If you’re a fan of sci-fi, you’ll expect robots will one day rule the world.
Thanks to tech startup Mobot, we’re a little closer to that day. The young
company is changing QA for mobile apps by using robots to test apps in ways that
imitate the way that humans actually use those apps.
Mobot founder and CEO Eden Full Goh joined startup coach Roland Siebelink on the
latest edition of the Silicon Valley Momentum Podcast. Eden shared the
inspiration for Mobot, the secret sauce that makes the company special, and
lessons she has learned since founding the company in 2018.
- Reflecting on Eden’s choice to drop out of college to accept a Thiel Fellowship.
- Why Mobot’s pricing resembles a cell phone plan.
- How Mobot is learning about the dynamics of different teams working together.
- Building a company culture at Mobot in Eden’s image.
- Why you shouldn’t be disqualified from doing something just because you haven’t done it before.
- The lens through which founders should take advice.
Roland Siebelink:Hello and welcome to the Silicon Valley Momentum Podcast.
My name is Roland Siebelink, and I'm a coach and advisor to tons of startups and
scaling companies. And I'm very happy today because I have an awesome founder
with me once again. Her name is Eden Full Goh and she is the founder and CEO of
Mobot. Hello, Eden. Thank you for joining us.
Eden Full Goh:Thanks for having me, Roland.
Roland Siebelink:Absolutely. It's such an honor. Eden, we always start
right away with tell us what you do and what difference do you make in the world
for which target group?
Eden Full Goh:Mobot is a mobile app testing service that combines the best
of humans with automation. And what makes us different is we will actually
execute physical device testing for mobile apps on actual iOS and Android phones
and tablets. And the reason that we're able to do this is we actually use
mechanical robots to do the work. I was inspired to start the company because I
used to be a product manager and I had to do a ton of manual QA. I couldn't find
a good solution for this out there, so I had to start Mobot.
Roland Siebelink:You already mentioned your background as a product
manager, can you expand a little bit on that? What was that company that had you
do all this awful manual QA work?
Eden Full Goh:I started off my career at Palantir Technologies, and I
helped them build a number of different web and mobile applications across the
government healthcare and energy sectors. That was my first exposure
specifically on web apps where I realized it's very easy in this day and age to
make a new web app, deploy it, create a CI integration, and have tests that are
running in the cloud. And you can know right away, did the code you create break
something or is it ready to go? And you can push it out. And this can happen in
a matter of minutes, if not hours sometimes.
But on the mobile side, I realized that it doesn't work the same. As I moved
into the next company that I worked at - it was a company called Butterfly
Network. They recently went public. They build a portable ultrasound device that
allows you to plug an ultrasound probe into an iOS or Android phone.
And it was here where I worked very closely with other engineers on this mobile
app that I realized it's very different building mobile apps than it is building
a web app. It's a lot slower. You are subject to the approvals of the Apple app
store and the Google Play store. You also do have to go through and build all
these custom tests using Apple and Google's testing frameworks, XCUI and
Espresso, and fundamentally, you have to play by their rules.
Fundamentally, what we do at Mobot is we believe that because you're using a
mobile app in a physical way on a tactile device, you actually should test
exactly like a human would use the app. And that means having a finger, a
mechanical finger up on a screen, and it will actually be faster to set up a
test that way.
Roland Siebelink:Was the robotics part of this startup new to you? Or did
you have experience with that as well?
Eden Full Goh:I studied a little bit of mechanical engineering in college.
And prior to my career as a product manager, I did start my own nonprofit where
I actually tinkered a lot with hardware. I developed a solar panel tracker
system that would passively rotate a solar panel to follow the sun. And I spent
quite a bit of time working on that and growing a nonprofit around that after I
received a Thiel Fellowship to drop out of college.
In my previous background, I tinkered a lot with hardware. I really loved the
interaction and the intersection between hardware and software. Very naturally,
looking back, I'm not surprised that the startup I ended up building combined
both of those interests.
Roland Siebelink:Ever regretted getting out of college?
Eden Full Goh:It's interesting. I think there are definitely days where I'm
like, "I wish if I had taken this class or that class, maybe I would know the
answer to certain questions." But I also feel in 2021, you can Google or Stack
Overflow your way out of most challenges on the technical side. I think I've
also realized that there are specialists and engineers on our team that can
solve a lot of these problems way better than I ever could. Hire people smarter
than you, and that helps solve a lot of these problems.
