# The Rise and Fall of iRobot: Consumer Robotics Lessons

**Podcast:** How I Built This with Guy Raz
**Published:** 2026-04-13

## Transcript

Certainly the dream was, let's be honest, we were promised robots.
We we humans were promised robots.
Yes.
We were gonna build robots.
And everyone asked for a robot vacuum cleaner from the very first day.
Wait, when you would meet people, they'd say, oh, when is Rosie, like from the Jetsons, gonna be in my house?
When are you gonna clean my floor?
Welcome to How I Built This, a show about innovators, entrepreneurs, idealists, and the stories behind the movements they built.
I'm Guy Ross, and on the show today, how Colin Engel set out to make robots part of our daily lives and brought us one of the most iconic home helpers ever, the Roomba.
If you think about it, there are very few products that manage to cross over from being purely functional into something that feels almost cultural.
And for the past 20 years or so, one of things that don't just solve a small, but actually take on a personality of their own.
Something that people don't just use, but talk about and share videos and form a kind of relationshipnotic robot that quietly makes its way across your living room floor and sweeps up dirt.
Now, what's interesting about the Roomba isn't just that it works, it's that it almost feels alive.
It bumps and turns and adapts to every corner of your house.
It's been parodied by Dave Chappelle in a famous Pepsiad.
It was a running gag on parks and recreation, and of course, it's inspired an endless stream of videos of cats happily riding around on top of it.
But here's the thing: the Roomba didn't come out of a consumer electronics company.
It came out of a robotics lab, a company that in its early years was building machines for NASA, for the US military, and for some of the most dangerous environments on Earth.
In fact, one of the company's most important creations was a robot that saved lives in Afghanistan and Iraq by helping soldiers identify and disarm landmines and roadside bombs.
When Colin Engel and his co-founders started iRobot in 1990, they didn't have a business model.
They didn't have capital.
They didn't even really know who their customer was going to be.
But what they did have was a belief that robots would one day live with us and could change the world.
And for years they chased that idea, taking on government contracts and building whatever they could just to survive.
Until one product, the Roomba, finally cracked things open.
And what followed was one of the most unlikely consumer success stories in the past 20 years.
A product that didn't just create a category, but in many ways invented the idea of consumer robotics itself.
At its peak, iRobot would sell tens of millions of these machines around the world and dominate nearly 70% of the global market in robot vacuums.
But over time, the success of the Roomba began to unravel with rising global competition.
And a company that once defined the future suddenly found itself struggling to survive.
But that part of the story, we'll get to it a bit later.
Because the roots of iRobots start much earlier with Colin.
He grew up in upstate New York.
At a young age, he was already building and fixing things.
My mom would say, bring your milk in from the living room, and I would say, okay.
And then I would embark on a two-day-long project to create a gantry cream to automatically lift that glass of milk and um parry it on cables into the kitchen and really shut down the house for a few days.
So you were really more of a the kind of kid who would take things apart, maybe like, or go to Radio Shack and solder things together.
So yes.
But if you asked me, are you a jock or a nerd?
I would say, Well, I'm a smart jock.
Because I love sport.
And I played um soccer, I was a wrestler, I played hockey.
During the summers, I did canoeing, and so that I was a whitewater uh canoeer and and wilderness guide.
So the physicality of my existence was balanced by my love of building and creating.
And uh to me it was a wonderful and healthy balance.
And so uh it's not surprising to me that you, when you decided to go to college, you decided to major in electrical engineering.
You went to MIT in in 80, 1985, and uh, and that was your passion, electrical.
I mean, this is I imagine this is what you wanted to do with your life.
I you know, I went to MIT to major in whatever would let me build the coolest stuff.
You know, one of the most amazing things, serendipitous things that happened to me in my career is that I was walking back from class one day between my junior and senior year, and the guy I was with was in my fraternity, Joel Cayley was his name, and and um, you know, I was making conversation.
Hey, what are where are you going?
He said, you know, I'm gonna go apply for a summer job at a robot lab as part of this Europe program.
And and uh, you know, I asked Joel if I could come with him and maybe apply for a job too.
And you know, you kind of give me a dirty look like I'm going to this to apply for the job.
But, you know, he couldn't really say no, so he said, okay.
And so I followed him to this lab.
The application for the lab uh was actually really interesting.
They gave us all a blank piece of paper, and they said, Well, what have you built?
Just write it on the paper, and when you're done, hand it in.
So I started writing, you know, a G I built a gun out of Lego that could fire a piece of Lego track through plywood.
You know, I built a canoe, an hour went by, and I was still writing things on this paper that I had built.
