ve seen increasingly is people building applications that leverage transcripts. Allow people to share moments inside the video using quotes. You can tweet out a quote or link back to that moment inside the video. So, there’s a social functionality there.

That all dovetails with the publisher benefits. For them, and guys like you who spend a bunch of time and money and effort to create a video, but how does it get discovered. It’s all about sort of going viral. In a way, that’s sort of like throwing darts. How do you get people to share video? An organic search isn’t that much of a driver because Google and the rest of these technologies on the Internet were built on the presumption of there being some sort of text that they can crawl and understand to organize the web.

Well, video is not a text document, so by layering a transcript with tags on top of video, suddenly it’s discoverable in search engines. Then, you get that long-term search traffic. On top of that, you can do more targeted contextual video advertising using a transcript combined with time codes. So, publishers get more eyeballs, and they turn those eyeballs into more money.

Matthew Wise: Now we’d like to kind of understand a little bit more about some metrics or traction to date. So, where are you guys at Speaker Text?

Matt Mireles: We’re in the process of collecting data to see what exactly Speaker Text has on engagement and viewership. Some of the earlier data that we’ve got to date is pretty amazing. One client that we can’t really name, reported a 160% increase in time spent watching videos. Once they added Speaker Text, their videos went from being watched on average of two to three minutes to five to six. In terms of the traction side of things, we can’t announce the name yet, but we actually just signed a really big client, one of the major video platforms. They’re going to be using Speaker Text to power up transcriptions for a global customer base. So that’s really, really exciting.

Matthew Wise: When did you guys launch?

Matt Mireles: There’s been sort of a rise and fall and sort of rebirth of Speaker Text. We sort of launched a product in January 5th, 2010 at the New York Tech Meetup. We started off in New York City, and the idea was very similar to what we’re doing now. It was an interactive transcript [inaudible 05:09] that lets you share quotes from inside a video. Over time and as the team changed, we sort of got into the market more.

What we realized is that our real value proposition, the core of what people want and value from us, and what we can do especially well is actually the transcription component. The company almost died in the meantime.

Matthew Wise: Wow.

Matt Mireles: There were big team changes. At one point, we had $1300 in the bank. We moved to Pittsburgh, were broke. We had tried to raise money. Totally, everything fell apart. We applied to Y Combinator; got rejected from Y Combinator. People said we like you, but we don’t have any faith in you and your team to actually build anything.

So, we moved to Pittsburgh to my co-founder and CTO, Matt Swanson’s apartment. He had just finished graduate school at Carnegie Mellon. I lived in the dining room. Tyler, our third co-founder lived in his old roommate’s room. For two months, we just locked ourselves in the apartment essentially. Literally, there were like days where we wouldn’t leave for days on end except to go like across the street to the gas station and buy some sort of snack or something. It was insane, coding from 11:00 a.m. to 5:00 in the morning [inaudible 06:43].

We were trying to raise investment, and our credit card was maxed out. These motorcycles would drive by the window, gangs of motorcycles, so we would be on the phones with investors, but we didn’t actually have a phone, we had Skype on the laptop so we’d be huddled around the laptop and it’s got these crappy speakers, you can barely hear in general and then you have semi trucks and motorcycles driving by. So, we said, “OK. We need an actual telephone with a speaker phone.” So, we go to Radio Shack and pick out the cheapest phone with a speaker phone they have, and our credit card gets declined, and we can’t even get a phone. We came so close to running out of money.

I’d been a paramedic before in New York City, and that’s how I financed my college education, and the early days of Speaker Text, and the plan was to say, OK. We could make rent through June, but then after that we were out of money. And so, I was going to go back to New York, sleep on friends’ couches and work as a paramedic and essentially send remittances to the guys back in Pittsburgh. That was the plan, and at the end of June we raised a very tiny, small amount of money from a couple of angel investors.

And then, suddenly we were able to hire a graphic designer and get a logo on 99 Designs and hire some college kids from Carnegie Mellon to do some free lance work. Ultimately, we moved out to California and launched the product. After then after that, now we’ve raised just a little over $600,000 in an angel round lead by Mitch Kapor, with Dave McClure, the CEO of Cloud Flower participating. Now, it’s a totally different story, but it came very, very close.

  • anonymous

    …so where is the online video transcription? ;)

    • FounderLY

      we produce videos and get them transcribed on a rolling basis.  we hope to have it up soon.


FounderLY Photos

			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY

Press Mentions #1

Press Mentions #2