But definitely in the early days - I remember a few years ago, whether or not to
dropout of college was a much more controversial topic. And I think the answer
there is, if you have something better to be doing than college, you should
probably do the thing. But if you are still figuring out what you want, which a
lot of young people are, you should probably stay in college because that is the
best way to expose yourself to other people, internships, classes, professors
that can provide you with that opportunity.
Roland Siebelink:Going back a little bit to Mobot, you explained why you
guys do what you do. What's the reason for being? As you build a startup like
this, what do you want to be world-class at? What's the very core of Mobot's
competencies that you're building up versus what's the stuff where you say,
that's more adjacent, that's something we would outsource?
Eden Full Goh:I think I've realized, in QA, there's this false aspiration
people have where, "Oh, we want to automate everything." I think I've realized
in the last couple of years building Mobot that that is a false ideal or it's
not realistic. What we really need to aspire to do is efficiently leverage human
insight and creativity when it makes sense to, and then also efficiently
leverage technology and automation where we can.
What makes Mobot work and our secret sauce is the orchestration and strategic
leveraging of people supervising robots and then having the robots do the actual
repetitive work. But then when the robot stops and is confused by something,
that's when we have a human jump in, intervene, update a test plan, give
feedback to the robot, re-train the robot, and then keep going. That
orchestration of humans plus robots and different test steps - you might have a
robot run a test continuously for several minutes or hours, and then only have a
human interspersed in there strategically. Our goal is how can we reduce that
intervention more and more over time?
Roland Siebelink:How are you imagining Mobot making money over time? What's
the underlying business model that you see? Where will you be unbeatable and a
service that your customers can not live without?
Eden Full Goh:Ultimately, we are providing a service, right? The words
"technology-enabled service" have often a connotation where it's like, "Oh, some
outsourced human somewhere." But I think Mobot is taking that definition and
flipping it on its head. We have robots doing the majority of the work. But then
you have humans that provide that insight and that feedback. And I think that's
really how you build a business that will be scalable as a service.
We're not quite traditional SaaS. But how we ultimately price ourselves is how
much of our service are you needing? A lot of the way that we measure volumes of
work at Mobot is we measure it in terms of the number of actions or steps that
the robot has to take. A lot of how we price is based like a cell phone plan
where you elect a tier of you think you will be running this many devices, this
many test cases, that's going to be roughly this amount of work. And that
becomes a monthly subscription.
And the reason we're able to price this way is because engineering teams do have
a cadence. A lot of folks are following agile or scrum methodologies. They are
releasing every one to two weeks. They have a schedule where Wednesday we cut a
build or Friday we submit a bill to test flight or the app store. And so there
is a way for a Mobot to then estimate their usage and work with the customer to
project what their QA strategy will be.
Roland Siebelink:Yeah. That engineering team is also, it sounds like, who
you define to be your core customer? Or is there even a more narrow profile that
you have in mind as you take this to market?
Eden Full Goh:I think we want to work with engineering teams that
prioritize mobile as a product. I think there are certainly companies out there
where the web product is the most important and the mobile app is just this side
thing that gets made and neglected. But what we've really seen work well is
there are certain companies and products out there where mobile first, right?
And really, Mobot is designed for mobile-first companies because no one else is
serving this population of engineering teams.
I think there are over 2 million apps in the app store. We're not going after
all 2 million apps, right? I think there's a certain kind of engineering team
that you're not a solo indie developer by yourself. You probably wouldn't be
able to afford Mobot. What we really want to target are engineering teams that
maybe they have 20, 25 engineers. They have a release cadence. They're following
a schedule. They're looking to invest in their dev ops tools. They care about
being efficient and having a good product.
I would say that's probably only a smaller portion of the 2 million apps that
are in the app store. It's not so much an artificial constraint but just being
realistic about - yeah, there are 2 million apps, but not all of them are good.
We only want to go after the ones that are good.
Roland Siebelink:That brings us to your team. I think you mentioned you had
slightly below 20 people in the team at this stage, is that right?
Eden Full Goh:Yeah. We're 17 right now.
Roland Siebelink:Seventeen. Okay. That is awesome. Sounds about like the
age, right? Age of 17 and the middle of adolescence and a growing team like
that. How does it feel at the moment? What are some of the biggest things you're
working on as a team to make the team work better?