And I had this epiphany saying, this might be the place for me, because gee, I'm at MIT and everyone is already done writing down the stuff that they've built, and I'm I'm not done yet.
And uh, you know, I handed my paper and I got this job.
And and this lab was run by a guy named Rodney Brooks, who's many people who know anything, you know, something about robotics knows he's sort of a considered to be a pioneer in this field.
And the job was to be like a researcher at this lab, or to was it a paid job, or was it so it was it was paid, and you know, they had projects.
For example, this professor Rod Brooks said, okay, Colin, we've got this robot Seymour.
It has a 12-inch wide base, could drive around the floor in the AI lab.
And he said, I want you to enable it to open doors.
That was a pretty tall ask for an undergraduate at MIT at the time.
No one, you know, there was barely any robot arms around, much less arms that could reach out and grab handles and open them.
But I said, okay, I'll I'll I'll work on this project.
And, you know, I I did the obvious thing.
Uh I built a candy machine.
So I decided that the right way to open doors wasn't to build a robot arm because that was hard.
The right way to do it was to put a candy machine on this robot and drive up to the door and offer candy if anyone would open the door for it.
Part of me was being a wise ass, but you know, the the this idea that robotics isn't a thing, it's a toolkit was uh an early lesson that I was taught because when you say robot to most people, they think humanoid.
But given the journey I was on, robotics was a box of parts that allowed you to make smart machines that were capable of doing interesting things and to solve a problem.
And I know that uh at some point, at least you were involved in helping Rod build something called Genghis.
And and anyone who's been to the Smithsonian, I think it's on display there.
It's it was a it's an insect like robot uh that can walk on six legs.
Uh and this was kind of a a revolutionary robot that was created at the time.
Yes.
So Rod uh was interested in planetary exploration.
And this little silly robot was able to successfully climb over very rugged terrain more successfully than robots that were using supercomputers.
And the AI processor on this robot was an eight-bit microprocessor with 256 bytes of RAM.
Not kilobytes, bites of RAM.
Wow.
So we decided we like working together.
And uh so I was I was about a year and a half into the master's thesis.
Rod said, he said, hey Colin, I I got something to ask you.
I said, okay.
And it's like, well, I'm thinking about starting a robot company based on behavior control.
And I said, okay, I'm in, I'll run it.
That was that was easy.
That was, I mean, yeah, it was it.
I didn't he wanted you to run it or or you volunteered to run it, or kind of both.
He was telling about it because I think he wanted to be involved.
But he didn't probably have the time, right?
Because he's also teaching classes and re and and publishing.
I mean, he he couldn't run a business as the CEO, presumably.
Right.
Yeah.
All right.
So so this company that he he wants to start would would of course become iRobot.
Uh, and it was gonna be him and you and and another student from MIT, Helen Graner.
Uh, and this is uh 1990.
So you you're like in your early 20s.
Uh and and so what was your sense of of what this robotics company would be?
There was no core application.
There was a need at the time uh where people would say, Hey, can you build us a robot like Genghis?
Uh that we could go and program?
Because it takes a long time to build a robot.
Who would ask for that?
So Rod had a pretty good network.
You know, it would be like Boeing or Mitsubishi.
So the there was no business model in the sense that you didn't go out and raise money, right?
I mean, it was sort of like, hey, uh I'm I'm connected.
Rod was like, I'm connected, and let's let's just do the research, let's have our research funded by contracts from the government or from corporations who may want, you know, a copy of this robot to tinker with, and maybe they'll hire us to do some contract work for them, and that can finance what we're actually trying to build.
Is that is that fair to say?
Yes.
We did not take venture capital until year eight.
Right.
So in the beginning, we had designed these robots and Mitsubishi would say, I want to buy uh a legged robot, and we would say, Okay.
But what would they want to do with it?
Just to see it, just to have it?
They wanted to do AI research, but had no robot to do research on.
I see.
And we had a line of a wheeled robot, a tracked robot, and a legged robot that we would sell for between five and ten thousand dollars.
We would get paid half of the money up front.
We would use that money to go out and buy the parts.
Uh I had invented a way of building uh the robots.
They were kind of origami robots.
Um they were made out of bent sheet metal.
Okay.
Uh it worked, it was a really quite an ingenious early way of of inexpensively building a machine.
But we really lived hand to mouth for quite some time.
I think it was six and a half years before I started a month with enough money in the bank to make payroll.
Wow.
And it it didn't seem like you were you guys were starting a consumer products company.
I mean, it seems like you the the business was going to be for military applications, maybe space, military, maybe some industrial, but not this you weren't in any way thinking about consumer products or customers in that way.
You were probably thinking about building a business that was going to work with governments.
