August 18, 2025 • 48 mins
In this episode of Investing in Accessibility, hosts Kelvin Crosby and Chris Maher dive into the evolution of accessibility tech with Josh Miller, Co-Founder and Co-CEO of 3Play Media—a leading media accessibility company that started at MIT and grew into an industry trailblazer.
Together, they unpack how 3Play Media emerged from a real-world need to caption complex academic videos into a scalable tech-driven company that combines AI innovation with human precision. Josh shares his entrepreneurial journey—from an early startup that didn’t succeed, to building a company with strong co-founder dynamics, to navigating growth, acquisitions, and private equity investment—all while staying mission-focused.
In this episode you’ll hear:
- The story behind the founding of 3Play Media and how it identified a niche before the market even existed.
- Why accessibility must be built for both the buyer and the end user—and how balancing compliance with value is key.
- How AI has revolutionized speech recognition, captioning, and media accessibility—and where it's going next.
- The importance of customer workflow, scalability, and being AI engine agnostic.
- What it takes to grow a successful company in a mission-driven market, including how to say no to misaligned opportunities.
Josh also offers powerful insights for early-stage founders on execution, differentiation, and staying true to the mission in the face of rapid tech evolution and market shifts.
Whether you’re an entrepreneur, investor or accessibility advocate, this episode is packed with lessons on how to scale with impact.
Links & Resources:
Josh Miller: LinkedIn
3Play Media: Website
2025 State of ASR: Report
COMING SOON!
American Sign Language (ASL) and Captioning for each episode will be provided on our YouTube channel. Go to handle @SamaritanPartners.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Kelvin Crosby (00:06):
Welcome to Investing in Accessibility, a
Samaritan Partners podcast.
We're not waiting for change,we're investing in it.
Join us as we speak withentrepreneurs and thought
leaders that are focused oncreating a more accessible world
(00:28):
.
Hey, It's so good to see you,even though I can't see you.
It's another beautiful day inthe neighborhood and I'm so
excited that you're here atInvesting in Accessibility.
And we're going to be talkingabout the captioning effect.
I think we're going to upgradeit from the curb cut effect to
the captioning effect today, butbefore we get there, let me
introduce you to my co-host,Chris Maher.
(00:50):
How you doing, man?
Chris Maher (00:52):
Good, Kelvin, it's good to see you and to be with
you, my friend.
We've both been a little bitbusy.
Kelvin Crosby (00:57):
Oh yeah.
Chris Maher (00:58):
We've been burning the candle at both ends, and
that might be a separate episodewhere we dig into that.
But how are you, buddy, you'vegot so much on your plate, right
now?
Kelvin Crosby (01:05):
Oh, I'm trying to survive between parenting,
between running two differentbusinesses, launching a new
series on the DeafBlind Potterplatform and then working the
See Me Cane and just trying toget the last version of the
prototype done.
I'm just like I just need tobreathe .
Chris Maher (01:28):
Your ears must have been earlier this week because
I was at that conference,Disability:IN and I ran into our
friend, Troy from Aira, whointroduced us so, so grateful
for that, and Mike May and wewere, we were talking about you
and and it was all glowing andso your ears must have been
burning a couple days ago whenwe were talking about you at the
(01:49):
conference.
Kelvin Crosby (01:50):
Well, that's awesome.
Well, I tell you, we had anepisode a couple months ago that
we talked about upgrading fromthe curb cut effect to
captioning effect, and I'mexcited about today's
conversation.
So why don't you introduce usto our guest?
Chris Maher (02:08):
You got it.
So our guest today is JoshMiller, who's the co-founder and
CEO of 3Play Media, and beforeI kind of talk about how we
connected, let's welcome him tothe show.
So, Josh, welcome to thepodcast.
Josh Miller (02:19):
Thank you.
Appreciate you guys having meon.
I'm excited to have theconversation.
Chris Maher (02:23):
Well it's good to have you here, and so we, as we
were talking before we startedrecording, we met at M- Enabling
and I think it was the fall ofOctober 2023.
And I think I was walking bythe 3Play Media booth and I
caught you there and you weregracious, and I literally just
launched the Samaritan Fund.
And I think you spent like 15minutes chatting with me and
(02:47):
telling me like, oh, meet thisperson and check out this person
.
So you were incredibly graciousand kind to me at that time and
then we stayed in touch and webumped into each other at the
conferences.
And then again your kindness,recently you reached out because
you recommended me to moderatea panel for a conference in
Indianapolis in the fall, whichI'm going to do by the way, and
so thank you for that.
But yeah, no, you've just been,you've been kind and generous
(03:09):
and have become a friend overtime, and so it's great to have
you on the podcast and to diginto your business.
Josh Miller (03:16):
Oh, that's great and, I think, for everyone.
I mean, I'm a huge believer inentrepreneurship and innovation
and the more we can do,especially in the accessibility
space, to promote that growthmindset and innovation here is
such a good thing for all of us,and I know that we've certainly
benefited from just generalawareness growing in the market
(03:38):
and if we can do more of that tohelp each other, we should.
Chris Maher (03:41):
Yeah, love it.
And so today we're going to getinto a number of topics, but at
a high level,big reason why Kelvin and I
wanted to have Josh on thepodcast and to talk about his
business is, as this podcast isthe intersection of
entrepreneurship, accessibilityand impact investing, part of
(04:02):
what we want to do is certainlytalk about early stage
innovation right, that's justkind of getting off the ground,
but we also need to highlightthe success stories, and 3Play
Media, and you and your team,are absolutely one of those
examples.
And so we need to start shininga light on that, because we
don't have a ton of them so far.
Hopefully we're going to have alot more, and so we need to be
(04:23):
talking more about the companiesthat have scaled to that next
level, or the next severallevels, as examples for private
capital that hasn't yet startedto look at this space as an
investment opportunity.
So with that being the tablesetting let's first start off,
Josh, because it's a little bitunique, I think, in terms of how
the business started.
Talk about, maybe, yourbackground first prior to
(04:48):
starting the business and, ifyou want to get into maybe your
connection or lived experiencearound disability and then how
the founding of 3Play Mediahappened, would be great, let's
start there?
Josh Miller (04:59):
Yeah, definitely, and I appreciate very, very kind
words.
I think we would say we've madea lot of progress but we still
have a lot to solve, butdefinitely excited to talk about
it.
So we started the company atMIT Sloan.
There were four of us whoco-founded the business.
Two of us continue to run ittoday and the other two are
still good friends and areactively in conversations about
(05:20):
how things are going.
I will say right off the bathaving four co-founders is a
real blessing, because you'vegot four people who are just
working their butts off, we'llsay, and just doing anything to
see the business forward andmake progress together.
Certainly, who you have as yourco-founders matters and making
sure that that alignment exists,but we were really lucky to
(05:43):
have a tight group of people whoare just so focused on getting
to the next step and pushingthings forward together.
So how did we get there?
So we met at MIT Sloan.
Personally, I was in aconsulting role out of undergrad
.
I had had some exposure to thestartup world and
(06:04):
entrepreneurship both inundergrad and outside of school,
and so I had the startup bug.
I also had the pleasure ofbeing part of a startup that
didn't have a very happy endingand realized quickly how much I
had to learn.
So my mission out of undergradwas actually to just learn, and
so that's why I went intoconsulting.
(06:24):
I figured that would give mesome general business knowledge.
Certainly got that and maybemore so which I
appreciated.
I also knew that I needed adifferent experience and to give
me an opportunity, a little bitmore of a platform, to get into
the entrepreneurship world andthe tech world.
So that's how I ended up at MITSloan.
I was part of theentrepreneurship and innovation
(06:46):
program there.
Chris Maher (06:47):
And one one quick comment on that experience early
on with the startup that let'snot call it, that didn't fail,
but that didn't quite succeedthe way you wanted to.
Those experiences are socritically important as you move
forward those learnings Right,and we can get into that a
little bit later but I raised myhand.
(07:08):
I'm part of that group too,where I've had some ones that
have done well and succeeded andothers that have not.
But those ones that did not, Imight've learned more in those
companies than the ones that youknow did really well,
Josh Miller (07:22):
You're so right and it's you know, it's hindsight
2020 situation, right, where youyou realize the things that
went wrong and the flags alongthe way and even that, you still
, hopefully, came away withsomething that you've learned
from and grown from it.
In the moment, it's prettypainful.
Yeah, a hundred percent, likeKelvin and I, just the other day
(07:42):
we hopped on the phone becausehe's going through some stuff
with with his company, See MeCane, like exciting stuff but
you know it's first time for himgoing through that, although
he's run some other companies,cause this is a little bit
unique and and you knownavigating that and Kelvin's got
this collection of kind ofadvisors that he's tapping it.
Like he's being really smartand thoughtful about it.
(08:06):
I mean, Kelvin, I don't know ifyou want to touch on that.
Kelvin Crosby (08:08):
Well, I can't tell all the details, but I mean
one of the things, it's likeyou've got to have a team around
you, because I wouldn't havegotten to where I am.
I just think about when youinvent a product, like with my
end.
This is totally off topic, butI mean we're on topic, but I
don't know.
But it was one of those thingswhere it's like if I didn't have
(08:29):
my team, I wouldn't have beenable to draft up what I needed
to draft up to make theseconversations happen.
And yeah, I mean it's excitingtimes and hopefully in the next
couple months I'll be able toshare with you guys what's going
on and we'll table that for now.
Josh, I think one of theinteresting things that you
(08:51):
talked about is working out thefounders, like building up that
core relationship, working hard.
And would you say, as you wentinto this and started building
this, like what were you guys'sexperience in the captioning
industry or in making mediaaccessible and all of that?
Josh Miller (09:14):
It's a great question, so the the honest
answer is that none of us hadreal experience in the
accessibility world directly,and what I can say is that we
all recognized it as a very realissue that needed to be solved,
and I think we all haddifferent experiences on a
personal level.
Family members who might havehad severe hearing loss or sight
(09:37):
loss that we've been a witnessto but certainly not experienced
firsthand.
But even then that is anexperience, right.
Working with a family memberwho all of a sudden doesn't hear
the same way that they did twoyears ago, I mean that's a thing
that you have to work throughand figure out how to navigate.
(09:57):
So we were aware of it.
I don't think we fullyappreciated how poorly addressed
it was in the market, certainlywith technology, and so you
know,
Chris Maher (10:08):
And Josh, this was how many years ago.
What year is this?