Eden Full Goh:I think we've realized that every function - we have a sales
team, we have an operations team, and we have our engineering team. Every team,
at different times during the year, will go through periods of optimization, or
sometimes you have to rebuild things and pay down tech debt. Things like that or
operational debt process.
Every team is on their own path, right? Right now, we're about to roll out new
robots that have more automation in place. And so, the engineering team is very
busy fleshing that out. And the operations team is in a position where they've
built out enough processes in terms of knowledge, transfer, training that people
who are going to oversee the robots, that they're ready to receive the
technology and start to operationalize it.
And so, once the robots are deployed, the engineering team will be in
maintenance mode to receive the feedback from operations. But then the
operations team needs to learn how to actually use this new technology and roll
that out so that we can deploy the service to the customers. And so, there's a
cross-functional dynamic that has to happen as we really learn how the teams
Every team is going through their own construction and deconstruction at
different points in the company. But I think we're at a stage right now where
we've seen that cycle play out a couple of times on different teams that it's
exciting to know that we're not reinventing the wheel the way we were a year or
two years ago. The company, when we were four people, and the company when we
were 10 people, or even pre and post pandemic, those are like very distinct
stages of the company.
Roland Siebelink:One of the CEOs I worked with was complaining to me one
day and he said, "You know, Roland, I never knew that being a CEO is like
herding cats." How do you feel about that statement? If you recognize some of it
in it, what would you say back to him?
Eden Full Goh:Yeah, I also didn't realize how much being a CEO is like
herding cats. Hopefully, the cats that you're herding are smarter than you and
they are also herding their own set of cats. I think I realized, yes, it can be
messy and unexpected in some ways.
Once we had customer requests coming in and we know this is our technology, this
is what needs to be done. Some of that naturally plays itself out. And if you
have the right people, they will propose improvements and changes and
opportunities, and they will be proactive and do a lot of that work for you as
well. But it's a lot of chaos. I will be honest. I completely agree with the
Sometimes, also myself, I don't always know what's best for the company. I have
definitely gotten called out a few times by my team where they felt like they
had to herd me. They had to direct my attention, my focus, and give me feedback.
I think it's an evolving journey where sometimes I know what's best for the
company because I have the original vision. I'm close to our customers. And
sometimes, someone else might know best. And I think that give and take is
really important when it comes to herding cats.
Roland Siebelink:Yeah, absolutely. That's a great answer. I love that. I
loved also that you started with "Well, hopefully the cats are smarter than you
are." That's a great way of thinking about it.
When you do find yourself in such a situation of uncertainty, when you're not
quite sure what to do, when maybe you get some flack from your team, I wanted to
talk a little bit about how you deal with that, but I also wanted to bring up
something that came up in the pre-conversation we had. Do you feel that as a
female founder, that is something harder to deal with than what you see from
fellow male founders or people in other positions? To the degree you can comment
on that, of course.
Eden Full Goh:Yeah. I've definitely gotten feedback before that. Sometimes
I'm more empathetic than I need to be. I can be very in touch with my emotions,
other's emotions, and I care very much about the culture of the company. I'm
very proud of that though, because I think we've created a team and a workplace
and a culture that people actually enjoy. We're all a little bit quirky and
nerdy at Mobot. And I think that culture comes from the top. I know that culture
can come from me. And it comes from the people that I hire who then also
perpetuate a good culture at the company and across the team.
I think that is important. But it is a double-edged sword where sometimes I
don't always make the best decisions because I feel so passionately about this
customer. They were early, so I'm willing to do good things, go above and beyond
for them. And then the team is like, "Eden, there's no need to do that. That's
not good. Or maybe that's not even scalable. We should think about that more
carefully before we promise things to customers." There's definitely times where
my loyalty to a customer or my fondness for a teammate or something can
definitely manifest itself in decisions. But I think being aware of that is
important and being upfront about that is important.
A lot of the way that we try to figure out how to work with each other at Mobot
is we actually, when a new person joins the team we actually have them create a
user manual - I think I borrowed this practice from a different operating
partner and investor that I was working with - where you write out here's what I
like, here's what I don't like. I don't like Slack messages or I do like Zoom
calls. Don't text me after 8:00 PM. Or I'm very emotional, so when you come to
me, try to come with data. Things like that. Quirks that we all have.