Josh Miller (10:12):
We all started getting to know each other in
2007.
And I think the legal paperswould say we started in 2008.
Some of the initial projectswere starting in 2007.
, and so the first conversationsthat actually got us into this
was one of my co-founders had arelationship with some people at
MIT OpenCourseWare, and thiswas the first group putting
(10:33):
lectures up online for free.
And, probably not a surprise,MIT content is not exactly
simple.
There's some pretty complicatedlanguage in there, and one of
the things that they were veryproud of is just how real these
lectures were.
And the curveball they gotthrown was that one of the
(10:54):
funding institutions that werewriting these grant checks to
get the program off the ground,kind of out of the blue, said
hey, by the way, we expect youto make the content accessible,
and of course, that was met with.
What does that mean?
Because even they who wereputting video up online, they
didn't understand it.
They weren't thinking aboutthat because nobody was.
YouTube was still pretty new atthat point and the only place
(11:16):
that you could find captionswere on broadcast television,
and even then it wasn't, it wasalmost ubiquitous, but not quite
.
The streaming world had nottaken off.
So we were in the early, earlydays of this, and so they
started going out to thetraditional broadcast television
, film production typecaptioning companies, some of
(11:37):
which still exist, and trying toget quotes to do this, and it
was going to be half theirannual budget.
So, right off the bat, they arefreaking out because they
didn't know what to do, becausethere's no way they can afford
this.
But that was the conversationthat kicked us all off.
Is us learning about this, whatwe would call absurdity on
paper?
Right, and what we had to do islearn about it.
(11:58):
We had to figure out why is itso expensive?
And at the same time, westarted working with a professor
in the spoken language lab atMIT who was part of the
artificial intelligence groupthere and looked at different
ways to use speech technology.
So how could we use speechrecognition to create captions?
And the hope was oh, that's theanswer.
(12:20):
We're going to use speechrecognition, we can do a few
things to it and, boom, we'llsolve this.
We were very far from that,especially back then, very
different from what we see in AItoday.
It was nowhere near usable forthis use case, and what we
started to see was that, one, weneed to do something more
(12:40):
interesting here and there's anopportunity to build something.
And two, if all these otherschools are talking about
putting lectures up online andhaving recordings online, this
is going to be an issue andnobody's solving this well, so
maybe we should see if there'ssomething here.
That's literally kind of theconversation we were having.
Chris Maher (13:00):
So the problem came to you?
Josh Miller (13:03):
Yeah, we were very, you said it, we were very
unusual.
We had a problem in a customerbefore we had a product.
I mean we basically, we saidwe're going to figure out a way
to get it to a third of theprice of what you just got
quoted, and like that's how westarted.
Chris Maher (13:18):
And what's amazing about that is that if you could
build that solution, you werestarting off with product market
fit, which is just socritically important.
Josh Miller (13:27):
Yeah, and we can talk about that.
I mean, I think the risk ofdoing something big for one
customer is that exact question,right?
Is this solving their problemor is this solving everyone's
problem?
I think the good news was weactually had found something
that was going to generally be aproblem for a lot of people.
How we solve it was going to bethe big question.
(13:50):
So we basically took that asinspiration and built our first
piece of software.
That was really about editingspeech recognition output, and
the concept still applies today.
We've certainly evolved it overtime, but this idea of you
start with a speech recognitiondraft and then correct it and
then have another method to QCit before it goes back to a
(14:10):
customer and there's a wholebunch of machine learning and
natural language processing thathappens in between
automatically to create theright line breaks and put in all
the other things that arerequired for captioning so that
we can do it more efficiently.
And so you know we think aboutthis concept, what we call it
XRT.
So how much time does a humanneed to contribute to fully
(14:32):
caption, say, an hour of content.
So in the past and the way wemeasured it and analyzed it, it
was getting around like a 10 Xmultiple, so 10 human hours per
one hour of video.
Kelvin Crosby (14:43):
Wow.
Josh Miller (14:44):
And we've gotten that down much, much, much lower
.
That's ultimately why the pricewas so high.
Chris Maher (14:49):
Yeah, I was just going to say, hence why that
initial it's 50% of their annualbudget.
Josh Miller (14:55):
And so I'll shout out Larry Goldberg, because we'd
be.
You know, Kelvin, you said it,the curb effect, or curb cut
effect being renamed thecaptioning effect.
Larry was instrumental ingetting captions off the ground
period for all of us way back inthe 80s, and he was nice enough
to allow us to come to theAccess Center at WGBH when we
(15:18):
were just getting started, andlearning to just see how
captions were created.
He was nice enough to walk usthrough it.
I think he questioned why hewas showing us all of this, but
he's remained a mentor and afriend for us, and it's just a
great reason we got off theground.
Chris Maher (15:35):
That story warms my heart.
He's the best.
We, as you know, we had him onthe podcast recently and, yeah,
I mean he's a pioneer in thecaptioning space and and even
back then he was mentoring andadvising and that's the bulk of
what he does today for earlystage companies which yeah he's
(16:00):
the best.
Josh Miller (16:00):
He's amazing.
I mean, he has the mission inmind all the time and he knows
that he can't alone solve it andso very much appreciate that.
Chris Maher (16:04):
Yeah, in the early on.
It's a combination.
It's this kind of hybrid modelof software and humans that are
doing it more quickly and, let'ssay, as accurately or maybe
even more accurately, can youtalk about the arc of the
technology aspect of that, howthat has evolved over time and
(16:24):
obviously AI is becoming evenmore and more a part of that.
We'd love to hear kind of howthat has evolved.
Josh Miller (16:30):
Absolutely, and so it's, you know, lots of pieces
to it.
So we look at this as almostlike two operational challenges.
One is the actual process ofcreating captions or whatever
output it might be these days.
It could be subtitles, it couldbe audio description, but the
point is actually beingefficient in creating a high
(16:50):
quality output.
Then there's doing thatthousands of times over, so
that's all internal operations,then there's the operations for
the customer, and so we had toalso figure out okay, great, we
just did this more efficientlyfor us, but is it easy for them?
And so that was the other piecethat came out really quickly
from MIT and from all theschools we spoke to is that if
(17:14):
it's not easy, we're just notgoing to be able to do it?
We don't have the we don't haveresources, the human power to
sit there and move files around.
That's just not a possibility.
So very early on we thoughtabout how do we connect systems,
how do we connect videoplatforms to actually make this
easy, and like no joke, we putsurveys out early on and we
(17:36):
found that the barrier half thetime wasn't budget, it was
actually process.
So we thought about that andthat's evolved now quite a bit
as well.
But to your point about AI, Imean, so for a long time we
thought about how AI or backthen, machine learning, so as
speech recognition got better,that essentially would reduce
the human time required tocaption effectively for a few
(17:58):
reasons.
One, obviously fewer things tocorrect, but we could even be
smarter about how to do thecorrections and how to use the
data that we would get from thespeech recognition output and
really kind of focus on certainareas and allow people to do
research certain ways andcollaborate in certain ways.
So there's a whole bunch ofthings that we continue to
(18:18):
evolve in terms of those editingtools.
But then we also built more andmore tools and heuristics
around the process of qualitycontrol and deadline management
and all these other things wehad to consider as we scaled the
business.
With the first few customers wecould just say, okay, you'll get
it back, great.
But we really had this visionof creating an environment of
(18:42):
on-demand and limitless scale,at least in the eyes of the
customer.
It's impossible to havelimitless scale with people, but
from a customer perspective, ifthey could come into the system
and upload a file and pick theturnaround or the deadline
without talking to us.
That's when we knew we actuallyhad something special.
(19:03):
Now we still want to be thereto support them, of course, but
the point is they needed to feeltotally empowered to do what
they needed when they needed it.
Ai has come a very, very longway and I think there's a lot we
could talk about there, andit's certainly changed the way
we look at certain things, andit certainly changes even our
ability to go after otherservices which I think is the
(19:24):
interesting part.
It's both threat and opportunity, depending on how you look at
it.
It also depends on the use case.
Kelvin Crosby (19:30):
Yeah, I mean, I think one of the things that
you're explaining and I think Iwant people to see is that you
were looking at the customerexperience in two different
areas.
The end user and the personthat's going to be paying the
bill.
And I think that's somethingthat a lot of people in
accessibility they go one way orthe other.
(19:51):
And and really thinking about,like even with my company right
now, like I'm having to thinkabout, like who can actually buy
my product, right, and who isthat?
But my end user are blindpeople, right, and so that's
where it's like you've got toreally think that through, and I
think people forget that,especially when they're
developing a product like youwere talking about that.
(20:14):
I think that's something thatis very kind of eye-opening and,
in a lot of ways, like okay, sothere's that.
So that's how you're going toscale in that way, and I would
love to get into thisconversation with you and really
kind of look at, all right, soyou've got your software figured
out, you got the customerexperience figured out, now
(20:35):
you've got to scale.
How were you going to do that?
What was that going to looklike.
To get to that point.
Josh Miller (20:42):
So you just hit on something I think is so
important, especially in theaccessibility space, and it's
very applicable to us, wheremany times, most of the time, we
(21:03):
are selling to a person or ateam that has been tasked with
implementing, but they'recertainly not the ones
experiencing the end product,and that is challenging,
especially if they have noexperience with any kind of
disability.
They don't really appreciatequite often how important it is
to do it a certain way.
So that is a very realchallenge and I think gets into
this whole idea of like are yousolving for compliance or for
value?
And the goal is can you combinethem, right?
(21:27):
Can you actually solvecompliance and add value at the
same time?
And that's why we talk aboutthese things like the workflow,
the customer experience, andthink about how can we save them
time, especially in a casewhere they at least have
committed to doing this, theyknow they're going to do it.
How do we make their lifeeasier?
So I think we can find ways toadd value there.
Certainly, when we think aboutaccessibility versus, say, like,
(21:48):
localization, there is a verydifferent conversation and team
thinking about these two issuesand how they go to market and
how they prove the ROI or getbudget for that.
So, I agree, it's a very realissue and if we back up even to
when we were starting thecompany, one of the biggest
issues we faced was people justdidn't understand this stuff.
(22:10):
So they certainly didn'tunderstand the value of it.
Even the compliance value?
Right, because there is valuein being compliant.
People always like to think ofit that way, but there is value.
So we, we set out on this ideaof educating the market and in
saying, hey, we need to helppeople get there because they're
not ready now.
(22:31):
So we're going to start writing,putting a blog together, and
we're going to host webinars,and we're going to go to
conferences and we're going tobasically start doing everything
we could to get the word out asto why this is a good idea.
And I will even say one of thefirst things we talked about,
before we even had a secondcustomer, was the idea that
(22:51):
captions, by definition, istime-synchronized text, and that
is a very rich tool for videosearch.
So we have tools, still to thisday, where it allows you to
click on a word, jump to thatpart of the video search through
a whole video library, and it'sstill shocking to me how few
organizations have implementedthese concepts, whether it's
(23:12):
ours or not.
I mean, it's such a great userexperience, but it's potentially
challenging to implement from aUX perspective if you don't do
it right.
So that was something that gotus really excited, though,
because that was an opportunityto change this compliance and
cost conversation to a purevalue and user experience
(23:34):
conversation, which was, youknow, I think, something that we
still strive for is like how,why is this a valuable thing.
And it certainly gets into thiscurb cut concept or conversation
that most captions are on allthe time.
Right, captions are on becausepeople have finally recognized
the value of them.
(23:54):
And I think when, for us, oneof the first times we realized
it was happening was whenFacebook released the capability
to have captions displayautomatically in the feed when
video sound or the sound was off.
And so all of a sudden you hadall these people watching video
(24:16):
and realize, at work or we'llsay on the subway, right, you
know, on their way to work,basically realizing like, oh
crap, I really like captions,right, and like these are great.
And so you know, certainly theyou know, netflix and Amazon and
Hulu and all the streamers havehelped in this by making their
libraries compliant orwell-captioned, so now they're
(24:40):
available everywhere.
And the idea of having goodcaptions on content is kind of a
table stakes now which isamazing.
Chris Maher (24:47):
So a few things just to kind of summarize there
because that was just a masterclass for an early stage founder
, an executive team, because yougot to do a lot more than just
have a good idea, and so much ofwhat you just talked about is
execution.
And, at the end of the day, youcan have a great idea, but if
(25:09):
you don't have a good team andexecute well, that great idea is
probably not going to scale andturn into anything.
And you could have a mediocreproduct with a really good team
and really good execution andyou could go crush it.
And so you've done both.
But that whole idea ofcompliance versus value and
really getting to where theystart to think about it as more
(25:30):
as value than compliance or insome cases, like, oh, you're my
insurance policy, so I don't getsued.
And so I think that wasimportant for you, and part of
that was understanding yourcustomer's workflow and their
processes.
And, by the way, this isn'tjust like an accessibility
industry thing, that's everyindustry.
I could give you so manyexamples of companies that have
(25:53):
said, yeah, your solution'sbetter, it just doesn't fit our
process and workflow internally,so we're going to buy the
product that's not as good, butit just fits what we do better.
That happens all the timeacross many, many industries.
The educational part, likeeducating the market and your
end buyer.
(26:13):
It's your buyer in this caseand the people who are executing
it versus the user, criticallyimportant, and you guys invested
in doing that, creating content, and we're going to talk about
research later.
It started with a blog, butyou've taken another level there
.
And then you identified how youcould make it beyond just
disability, right, that it'sgood for the general market, and
(26:34):
so you've expanded your TAM.
Right, that total addressablemarket, and so, like, you just
checked like half a dozen boxesabout good execution.
So, for the entrepreneurs thatfollow the podcast, God, I hope
they listened to this episodebecause I think the insights
that you shared are socritically important for an
early stage company from anexecution standpoint.
Josh Miller (26:53):
I appreciate that.
I mean, so you just hit onsomething I think is so
important, which is this ideathat the best product doesn't
always win.
And I will never forget hearingthat.
And one of the other things,and why it matters, is we cared
so deeply about having the bestproduct and doing it in a way
that was also balancing the cost, so we never tried to play the
(27:15):
price game.
That was not ourdifferentiation, but thinking
about our differentiation andhow could we differentiate has
been central to everything werelease.
So we care deeply that we aredoing it differently somehow,
some way.
Now that could be operationally, so that we can do it faster.
It could be the way thecustomer receives it, but the
goal is not to just be thecheapest because that that
(27:38):
doesn't end well, and we seewhat is happening with AI.
Chris Maher (27:43):
It's a race to the bottom.
Josh Miller (27:45):
Exactly, and so that that is something you learn
in business school for sure.
So, with that in mind, you know, and knowing that the best
product might not always win, Imean certainly you know, myself
and my co-founders cared verydeeply about not letting that
happen, to not have the bestproduct win, and so we listened
to our customers very carefully,and then from there, listening
(28:09):
to your customer can be hugelyimportant and beneficial, but it
can also be dangerous, right?
Because if you act on the wrongthing, that's going to take you
down the wrong path.
So we're constantly looking forpattern recognition and
understanding.
Okay is this a thing just forthem, or is this a thing for
everyone?
Chris Maher (28:29):
Is it the exception or is it the rule, right?
And that sometimes it's hard todifferentiate between those,
especially when it's a bigcustomer, right?
Josh Miller (28:38):
Especially early, when revenue matters a lot.
So that is something you haveto be really careful about.
It'll pay those bills tomorrow,but what happens in six months?
What is it setting us up for?
We absolutely had someconversations where we had.
I mean, there was one time wehad a project put in front of us
(28:59):
with pretty much guaranteedcash for several months.
It was absolutely a differentbusiness and we walked away from
it, even though we needed thatmoney.
Chris Maher (29:08):
Good for you.
That's really hard to do.
Josh Miller (29:12):
It was hard to do, but it was the right choice.
Chris Maher (29:14):
Yeah.
So there's two more topics I'dlike to get into, one of which
is AI.
Go a little bit deeper on AIand how that's really
transforming your business, butalso, as you see, across the
broader market.
And then we'd love to, afterthat, get a little bit into, as
much as you can share, about howyour business is going, because
(29:37):
you've done a good job ofraising money.
You've acquired some companies.
You are a success story in thatrespect, and I know you feel
like it's more about, hey we'vebeen able to progress.
But I'd like to get into thatas well, with as much as you can
share.
But let's start off with the AIand how it's impacted your
business and just what you seein the broader market going
forward.
Josh Miller (29:53):
Yeah, absolutely, and I think the stories are
linked.
They're very much linked in theevolution of AI is both kind of
informing our future and alsokind of the opportunities in
front of us.
So when we started we talkedabout it as machine learning and
it wasn't really AI at thatpoint.
(30:14):
It was pretty much there's asound wave and we have to get
text out of that sound andhopefully it'd be accurate and
there were things we could dowith modeling to improve it.
At the time we'd see what wecall word error rates, you know
15, 20, 25% on pretty regularly,which means accuracy rates or
(30:34):
the inverse of that are, youknow, 80% or so.
That's changed dramatically.
So now we're seeing accuracyrates very regularly above 90%,
which is significant, and it'scaused this kind of perception
change, if you will, kind oflike what is good enough and how
good is it actually.
AI, I mean to simplify it whenwe think about this.
(30:59):
Ai has been part of ourbusiness from day one and it's
funny because in those earlyyears people didn't trust speech
recognition.
So especially for this use caseand I think that's important is
understanding kind of theapplication of it.
That the idea that machineswere creating captions people
had seen it fail enough that ifwe said we were using speech
(31:20):
recognition to create yourcaptions, they'd say then we're
not working with you.
And so we actually hit it insome ways, we would find ways to
show that we were moreinnovative, but and we had to
stress the fact that there was ahuman involved all the time,
throughout the file, everything.
So that's evolved.
It's like the opposite.
Now it's like if you're not AI,we don't want to work with you,
(31:42):
right?
So it's certainly flipped andwe're becoming much more forward
with how we use AI in ourbusiness or talk about AI in our
business.
But it is part of everything wedo and we view it as a tool.
So our whole thing is AI isincredibly powerful.
You have to be very thoughtfulin how you implement it and how
you apply it, and if you applyit well, you can do some pretty
(32:06):
cool stuff with it and you cangain some very real efficiencies
.
But the beauty of AI is alsothe threat of AI, right.
So it can do unbelievablethings.
It's very, very powerful.
It does it very, very fast.
And if you look at some ofthese chat agents, And if you
(32:28):
look at some of these chatagents, the challenge is they're
so confident in their response,right, .
Oh of the response is like thisis the answer, oh great, well
then I'm done.
accurate unless someone checksit.
So, for the customers that wework with, so for the customers
that we work with, where the toor a brand to adhere to, or
(32:48):
certainly a direct studentaccommodation that you know
someone who truly needs it isthe one person experiencing it,
and it better be right Then youknow we need to make sure that
we're .
doing this right, and so that'swhy I say the use case and the
application matters so much.
So in our world, we view humansas critical to the process.
(33:09):
So the challenge is how do youbuild tools to insert the human
in the most efficient way?
And so we're constantlyevolving our tools.
As we think about whether it'sthe same process or new service,
we're constantly looking at howdo we continue to make the
human more efficient to do areally good job while using the
best of AI, and so we'veactually stayed AI engine
(33:30):
agnostic from the.
beginning AndWe've basically said there are
enough dollars being spent byvery, very large tech companies
that we can't compete withdirectly, but we can benefit
because the price has become solow that the real secret sauce
is how do you source the rightpeople and train the right
people and monitor the rightpeople and then put them on top
(33:52):
of the AI in a really effectiveway.
That's the hard part andcertainly the most expensive
part.
So if we can get the AI and ifthat keeps evolving, we can
basically swap in and outdifferent models and make sure
that we're using the best AI allthe time.
Kelvin Crosby (34:08):
I mean this is fascinating because I mean, I
think, when I use AI for myvideos and everything I do and I
humanly edit it, but what'sinteresting, as we grow and
using resources, and I thinkthis is something that as
entrepreneurs, as individuals,like in startups, it's like use
(34:31):
it, don't be afraid to go, useanother product to empower your
product.
I still like, for example, likein my, my company, See Me Cane.
My tube for the See Me Cane wasfor a completely different
product and then I was like canI get this tube this long and
(34:53):
this way?
And I tested it, boom, perfect,it works like a blind cane.
And I'm like, all right, let'sbuild out the electronics.
And then the rest is history.
And I think that's somethingthat's really crucial as you are
evolving through this processis using other resources to
(35:13):
lower your bottom line, and Ithink that that's something
really powerful.
Josh Miller (35:18):
Yeah, it's a really good point, I think you know,
recognizing where your strengthsreally are and kind of what
it's going to take to compete.