I share my quirks, my likes and dislikes. I read and learn the same from my
teammates. And I think that does really help us understand each other. I think
that's part of that process. I'm creating this culture where I'm open and honest
and vulnerable and that makes it so it's easier to work. But there's definitely
times where maybe I'm not as a cutthroat or as harsh as I could be. It's my
company. This is the way I'm going to run things. I don't know if it could be
different with a different CEO. But then, if it was a different CEO, it wouldn't
be the Mobot that I founded.
Roland Siebelink:That's right. Absolutely. And I think being able to
embrace yourself as you are and just being authentic about it is ultimately the
best form of leadership. More than trying to be someone you're not.
Eden Full Goh:I think we all have different working styles and different
personalities. This is why it's important to hire people of different
backgrounds. Not everyone on our team has a PhD in engineering or 40 years of
experience in business development. I think it's so important to, during the
interview phase, get to know people. Look at their side projects, their hobbies.
What other prior deliverables or outcomes or milestones were they a part of at
their prior job? And what does that tell you about that person and their
Just because on paper they don't have the perfect resume, that doesn't mean that
they can't do the job. And one of the things that I fundamentally believe at
Mobot - and this is actually something that our VP of engineering said one
time - no one should be disqualified from doing something just because they've never
done it before. I really try to embody that. We give people the opportunity.
And I feel the same way. Our early investors - I've never built a robotics
company before. I dropped out of college. What could I possibly know about being
a founder? But being given that opportunity and given the chance to iterate and
make mistakes and learn quickly - the trick is to learn quickly. You're not
expected to be perfect. But I think what we really want to do is give people the
opportunity. You never know what kind of unexpected insights can pop up.
Roland Siebelink:What is one thing that if you could just tell founders one
thing they should do or learn or focus on, what would it be?
Eden Full Goh:When you look at books and resources and podcasts and tips
out there, I think it's always important to level set with the person giving
that advice. What size of company are they running? Can that advice actually be
applicable to you? Because what I've realized works well for one founder in one
industry - or maybe they're a Series B startup and you're a pre-seed startup -
following the advice of that other company in that industry, or in that country,
even, it may not directly apply to you. And I think it's important to view every
piece of advice you get from anyone from that lens. Because I have often tried
to follow other people's advice too much without really being more thoughtful
about, "Hey, does this actually apply to me? Is this in alignment with my
An example is "hire fast, fire fast," right? Yes, you should take risks and
take a chance on people. But you also should not blindside them and you should
not be cruel to people. I think you have to look at advice from different lenses
in that way. And I think that's something that I've also learned is that you
can't just blindly follow that advice because you saw it in some startup book or
some podcast and someone else said it. It really has to apply in your particular
industry, your particular company, size and stage. Does your sales process work
like theirs? Do you have a traditional SaaS business or is your business more
services based? Is it B2B or B2C? All of those nuances are actually really
important. There's no one size fits all advice.
Roland Siebelink:That's absolutely right. And I wish more founders would
understand that and not all try to emulate. Google as a 50,000-people company
when they are just 10 or something like that. That's just one example. Very good
situational use of advice, right? Extremely important points.
Eden, this has been an awesome interview. How can listeners to this podcast help
you? What do you need most from the people out there and where should they go
Eden Full Goh:We at Mobot are really trying to learn as much as we can
about all the different kinds of mobile apps that are out there. Different
industries, use cases, how people actually leverage mobile products in their
day-to-day life. And as the industry evolves, we've realized it's not just open
your phone and use it. People are using it on the go. They're using it in their
car. They're using it with a wearable fitness tracker. And so, really what we're
trying to do is collect as many of these use cases and collect data on that as
we can. If you are someone who works at a company with a mobile app or a mobile
product, would love to connect with you, see if there's a way that Mobot could
be helpful to you. But also just a learning process for us as we try to
understand the mobile industry and really think about how we can innovate for
this space in a mobile-first way, which is a little bit different.
Too many companies out there are thinking about mobile as an afterthought. And
we really don't want to approach it that way. We want more data to be able to
Roland Siebelink:That's awesome. Okay. And so, anyone listening who knows
me and wants to get an intro to Eden, I'm happy to provide. Please just contact
us there and we'll also provide LinkedIn contacts and the website context in the
Thank you once again, Eden Full Goh, the founder and CEO of Mobot. This has been
an amazing interview and I really thank you for your time.
Eden Full Goh:Yeah. Thank you.
Roland Siebelink talks all things tech startup and bring you interviews with tech cofounders
across the world.