And we've seen several companiesin different spaces, both in
the captioning, accessibilityand in localization, already
full or or kind of move on fromfrom this path because they
(35:39):
tried to build their own engineand it was just so expensive.
It's very hard, and thensomeone surpasses them very
quickly and they have to figureout who are we.
Chris Maher (35:51):
Yeah, your model, and I said where AI and humans
to create your own solutions.
But now you're taking that andtaking that same methodology and
framework by taking all thestuff that you're building with
(36:12):
your software and people andconnecting it to third parties
like the Googles and theMicrosofts and the open AIs that
have built, as you said, putbillions of dollars into these
engines, that the way you getthe best value out of them is
you figure out how to use themto your specific customer or use
cases.
And so now you guys, yourspecial sauce is that technical
(36:35):
and human integration within andacross your business to
external parties and that's noteasy to do, that's super hard.
And also being agnostic you cankind of cherry pick the best in
class, right?
Josh Miller (36:53):
Yeah, it's become really interesting in the last
couple of years with Gen AItaking off, and especially with
a good example is voicetechnology.
So text to speech and what itcan do, whether we're talking
audio description or dubbing inmultiple languages.
That you know, certainly in themedia space, where they care
deeply about the quality of themix and the voice, people want
(37:15):
to know that you're using thebest voicing, and we have to be
able to say we.
And so the service we're nowproviding in a funny way is
actually being your curator ofAI technology, and that's a very
real thing and so you're wellaware that we do this state of
ASR report, which is state ofautomatic speech recognition.
(37:37):
We've done that for severalyears now and we're literally
benchmarking all the enginesagainst each other, because it's
both beneficial for us, butit's actually really interesting
data to share with the marketto see into a level set that
certain engines perform betterin certain cases.
Kelvin Crosby (37:52):
Right.
Josh Miller (37:53):
The content, and it all depends on how they've been
trained, and so that issomething that we intend to
expand.
Now that we're getting more andmore into localization.
We're seeing some fascinatingthings with text-to-speech in
these voices.
Some do really well in certainlanguages, some do really well
with emotion, so we have to takeall that into account and let
(38:14):
our customers really get valueout of the fact that we're
paying attention to that forthem.
Chris Maher (38:19):
Yeah, and the byproduct of you doing that sort
of research right is, one, itreinforces your thought
leadership in the marketplace,which then leads to credibility
for you, right, with themarketplace, but it's also
driving your product roadmap andmaking your products better.
So it's super smart.
Josh Miller (38:36):
Exaxtly.
So that was one of the thingsthat we figured out early on
with the state of ASR report isprobably now five or six years,
at least, that we've been doingit.
We were doing it anyways.
We were benchmarking the enginesanyways because it was helping
us create more efficiency in ourprocess, and then we realized
people want to read this.
People actually find thisreally interesting and it's one
(38:57):
of our most sought after piecesof content every year, so that's
been very, very helpful for us.
Chris Maher (39:05):
Maybe, as we start to wrap up here, let's touch on
with what you can share aroundhow the business has grown over
the years, and that's throughhaving to raise capital, making
some important decisions aboutwhether it's targeting certain
markets or go-to-marketstrategies.
You've acquired some companies.
You got into a relationshipwith selling a portion of the
(39:26):
company to private equity.
What sort of insights can youshare about that process and
things that you would want totell the younger you or an early
stage entrepreneur that's justembarking on that journey?
Josh Miller (39:38):
Yeah, it's a good question.
I mean, so we grew organicallyfor a long time.
We started the business in 2008.
We did an angel round in 2010,small one and then grew
organically and we wereprofitable early on, which we
cared a lot about and our earlyinvestors felt like that was
very important.
And lo and behold, it matters.
(40:00):
So, that was great.
We had a certain level ofstability for a very, very long
time.
Chris Maher (40:07):
Can I ask a quick question about that decision to
focus on profitability versusgrowing top line revenue?
Did you have investors and aboard that were supporting you
on that, or did you have tofight them on that?
Josh Miller (40:22):
So it's a great question.
I think, early on, part of thechallenge we faced was being in
an emerging market.
And being especially in onethat was still relatively niche
is the honest truth.
As much as I'd like to thinkthat the captioning market is
billions and billions of dollars.
It's not, and that was part ofthe challenge we had, is kind of
(40:44):
how fast can we actually grow?
And we had to be honest withourselves about that, both in
terms of our capabilities butalso the market.
And that is something that isso important and as we think
about our future endeavors, themarket opportunity matters.
So that was part of it.
Part of it was driven by thetiming and the market size and
(41:04):
being real with ourselves aboutif we were to invest in more
revenue growth, could weactually achieve it?
Was it realistic?
So we were trying to balancethat and be really thoughtful
about that.
We never played the game ofhire 20 sales reps and see who
survives and like all that kindof stuff.
We just didn't do that.
It wasn't really in our DNAeither.
(41:25):
So we we kind of grew with themarket and, as I said, you know
we're trying to educate themarket at the same time.
So it was it was a prettymeasured approach.
We had a couple of keyinvestors who cared deeply that
we be profitable.
I mean, their experience wasactually with much larger
companies where profitability isvery, very important.
So that was in their DNA.
So it kind of like all cametogether and that's how it
(41:48):
worked and I think it served uswell.
I think understanding the marketyou're in and the type of
business you have and whether itis capable of scaling quickly
is so important.
So I don't know that we wouldhave done things that
dramatically differently.
We certainly saw competitorscome into the market, in some
(42:08):
cases, come and go in thattimeframe, but you know, I think
we did a good job ofmaintaining who we were in the
market and certainly theperception of us in the market,
and in making sure that westayed true to ourselves and
continuing to deliver a reallygood product.
I mean, I think one of thethings about the accessibility
market which I hope peopleappreciate, and even our
(42:31):
competitors, is doing a poor jobit just feels really messed up,
honestly.
Like you gotta care, and if youcan't get yourself to care,
then you're in the wrongindustry.
And so that was something thatwe felt and cared about and you
know, lucky for us, we're ableto hire a really great team
(42:53):
around us that also felt thatway.
So if you look at our teamtoday even, we have a really
passionate team.
People really care about makinga difference and doing this the
right way.
So I do think that's somethingthat we've kept true to
ourselves even throughinvestment and acquisition, and
I think that's a hard thing todo with acquisition certainly,
(43:16):
so kind of skipping a few stepsthere.
We partnered with a privateequity firm in 2019.
And there were a few reasonsfor that.
I mean, part of it was thismarket was evolving, this market
was actually growing andproving to be a real market and
we needed to go a little faster.
We needed to have a little bitmore expertise on our board and
(43:39):
we had a few you know dynamicsof you know our angel investors
had been with us for nine years,so we needed to create some,
you know, an off ramp for them.
And so there was a whole bunchof things happening at the same
time and it made it for to bethe right decision to to bring
on a new financial sponsor andthen with that that opens the
doors into a whole bunch ofother things.
So we had the opportunity toacquire a couple of companies.
(44:01):
It was kind of two at the sametime in CaptionMax and National
Captioning Canada and that wasat a time when we were feeling
really strongly that we wantedto expand into this live
captioning use case and saw thatas a market expansion
opportunity.
I will say, tying back to AI,the timing of it was interesting
in that several months later,ChatGPT came out and all of a
(44:26):
sudden, and even in speechtechnology in our measurements,
we saw a step function change inwhat was possible and we kind
of looked at ourselves and saidoh wow, okay, this is
interesting timing.
And then, on top of it, theeconomy collapsed and there was
a media strike and the resourcesthat we acquired were very
(44:49):
heavy in the media industry.
So there was a lot happenedthat we learned a lot from, and
so we had some, we've beenthrough some bumps and I think
you learn a lot from that.
It's certainly blood pressurerises at times in ways that are
probably not the healthiest, butI mean, I will say we came out
the other end.
(45:09):
But you know, it certainlystarts to inform like where,
where are we going.
Right?
And really try to understandthe market and what is AI now
and how are people viewing AI?
Cause it's one thing to havethe best product, it's another
to make sure your customersunderstand it's the best product
and it's the best product forthem and for their use case.
So, that is definitelysomething that's happened with
(45:30):
AI, and the evolution of AI isthat people's perception have
changed and that they want it towork.
They want it to solve all theirproblems.
Whether it can, you know, it'sanother story sometimes, but
it's gotten a lot better and wehave to be honest with ourselves
about that and we have torecognize that and just be
pragmatic in the way we run ourbusiness and where we invest.
Chris Maher (45:52):
Yeah, that was awesome.
We could talk to you for hourshere.
We're going to have to bringyou back and have like a part
two of this.
Kelvin Crosby (46:07):
So hours Josh here we're gonna have to bring
you back and have like a part
So, for better for worse, wetry to be pretty open and
available, and so we have apretty solid web presence.
We put a lot of content on ourwebsite and on our social
channels with, like I said,trying to educate first and and
be part of the community.
You go to a number of events tobe part of the community.
I think that's really important, so a lot of information about
3Play is out there.
(46:28):
I'm on LinkedIn, I'm around.
I love having the conversationsabout other startup companies
getting going in this space.
I have made some greatconnections with other founders
that are doing some reallyinteresting things in this space
even even, in some cases, veryclose to what we're doing, which
is in some ways excitingbecause it creates partnership
opportunities.
But I think there's so much todo .
(46:50):
here And and again, if we cancontinue that conversation of
how we get from compliance tovalue with technology, I mean
it's huge and it's definitely asituation where rising tides
raises all boats, because peopleneed to understand this before
they're going to buy it.
So very, very big proponent ofthat so we're out there and we
(47:12):
we can be found.
Chris Maher (47:14):
.
We'll put a bunch of links inthe show notes for all of our
listeners and, josh Josh thankyou so much for taking time out
of your busy schedule.
Congrats to you and your teamand the company that you've
built, and we wish you all thesuccess moving forward, and
we're looking forward to havingyou back on the pod so we can do
a part two of this conversation.
Josh Miller (47:33):
But thanks for joining us.
We have a whole future to talkthrough, what's going to happen
next, so I appreciate you guyshaving me and I think what you
guys are doing and focusing onthis emerging world and the
accessibility space is soimportant.
So thanks for doing it.
Kelvin Crosby (47:47):
This wraps up Investing in Accessibility.
Go live beyond your challengesand we'll see you in two weeks.
Thank you for listening toInvesting in Accessibility, a
Samaritan Partners podcast wherewe invest in change for
(48:08):
accessibility, not wait forchange.
If you want to follow us, youcan find us on YouTube or
LinkedIn at @Samaritan Partners.
If you would like to invest inSamaritan Partners, email Chris
at chris@samaritanpartners.
com.
If you'd like to learn moreabout us, go to www.
samaritanpartners.
com.
You can take the first step ininvesting in change by giving us
(48:32):
five stars and sharing thispodcast with everybody that you
know, so we can spread the word,so that we can give access to
all by Investing inAccessibility.
Welcome to Investing in Accessibility, a
Samaritan Partners podcast.
We're not waiting for change,we're investing in it.
Join us as we speak withentrepreneurs and thought
leaders that are focused oncreating a more accessible world
(00:28):
.
Hey, It's so good to see you,even though I can't see you.
It's another beautiful day inthe neighborhood and I'm so
excited that you're here atInvesting in Accessibility.
And we're going to be talkingabout the captioning effect.
I think we're going to upgradeit from the curb cut effect to
the captioning effect today, butbefore we get there, let me
introduce you to my co-host,Chris Maher.
(00:50):
How you doing, man?
Chris Maher (00:52):
Good, Kelvin, it's good to see you and to be with
you, my friend.
We've both been a little bitbusy.
Kelvin Crosby (00:57):
Oh yeah.
Chris Maher (00:58):
We've been burning the candle at both ends, and
that might be a separate episodewhere we dig into that.
But how are you, buddy, you'vegot so much on your plate, right
now?
Kelvin Crosby (01:05):
Oh, I'm trying to survive between parenting,
between running two differentbusinesses, launching a new
series on the DeafBlind Potterplatform and then working the
See Me Cane and just trying toget the last version of the
prototype done.
I'm just like I just need tobreathe .
Chris Maher (01:28):
Your ears must have been earlier this week because
I was at that conference,Disability:IN and I ran into our
friend, Troy from Aira, whointroduced us so, so grateful
for that, and Mike May and wewere, we were talking about you
and and it was all glowing andso your ears must have been
burning a couple days ago whenwe were talking about you at the
(01:49):
conference.
Kelvin Crosby (01:50):
Well, that's awesome.
Well, I tell you, we had anepisode a couple months ago that
we talked about upgrading fromthe curb cut effect to
captioning effect, and I'mexcited about today's
conversation.
So why don't you introduce usto our guest?
Chris Maher (02:08):
You got it.
So our guest today is JoshMiller, who's the co-founder and
CEO of 3Play Media, and beforeI kind of talk about how we
connected, let's welcome him tothe show.
So, Josh, welcome to thepodcast.
Josh Miller (02:19):
Thank you.
Appreciate you guys having meon.
I'm excited to have theconversation.
Chris Maher (02:23):
Well it's good to have you here, and so we, as we
were talking before we startedrecording, we met at M- Enabling
and I think it was the fall ofOctober 2023.
And I think I was walking bythe 3Play Media booth and I
caught you there and you weregracious, and I literally just
launched the Samaritan Fund.
And I think you spent like 15minutes chatting with me and
(02:47):
telling me like, oh, meet thisperson and check out this person
.
So you were incredibly graciousand kind to me at that time and
then we stayed in touch and webumped into each other at the
conferences.
And then again your kindness,recently you reached out because
you recommended me to moderatea panel for a conference in
Indianapolis in the fall, whichI'm going to do by the way, and
so thank you for that.
But yeah, no, you've just been,you've been kind and generous
(03:09):
and have become a friend overtime, and so it's great to have
you on the podcast and to diginto your business.
Josh Miller (03:16):
Oh, that's great and, I think, for everyone.
I mean, I'm a huge believer inentrepreneurship and innovation
and the more we can do,especially in the accessibility
space, to promote that growthmindset and innovation here is
such a good thing for all of us,and I know that we've certainly
benefited from just generalawareness growing in the market
(03:38):
and if we can do more of that tohelp each other, we should.
Chris Maher (03:41):
Yeah, love it.
And so today we're going to getinto a number of topics, but at
a high level,big reason why Kelvin and I
wanted to have Josh on thepodcast and to talk about his
business is, as this podcast isthe intersection of
entrepreneurship, accessibilityand impact investing, part of
(04:02):
what we want to do is certainlytalk about early stage
innovation right, that's justkind of getting off the ground,
but we also need to highlightthe success stories, and 3Play
Media, and you and your team,are absolutely one of those
examples.
And so we need to start shininga light on that, because we
don't have a ton of them so far.
Hopefully we're going to have alot more, and so we need to be
(04:23):
talking more about the companiesthat have scaled to that next
level, or the next severallevels, as examples for private
capital that hasn't yet startedto look at this space as an
investment opportunity.
So with that being the tablesetting let's first start off,
Josh, because it's a little bitunique, I think, in terms of how
the business started.
Talk about, maybe, yourbackground first prior to
(04:48):
starting the business and, ifyou want to get into maybe your
connection or lived experiencearound disability and then how
the founding of 3Play Mediahappened, would be great, let's
start there?
Josh Miller (04:59):
Yeah, definitely, and I appreciate very, very kind
words.
I think we would say we've madea lot of progress but we still
have a lot to solve, butdefinitely excited to talk about
it.
So we started the company atMIT Sloan.
There were four of us whoco-founded the business.
Two of us continue to run ittoday and the other two are
still good friends and areactively in conversations about
(05:20):
how things are going.
I will say right off the bathaving four co-founders is a
real blessing, because you'vegot four people who are just
working their butts off, we'llsay, and just doing anything to
see the business forward andmake progress together.
Certainly, who you have as yourco-founders matters and making
sure that that alignment exists,but we were really lucky to
(05:43):
have a tight group of people whoare just so focused on getting
to the next step and pushingthings forward together.
So how did we get there?
So we met at MIT Sloan.
Personally, I was in aconsulting role out of undergrad
.
I had had some exposure to thestartup world and
(06:04):
entrepreneurship both inundergrad and outside of school,
and so I had the startup bug.
I also had the pleasure ofbeing part of a startup that
didn't have a very happy endingand realized quickly how much I
had to learn.
So my mission out of undergradwas actually to just learn, and
so that's why I went intoconsulting.
(06:24):
I figured that would give mesome general business knowledge.
Certainly got that and maybemore so which I
appreciated.
I also knew that I needed adifferent experience and to give
me an opportunity, a little bitmore of a platform, to get into
the entrepreneurship world andthe tech world.
So that's how I ended up at MITSloan.
I was part of theentrepreneurship and innovation
(06:46):
program there.
Chris Maher (06:47):
And one one quick comment on that experience early
on with the startup that let'snot call it, that didn't fail,
but that didn't quite succeedthe way you wanted to.
Those experiences are socritically important as you move
forward those learnings Right,and we can get into that a
little bit later but I raised myhand.
(07:08):
I'm part of that group too,where I've had some ones that
have done well and succeeded andothers that have not.
But those ones that did not, Imight've learned more in those
companies than the ones that youknow did really well,
Josh Miller (07:22):
You're so right and it's you know, it's hindsight
2020 situation, right, where youyou realize the things that
went wrong and the flags alongthe way and even that, you still
, hopefully, came away withsomething that you've learned
from and grown from it.
In the moment, it's prettypainful.
Yeah, a hundred percent, likeKelvin and I, just the other day
(07:42):
we hopped on the phone becausehe's going through some stuff
with with his company, See MeCane, like exciting stuff but
you know it's first time for himgoing through that, although
he's run some other companies,cause this is a little bit
unique and and you knownavigating that and Kelvin's got
this collection of kind ofadvisors that he's tapping it.
Like he's being really smartand thoughtful about it.
(08:06):
I mean, Kelvin, I don't know ifyou want to touch on that.
Kelvin Crosby (08:08):
Well, I can't tell all the details, but I mean
one of the things, it's likeyou've got to have a team around
you, because I wouldn't havegotten to where I am.
I just think about when youinvent a product, like with my
end.
This is totally off topic, butI mean we're on topic, but I
don't know.
But it was one of those thingswhere it's like if I didn't have
(08:29):
my team, I wouldn't have beenable to draft up what I needed
to draft up to make theseconversations happen.
And yeah, I mean it's excitingtimes and hopefully in the next
couple months I'll be able toshare with you guys what's going
on and we'll table that for now.
Josh, I think one of theinteresting things that you
(08:51):
talked about is working out thefounders, like building up that
core relationship, working hard.
And would you say, as you wentinto this and started building
this, like what were you guys'sexperience in the captioning
industry or in making mediaaccessible and all of that?
Josh Miller (09:14):
It's a great question, so the the honest
answer is that none of us hadreal experience in the
accessibility world directly,and what I can say is that we
all recognized it as a very realissue that needed to be solved,
and I think we all haddifferent experiences on a
personal level.
Family members who might havehad severe hearing loss or sight
(09:37):
loss that we've been a witnessto but certainly not experienced
firsthand.
But even then that is anexperience, right.
Working with a family memberwho all of a sudden doesn't hear
the same way that they did twoyears ago, I mean that's a thing
that you have to work throughand figure out how to navigate.
(09:57):
So we were aware of it.
I don't think we fullyappreciated how poorly addressed
it was in the market, certainlywith technology, and so you
know,
Chris Maher (10:08):
And Josh, this was how many years ago.
What year is this?
Josh Miller (10:12):
We all started getting to know each other in
2007.
And I think the legal paperswould say we started in 2008.
Some of the initial projectswere starting in 2007.
, and so the first conversationsthat actually got us into this
was one of my co-founders had arelationship with some people at
MIT OpenCourseWare, and thiswas the first group putting
(10:33):
lectures up online for free.
And, probably not a surprise,MIT content is not exactly
simple.
There's some pretty complicatedlanguage in there, and one of
the things that they were veryproud of is just how real these
lectures were.
And the curveball they gotthrown was that one of the
(10:54):
funding institutions that werewriting these grant checks to
get the program off the ground,kind of out of the blue, said
hey, by the way, we expect youto make the content accessible,
and of course, that was met with.
What does that mean?
Because even they who wereputting video up online, they
didn't understand it.
They weren't thinking aboutthat because nobody was.
YouTube was still pretty new atthat point and the only place
(11:16):
that you could find captionswere on broadcast television,
and even then it wasn't, it wasalmost ubiquitous, but not quite
.
The streaming world had nottaken off.
So we were in the early, earlydays of this, and so they
started going out to thetraditional broadcast television
, film production typecaptioning companies, some of
(11:37):
which still exist, and trying toget quotes to do this, and it
was going to be half theirannual budget.
So, right off the bat, they arefreaking out because they
didn't know what to do, becausethere's no way they can afford
this.
But that was the conversationthat kicked us all off.
Is us learning about this, whatwe would call absurdity on
paper?
Right, and what we had to do islearn about it.
(11:58):
We had to figure out why is itso expensive?
And at the same time, westarted working with a professor
in the spoken language lab atMIT who was part of the
artificial intelligence groupthere and looked at different
ways to use speech technology.
So how could we use speechrecognition to create captions?
And the hope was oh, that's theanswer.
(12:20):
We're going to use speechrecognition, we can do a few
things to it and, boom, we'llsolve this.
We were very far from that,especially back then, very
different from what we see in AItoday.
It was nowhere near usable forthis use case, and what we
started to see was that, one, weneed to do something more
(12:40):
interesting here and there's anopportunity to build something.
And two, if all these otherschools are talking about
putting lectures up online andhaving recordings online, this
is going to be an issue andnobody's solving this well, so
maybe we should see if there'ssomething here.
That's literally kind of theconversation we were having.
Chris Maher (13:00):
So the problem came to you?
Josh Miller (13:03):
Yeah, we were very, you said it, we were very
unusual.
We had a problem in a customerbefore we had a product.
I mean we basically, we saidwe're going to figure out a way
to get it to a third of theprice of what you just got
quoted, and like that's how westarted.
Chris Maher (13:18):
And what's amazing about that is that if you could
build that solution, you werestarting off with product market
fit, which is just socritically important.
Josh Miller (13:27):
Yeah, and we can talk about that.
I mean, I think the risk ofdoing something big for one
customer is that exact question,right?
Is this solving their problemor is this solving everyone's
problem?
I think the good news was weactually had found something
that was going to generally be aproblem for a lot of people.
How we solve it was going to bethe big question.
(13:50):
So we basically took that asinspiration and built our first
piece of software.
That was really about editingspeech recognition output, and
the concept still applies today.
We've certainly evolved it overtime, but this idea of you
start with a speech recognitiondraft and then correct it and
then have another method to QCit before it goes back to a
(14:10):
customer and there's a wholebunch of machine learning and
natural language processing thathappens in between
automatically to create theright line breaks and put in all
the other things that arerequired for captioning so that
we can do it more efficiently.
And so you know we think aboutthis concept, what we call it
XRT.
So how much time does a humanneed to contribute to fully
(14:32):
caption, say, an hour of content.
So in the past and the way wemeasured it and analyzed it, it
was getting around like a 10 Xmultiple, so 10 human hours per
one hour of video.
Kelvin Crosby (14:43):
Wow.
Josh Miller (14:44):
And we've gotten that down much, much, much lower
.
That's ultimately why the pricewas so high.
Chris Maher (14:49):
Yeah, I was just going to say, hence why that
initial it's 50% of their annualbudget.
Josh Miller (14:55):
And so I'll shout out Larry Goldberg, because we'd
be.
You know, Kelvin, you said it,the curb effect, or curb cut
effect being renamed thecaptioning effect.
Larry was instrumental ingetting captions off the ground
period for all of us way back inthe 80s, and he was nice enough
to allow us to come to theAccess Center at WGBH when we
(15:18):
were just getting started, andlearning to just see how
captions were created.
He was nice enough to walk usthrough it.
I think he questioned why hewas showing us all of this, but
he's remained a mentor and afriend for us, and it's just a
great reason we got off theground.
Chris Maher (15:35):
That story warms my heart.
He's the best.
We, as you know, we had him onthe podcast recently and, yeah,
I mean he's a pioneer in thecaptioning space and and even
back then he was mentoring andadvising and that's the bulk of
what he does today for earlystage companies which yeah he's
(16:00):
the best.
Josh Miller (16:00):
He's amazing.
I mean, he has the mission inmind all the time and he knows
that he can't alone solve it andso very much appreciate that.
Chris Maher (16:04):
Yeah, in the early on.
It's a combination.
It's this kind of hybrid modelof software and humans that are
doing it more quickly and, let'ssay, as accurately or maybe
even more accurately, can youtalk about the arc of the
technology aspect of that, howthat has evolved over time and
(16:24):
obviously AI is becoming evenmore and more a part of that.
We'd love to hear kind of howthat has evolved.
Josh Miller (16:30):
Absolutely, and so it's, you know, lots of pieces
to it.
So we look at this as almostlike two operational challenges.
One is the actual process ofcreating captions or whatever
output it might be these days.
It could be subtitles, it couldbe audio description, but the
point is actually beingefficient in creating a high
(16:50):
quality output.
Then there's doing thatthousands of times over, so
that's all internal operations,then there's the operations for
the customer, and so we had toalso figure out okay, great, we
just did this more efficientlyfor us, but is it easy for them?
And so that was the other piecethat came out really quickly
from MIT and from all theschools we spoke to is that if
(17:14):
it's not easy, we're just notgoing to be able to do it?
We don't have the we don't haveresources, the human power to
sit there and move files around.
That's just not a possibility.
So very early on we thoughtabout how do we connect systems,
how do we connect videoplatforms to actually make this
easy, and like no joke, we putsurveys out early on and we
(17:36):
found that the barrier half thetime wasn't budget, it was
actually process.
So we thought about that andthat's evolved now quite a bit
as well.
But to your point about AI, Imean, so for a long time we
thought about how AI or backthen, machine learning, so as
speech recognition got better,that essentially would reduce
the human time required tocaption effectively for a few
(17:58):
reasons.
One, obviously fewer things tocorrect, but we could even be
smarter about how to do thecorrections and how to use the
data that we would get from thespeech recognition output and
really kind of focus on certainareas and allow people to do
research certain ways andcollaborate in certain ways.
So there's a whole bunch ofthings that we continue to
(18:18):
evolve in terms of those editingtools.
But then we also built more andmore tools and heuristics
around the process of qualitycontrol and deadline management
and all these other things wehad to consider as we scaled the
business.
With the first few customers wecould just say, okay, you'll get
it back, great.
But we really had this visionof creating an environment of
(18:42):
on-demand and limitless scale,at least in the eyes of the
customer.
It's impossible to havelimitless scale with people, but
from a customer perspective, ifthey could come into the system
and upload a file and pick theturnaround or the deadline
without talking to us.
That's when we knew we actuallyhad something special.
(19:03):
Now we still want to be thereto support them, of course, but
the point is they needed to feeltotally empowered to do what
they needed when they needed it.
Ai has come a very, very longway and I think there's a lot we
could talk about there, andit's certainly changed the way
we look at certain things, andit certainly changes even our
ability to go after otherservices which I think is the
(19:24):
interesting part.
It's both threat and opportunity, depending on how you look at
it.
It also depends on the use case.
Kelvin Crosby (19:30):
Yeah, I mean, I think one of the things that
you're explaining and I think Iwant people to see is that you
were looking at the customerexperience in two different
areas.
The end user and the personthat's going to be paying the
bill.
And I think that's somethingthat a lot of people in
accessibility they go one way orthe other.
(19:51):
And and really thinking about,like even with my company right
now, like I'm having to thinkabout, like who can actually buy
my product, right, and who isthat?
But my end user are blindpeople, right, and so that's
where it's like you've got toreally think that through, and I
think people forget that,especially when they're
developing a product like youwere talking about that.
(20:14):
I think that's something thatis very kind of eye-opening and,
in a lot of ways, like okay, sothere's that.
So that's how you're going toscale in that way, and I would
love to get into thisconversation with you and really
kind of look at, all right, soyou've got your software figured
out, you got the customerexperience figured out, now
(20:35):
you've got to scale.
How were you going to do that?
What was that going to looklike.
To get to that point.
Josh Miller (20:42):
So you just hit on something I think is so
important, especially in theaccessibility space, and it's
very applicable to us, wheremany times, most of the time, we
(21:03):
are selling to a person or ateam that has been tasked with
implementing, but they'recertainly not the ones
experiencing the end product,and that is challenging,
especially if they have noexperience with any kind of
disability.
They don't really appreciatequite often how important it is
to do it a certain way.
So that is a very realchallenge and I think gets into
this whole idea of like are yousolving for compliance or for
value?
And the goal is can you combinethem, right?
(21:27):
Can you actually solvecompliance and add value at the
same time?
And that's why we talk aboutthese things like the workflow,
the customer experience, andthink about how can we save them
time, especially in a casewhere they at least have
committed to doing this, theyknow they're going to do it.
How do we make their lifeeasier?
So I think we can find ways toadd value there.
Certainly, when we think aboutaccessibility versus, say, like,
(21:48):
localization, there is a verydifferent conversation and team
thinking about these two issuesand how they go to market and
how they prove the ROI or getbudget for that.
So, I agree, it's a very realissue and if we back up even to
when we were starting thecompany, one of the biggest
issues we faced was people justdidn't understand this stuff.
(22:10):
So they certainly didn'tunderstand the value of it.
Even the compliance value?
Right, because there is valuein being compliant.
People always like to think ofit that way, but there is value.
So we, we set out on this ideaof educating the market and in
saying, hey, we need to helppeople get there because they're
not ready now.
(22:31):
So we're going to start writing,putting a blog together, and
we're going to host webinars,and we're going to go to
conferences and we're going tobasically start doing everything
we could to get the word out asto why this is a good idea.
And I will even say one of thefirst things we talked about,
before we even had a secondcustomer, was the idea that
(22:51):
captions, by definition, istime-synchronized text, and that
is a very rich tool for videosearch.
So we have tools, still to thisday, where it allows you to
click on a word, jump to thatpart of the video search through
a whole video library, and it'sstill shocking to me how few
organizations have implementedthese concepts, whether it's
(23:12):
ours or not.
I mean, it's such a great userexperience, but it's potentially
challenging to implement from aUX perspective if you don't do
it right.
So that was something that gotus really excited, though,
because that was an opportunityto change this compliance and
cost conversation to a purevalue and user experience
(23:34):
conversation, which was, youknow, I think, something that we
still strive for is like how,why is this a valuable thing.
And it certainly gets into thiscurb cut concept or conversation
that most captions are on allthe time.
Right, captions are on becausepeople have finally recognized
the value of them.
(23:54):
And I think when, for us, oneof the first times we realized
it was happening was whenFacebook released the capability
to have captions displayautomatically in the feed when
video sound or the sound was off.
And so all of a sudden you hadall these people watching video
(24:16):
and realize, at work or we'llsay on the subway, right, you
know, on their way to work,basically realizing like, oh
crap, I really like captions,right, and like these are great.
And so you know, certainly theyou know, netflix and Amazon and
Hulu and all the streamers havehelped in this by making their
libraries compliant orwell-captioned, so now they're
(24:40):
available everywhere.
And the idea of having goodcaptions on content is kind of a
table stakes now which isamazing.
Chris Maher (24:47):
So a few things just to kind of summarize there
because that was just a masterclass for an early stage founder
, an executive team, because yougot to do a lot more than just
have a good idea, and so much ofwhat you just talked about is
execution.
And, at the end of the day, youcan have a great idea, but if
(25:09):
you don't have a good team andexecute well, that great idea is
probably not going to scale andturn into anything.
And you could have a mediocreproduct with a really good team
and really good execution andyou could go crush it.
And so you've done both.
But that whole idea ofcompliance versus value and
really getting to where theystart to think about it as more
(25:30):
as value than compliance or insome cases, like, oh, you're my
insurance policy, so I don't getsued.
And so I think that wasimportant for you, and part of
that was understanding yourcustomer's workflow and their
processes.
And, by the way, this isn'tjust like an accessibility
industry thing, that's everyindustry.
I could give you so manyexamples of companies that have
(25:53):
said, yeah, your solution'sbetter, it just doesn't fit our
process and workflow internally,so we're going to buy the
product that's not as good, butit just fits what we do better.
That happens all the timeacross many, many industries.
The educational part, likeeducating the market and your
end buyer.
(26:13):
It's your buyer in this caseand the people who are executing
it versus the user, criticallyimportant, and you guys invested
in doing that, creating content, and we're going to talk about
research later.
It started with a blog, butyou've taken another level there
.
And then you identified how youcould make it beyond just
disability, right, that it'sgood for the general market, and
(26:34):
so you've expanded your TAM.
Right, that total addressablemarket, and so, like, you just
checked like half a dozen boxesabout good execution.
So, for the entrepreneurs thatfollow the podcast, God, I hope
they listened to this episodebecause I think the insights
that you shared are socritically important for an
early stage company from anexecution standpoint.
Josh Miller (26:53):
I appreciate that.
I mean, so you just hit onsomething I think is so
important, which is this ideathat the best product doesn't
always win.
And I will never forget hearingthat.
And one of the other things,and why it matters, is we cared
so deeply about having the bestproduct and doing it in a way
that was also balancing the cost, so we never tried to play the
(27:15):
price game.
That was not ourdifferentiation, but thinking
about our differentiation andhow could we differentiate has
been central to everything werelease.
So we care deeply that we aredoing it differently somehow,
some way.
Now that could be operationally, so that we can do it faster.
It could be the way thecustomer receives it, but the
goal is not to just be thecheapest because that that
(27:38):
doesn't end well, and we seewhat is happening with AI.
Chris Maher (27:43):
It's a race to the bottom.
Josh Miller (27:45):
Exactly, and so that that is something you learn
in business school for sure.
So, with that in mind, you know, and knowing that the best
product might not always win, Imean certainly you know, myself
and my co-founders cared verydeeply about not letting that
happen, to not have the bestproduct win, and so we listened
to our customers very carefully,and then from there, listening
(28:09):
to your customer can be hugelyimportant and beneficial, but it
can also be dangerous, right?
Because if you act on the wrongthing, that's going to take you
down the wrong path.
So we're constantly looking forpattern recognition and
understanding.
Okay is this a thing just forthem, or is this a thing for
everyone?
Chris Maher (28:29):
Is it the exception or is it the rule, right?
And that sometimes it's hard todifferentiate between those,
especially when it's a bigcustomer, right?
Josh Miller (28:38):
Especially early, when revenue matters a lot.
So that is something you haveto be really careful about.
It'll pay those bills tomorrow,but what happens in six months?
What is it setting us up for?
We absolutely had someconversations where we had.
I mean, there was one time wehad a project put in front of us
(28:59):
with pretty much guaranteedcash for several months.
It was absolutely a differentbusiness and we walked away from
it, even though we needed thatmoney.
Chris Maher (29:08):
Good for you.
That's really hard to do.
Josh Miller (29:12):
It was hard to do, but it was the right choice.
Chris Maher (29:14):
Yeah.
So there's two more topics I'dlike to get into, one of which
is AI.
Go a little bit deeper on AIand how that's really
transforming your business, butalso, as you see, across the
broader market.
And then we'd love to, afterthat, get a little bit into, as
much as you can share, about howyour business is going, because
(29:37):
you've done a good job ofraising money.
You've acquired some companies.
You are a success story in thatrespect, and I know you feel
like it's more about, hey we'vebeen able to progress.
But I'd like to get into thatas well, with as much as you can
share.
But let's start off with the AIand how it's impacted your
business and just what you seein the broader market going
forward.
Josh Miller (29:53):
Yeah, absolutely, and I think the stories are
linked.
They're very much linked in theevolution of AI is both kind of
informing our future and alsokind of the opportunities in
front of us.
So when we started we talkedabout it as machine learning and
it wasn't really AI at thatpoint.
(30:14):
It was pretty much there's asound wave and we have to get
text out of that sound andhopefully it'd be accurate and
there were things we could dowith modeling to improve it.
At the time we'd see what wecall word error rates, you know
15, 20, 25% on pretty regularly,which means accuracy rates or
(30:34):
the inverse of that are, youknow, 80% or so.
That's changed dramatically.
So now we're seeing accuracyrates very regularly above 90%,
which is significant, and it'scaused this kind of perception
change, if you will, kind oflike what is good enough and how
good is it actually.
AI, I mean to simplify it whenwe think about this.
(30:59):
Ai has been part of ourbusiness from day one and it's
funny because in those earlyyears people didn't trust speech
recognition.
So especially for this use caseand I think that's important is
understanding kind of theapplication of it.
That the idea that machineswere creating captions people
had seen it fail enough that ifwe said we were using speech
(31:20):
recognition to create yourcaptions, they'd say then we're
not working with you.
And so we actually hit it insome ways, we would find ways to
show that we were moreinnovative, but and we had to
stress the fact that there was ahuman involved all the time,
throughout the file, everything.
So that's evolved.
It's like the opposite.
Now it's like if you're not AI,we don't want to work with you,
(31:42):
right?
So it's certainly flipped andwe're becoming much more forward
with how we use AI in ourbusiness or talk about AI in our
business.
But it is part of everything wedo and we view it as a tool.
So our whole thing is AI isincredibly powerful.
You have to be very thoughtfulin how you implement it and how
you apply it, and if you applyit well, you can do some pretty
(32:06):
cool stuff with it and you cangain some very real efficiencies
.
But the beauty of AI is alsothe threat of AI, right.
So it can do unbelievablethings.
It's very, very powerful.
It does it very, very fast.
And if you look at some ofthese chat agents, And if you
(32:28):
look at some of these chatagents, the challenge is they're
so confident in their response,right, .
Oh of the response is like thisis the answer, oh great, well
then I'm done.
accurate unless someone checksit.
So, for the customers that wework with, so for the customers
that we work with, where the toor a brand to adhere to, or
(32:48):
certainly a direct studentaccommodation that you know
someone who truly needs it isthe one person experiencing it,
and it better be right Then youknow we need to make sure that
we're .
doing this right, and so that'swhy I say the use case and the
application matters so much.
So in our world, we view humansas critical to the process.
(33:09):
So the challenge is how do youbuild tools to insert the human
in the most efficient way?
And so we're constantlyevolving our tools.
As we think about whether it'sthe same process or new service,
we're constantly looking at howdo we continue to make the
human more efficient to do areally good job while using the
best of AI, and so we'veactually stayed AI engine
(33:30):
agnostic from the.
beginning AndWe've basically said there are
enough dollars being spent byvery, very large tech companies
that we can't compete withdirectly, but we can benefit
because the price has become solow that the real secret sauce
is how do you source the rightpeople and train the right
people and monitor the rightpeople and then put them on top
(33:52):
of the AI in a really effectiveway.
That's the hard part andcertainly the most expensive
part.
So if we can get the AI and ifthat keeps evolving, we can
basically swap in and outdifferent models and make sure
that we're using the best AI allthe time.
Kelvin Crosby (34:08):
I mean this is fascinating because I mean, I
think, when I use AI for myvideos and everything I do and I
humanly edit it, but what'sinteresting, as we grow and
using resources, and I thinkthis is something that as
entrepreneurs, as individuals,like in startups, it's like use
(34:31):
it, don't be afraid to go, useanother product to empower your
product.
I still like, for example, likein my, my company, See Me Cane.
My tube for the See Me Cane wasfor a completely different
product and then I was like canI get this tube this long and
(34:53):
this way?
And I tested it, boom, perfect,it works like a blind cane.
And I'm like, all right, let'sbuild out the electronics.
And then the rest is history.
And I think that's somethingthat's really crucial as you are
evolving through this processis using other resources to
(35:13):
lower your bottom line, and Ithink that that's something
really powerful.
Josh Miller (35:18):
Yeah, it's a really good point, I think you know,
recognizing where your strengthsreally are and kind of what
it's going to take to compete.
And we've seen several companiesin different spaces, both in
the captioning, accessibilityand in localization, already
full or or kind of move on fromfrom this path because they
(35:39):
tried to build their own engineand it was just so expensive.
It's very hard, and thensomeone surpasses them very
quickly and they have to figureout who are we.
Chris Maher (35:51):
Yeah, your model, and I said where AI and humans
to create your own solutions.
But now you're taking that andtaking that same methodology and
framework by taking all thestuff that you're building with
(36:12):
your software and people andconnecting it to third parties
like the Googles and theMicrosofts and the open AIs that
have built, as you said, putbillions of dollars into these
engines, that the way you getthe best value out of them is
you figure out how to use themto your specific customer or use
cases.
And so now you guys, yourspecial sauce is that technical
(36:35):
and human integration within andacross your business to
external parties and that's noteasy to do, that's super hard.
And also being agnostic you cankind of cherry pick the best in
class, right?
Josh Miller (36:53):
Yeah, it's become really interesting in the last
couple of years with Gen AItaking off, and especially with
a good example is voicetechnology.
So text to speech and what itcan do, whether we're talking
audio description or dubbing inmultiple languages.
That you know, certainly in themedia space, where they care
deeply about the quality of themix and the voice, people want
(37:15):
to know that you're using thebest voicing, and we have to be
able to say we.
And so the service we're nowproviding in a funny way is
actually being your curator ofAI technology, and that's a very
real thing and so you're wellaware that we do this state of
ASR report, which is state ofautomatic speech recognition.
(37:37):
We've done that for severalyears now and we're literally
benchmarking all the enginesagainst each other, because it's
both beneficial for us, butit's actually really interesting
data to share with the marketto see into a level set that
certain engines perform betterin certain cases.
Kelvin Crosby (37:52):
Right.
Josh Miller (37:53):
The content, and it all depends on how they've been
trained, and so that issomething that we intend to
expand.
Now that we're getting more andmore into localization.
We're seeing some fascinatingthings with text-to-speech in
these voices.
Some do really well in certainlanguages, some do really well
with emotion, so we have to takeall that into account and let
(38:14):
our customers really get valueout of the fact that we're
paying attention to that forthem.
Chris Maher (38:19):
Yeah, and the byproduct of you doing that sort
of research right is, one, itreinforces your thought
leadership in the marketplace,which then leads to credibility
for you, right, with themarketplace, but it's also
driving your product roadmap andmaking your products better.
So it's super smart.
Josh Miller (38:36):
Exaxtly.
So that was one of the thingsthat we figured out early on
with the state of ASR report isprobably now five or six years,
at least, that we've been doingit.
We were doing it anyways.
We were benchmarking the enginesanyways because it was helping
us create more efficiency in ourprocess, and then we realized
people want to read this.
People actually find thisreally interesting and it's one
(38:57):
of our most sought after piecesof content every year, so that's
been very, very helpful for us.
Chris Maher (39:05):
Maybe, as we start to wrap up here, let's touch on
with what you can share aroundhow the business has grown over
the years, and that's throughhaving to raise capital, making
some important decisions aboutwhether it's targeting certain
markets or go-to-marketstrategies.
You've acquired some companies.
You got into a relationshipwith selling a portion of the
(39:26):
company to private equity.
What sort of insights can youshare about that process and
things that you would want totell the younger you or an early
stage entrepreneur that's justembarking on that journey?
Josh Miller (39:38):
Yeah, it's a good question.
I mean, so we grew organicallyfor a long time.
We started the business in 2008.
We did an angel round in 2010,small one and then grew
organically and we wereprofitable early on, which we
cared a lot about and our earlyinvestors felt like that was
very important.
And lo and behold, it matters.
(40:00):
So, that was great.
We had a certain level ofstability for a very, very long
time.
Chris Maher (40:07):
Can I ask a quick question about that decision to
focus on profitability versusgrowing top line revenue?
Did you have investors and aboard that were supporting you
on that, or did you have tofight them on that?
Josh Miller (40:22):
So it's a great question.
I think, early on, part of thechallenge we faced was being in
an emerging market.
And being especially in onethat was still relatively niche
is the honest truth.
As much as I'd like to thinkthat the captioning market is
billions and billions of dollars.
It's not, and that was part ofthe challenge we had, is kind of
(40:44):
how fast can we actually grow?
And we had to be honest withourselves about that, both in
terms of our capabilities butalso the market.
And that is something that isso important and as we think
about our future endeavors, themarket opportunity matters.
So that was part of it.
Part of it was driven by thetiming and the market size and
(41:04):
being real with ourselves aboutif we were to invest in more
revenue growth, could weactually achieve it?
Was it realistic?
So we were trying to balancethat and be really thoughtful
about that.
We never played the game ofhire 20 sales reps and see who
survives and like all that kindof stuff.
We just didn't do that.
It wasn't really in our DNAeither.
(41:25):
So we we kind of grew with themarket and, as I said, you know
we're trying to educate themarket at the same time.
So it was it was a prettymeasured approach.
We had a couple of keyinvestors who cared deeply that
we be profitable.
I mean, their experience wasactually with much larger
companies where profitability isvery, very important.
So that was in their DNA.
So it kind of like all cametogether and that's how it
(41:48):
worked and I think it served uswell.
I think understanding the marketyou're in and the type of
business you have and whether itis capable of scaling quickly
is so important.
So I don't know that we wouldhave done things that
dramatically differently.
We certainly saw competitorscome into the market, in some
(42:08):
cases, come and go in thattimeframe, but you know, I think
we did a good job ofmaintaining who we were in the
market and certainly theperception of us in the market,
and in making sure that westayed true to ourselves and
continuing to deliver a reallygood product.
I mean, I think one of thethings about the accessibility
market which I hope peopleappreciate, and even our
(42:31):
competitors, is doing a poor jobit just feels really messed up,
honestly.
Like you gotta care, and if youcan't get yourself to care,
then you're in the wrongindustry.
And so that was something thatwe felt and cared about and you
know, lucky for us, we're ableto hire a really great team
(42:53):
around us that also felt thatway.
So if you look at our teamtoday even, we have a really
passionate team.
People really care about makinga difference and doing this the
right way.
So I do think that's somethingthat we've kept true to
ourselves even throughinvestment and acquisition, and
I think that's a hard thing todo with acquisition certainly,
(43:16):
so kind of skipping a few stepsthere.
We partnered with a privateequity firm in 2019.
And there were a few reasonsfor that.
I mean, part of it was thismarket was evolving, this market
was actually growing andproving to be a real market and
we needed to go a little faster.
We needed to have a little bitmore expertise on our board and
(43:39):
we had a few you know dynamicsof you know our angel investors
had been with us for nine years,so we needed to create some,
you know, an off ramp for them.
And so there was a whole bunchof things happening at the same
time and it made it for to bethe right decision to to bring
on a new financial sponsor andthen with that that opens the
doors into a whole bunch ofother things.
So we had the opportunity toacquire a couple of companies.
(44:01):
It was kind of two at the sametime in CaptionMax and National
Captioning Canada and that wasat a time when we were feeling
really strongly that we wantedto expand into this live
captioning use case and saw thatas a market expansion
opportunity.
I will say, tying back to AI,the timing of it was interesting
in that several months later,ChatGPT came out and all of a
(44:26):
sudden, and even in speechtechnology in our measurements,
we saw a step function change inwhat was possible and we kind
of looked at ourselves and saidoh wow, okay, this is
interesting timing.
And then, on top of it, theeconomy collapsed and there was
a media strike and the resourcesthat we acquired were very
(44:49):
heavy in the media industry.
So there was a lot happenedthat we learned a lot from, and
so we had some, we've beenthrough some bumps and I think
you learn a lot from that.
It's certainly blood pressurerises at times in ways that are
probably not the healthiest, butI mean, I will say we came out
the other end.
(45:09):
But you know, it certainlystarts to inform like where,
where are we going.
Right?
And really try to understandthe market and what is AI now
and how are people viewing AI?
Cause it's one thing to havethe best product, it's another
to make sure your customersunderstand it's the best product
and it's the best product forthem and for their use case.
So, that is definitelysomething that's happened with
(45:30):
AI, and the evolution of AI isthat people's perception have
changed and that they want it towork.
They want it to solve all theirproblems.
Whether it can, you know, it'sanother story sometimes, but
it's gotten a lot better and wehave to be honest with ourselves
about that and we have torecognize that and just be
pragmatic in the way we run ourbusiness and where we invest.
Chris Maher (45:52):
Yeah, that was awesome.
We could talk to you for hourshere.
We're going to have to bringyou back and have like a part
two of this.
Kelvin Crosby (46:07):
So hours Josh here we're gonna have to bring
you back and have like a part
So, for better for worse, wetry to be pretty open and
available, and so we have apretty solid web presence.
We put a lot of content on ourwebsite and on our social
channels with, like I said,trying to educate first and and
be part of the community.
You go to a number of events tobe part of the community.
I think that's really important, so a lot of information about
3Play is out there.
(46:28):
I'm on LinkedIn, I'm around.
I love having the conversationsabout other startup companies
getting going in this space.
I have made some greatconnections with other founders
that are doing some reallyinteresting things in this space
even even, in some cases, veryclose to what we're doing, which
is in some ways excitingbecause it creates partnership
opportunities.
But I think there's so much todo .
(46:50):
here And and again, if we cancontinue that conversation of
how we get from compliance tovalue with technology, I mean
it's huge and it's definitely asituation where rising tides
raises all boats, because peopleneed to understand this before
they're going to buy it.
So very, very big proponent ofthat so we're out there and we
(47:12):
we can be found.
Chris Maher (47:14):
.
We'll put a bunch of links inthe show notes for all of our
listeners and, josh Josh thankyou so much for taking time out
of your busy schedule.
Congrats to you and your teamand the company that you've
built, and we wish you all thesuccess moving forward, and
we're looking forward to havingyou back on the pod so we can do
a part two of this conversation.
Josh Miller (47:33):
But thanks for joining us.
We have a whole future to talkthrough, what's going to happen
next, so I appreciate you guyshaving me and I think what you
guys are doing and focusing onthis emerging world and the
accessibility space is soimportant.
So thanks for doing it.
Kelvin Crosby (47:47):
This wraps up Investing in Accessibility.
Go live beyond your challengesand we'll see you in two weeks.
Thank you for listening toInvesting in Accessibility, a
Samaritan Partners podcast wherewe invest in change for
(48:08):
accessibility, not wait forchange.
If you want to follow us, youcan find us on YouTube or
LinkedIn at @Samaritan Partners.
If you would like to invest inSamaritan Partners, email Chris
at chris@samaritanpartners.
com.
If you'd like to learn moreabout us, go to www.
samaritanpartners.
com.
You can take the first step ininvesting in change by giving us
(48:32):
five stars and sharing thispodcast with everybody that you
know, so we can spread the word,so that we can give access to
all by Investing inAccessibility.
Popular Podcasts
CrimeLess: Hillbilly Heist
It’s 1996 in rural North Carolina, and an oddball crew makes history when they pull off America’s third largest cash heist. But it’s all downhill from there. Join host Johnny Knoxville as he unspools a wild and woolly tale about a group of regular ‘ol folks who risked it all for a chance at a better life. CrimeLess: Hillbilly Heist answers the question: what would you do with 17.3 million dollars? The answer includes diamond rings, mansions, velvet Elvis paintings, plus a run for the border, murder-for-hire-plots, and FBI busts.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.