We didn't know.
Uh certainly the dream was let's be honest, we were promised robots.
We we humans were promised robots.
We humans were promised robots, right?
And and so that, you know, it was a little bit of a if not us who, if not now, when are we going to actually go and start to build these?
So we didn't know how to build consumer robots.
Everyone asked for a robot vacuum cleaner from the very first day.
In fact, if I introduced myself, hello, my name is Colin Angle.
The answer would be not good to meet you, Colin.
It would be when are you going to clean my floor?
Wait, sorry.
When you would when you would meet people, they would ask you what you would you did for a living.
You would say I build robots.
I build robots, and they'd say, Oh, when is Rosie, like from the Jetsons, gonna be in my house?
When are you gonna clean my floor, right?
Because they thought of Rosie from the Jetsons and that's what they wanted.
And you would hear this from the beginning.
From the very, very beginning.
But we just had no idea how to do it.
In fact, I would start to get a little snarky and say, Well, how much are you willing to pay?
Yeah.
Are you willing to pay five thousand dollars if you are?
Well, maybe I can do something for you.
And it would say, Well, no, no, I would I can't spend five thousand dollars on a robot vacuum.
If I could do that, I would hire someone to clean my floor.
But I guess I'm wondering is when you would hear that.
Yep.
It sounds like it would be a little bit annoying because actually a lot of what you were doing was for research for government, you know, or military or space purposes, right?
Which which you could argue from your perspective, you could think, well, we're actually doing something much more uh consequential.
Like we're building robots that could save lives or could explore space.
I mean, was that in the back of your mind?
You know, honestly, no.
Because the dream of where are the robots we are promised.
I wanted to build the robots for the rest of us.
And so that it was almost okay, I I'm building robots for the military, I'm building bo robots for the oil and gas industry because I need to pay the bills.
But one day we'll do we'll be in a better position.
And at the time iRobot was great at technology, but really didn't have much talent within the company to understand markets and understand pricing and understanding.
Yeah, you guys were a bunch of We're a bunch of nerds.
Science nerds.
You didn't know about marketing or distribution or we certainly didn't have it.
We didn't respect it even.
You weren't allowed to say marketing without saying weenie, you know, immediately afterwards.
I mean, it was it was a a very much an engineering culture.
Yeah.
So I mean, so given all that, how how are you actually going to become a real business?
Like how are you gonna come up with a a sustainable model?
The breakthrough for us, actually, was a business model that we discovered.
We would come up with a uh a way of working together where I would work at cost.
I would work at breakeven, but since I wasn't making any money, we would split the value that we ultimately created.
For this big company, this is nearly free.
They were getting MIT engineers working on a problem that they cared about at a third the price, and they only had to care about this value sharing thing if something good came out of it.
And they would own the IP they brought to the party.
We would own the IP that we brought to the party, and any IP that was jointly developed was jointly owned.
That was the deal.
That's pretty risky, but also since they were putting money into it, in my mind, they had to at least believe it was a good idea.
So Johnson Wax is a fine example.
This SC Johnson, right?
Is that what you're S.
C.
Johnson?
Huge massive company.
They own tons of consumer products that we use all the time.
Yeah, and and they've sort of divided into the consumer brand that we all know, and then the professional uh that develops cleaning chemicals for janitorial services.
Their problem was uh so they sell the chemicals that are used to clean floors.
Right.
And so the idea was if we created chemicals which were specially designed for robots to clean the floors, then Johnson Wax owns more of the solution and can create a differentiated value proposition, right?
Because maybe the chemicals are more expensive, but with the robot, the whole system saves the target or the shopping mall money, and this all improves their profit margins and and creates a good business.
And this was an autonomous robot you're talking about?
Absolutely.
So, okay, so I I'm obviously we're I think any everybody listening can start to connect the dots of where this is headed, but this is still 1995, 96, 97.
You're working on this thing.
Yep.
But uh I'm assuming they never commercialized it, it never went anywhere.
That's correct.
We weren't going to succeed as a business until something actually broke through and became right.
Because you were making no money.
We were paying ourselves.
We said I say making no money.
We were paying salaries, we were paying overhead, um, we were paying for parts, we were developing IP.
I mean, it was we were having a ball, no one left.
Everybody was just psyched to be in this lab tinkering.
We were on a mission.
Yeah.
And we were very successful at getting press and validation.
We got the cover of magazines, we're on the cover of National Geographic, we're in the cover of Scientific American, we were on the cover of popular science, like four times.
And even though you weren't making consumer products, it was because you were making products for I think a lot of the military applications, like well, remember, nobody was making robots for consumers.
That's true, right, right?
Right.
But you know, the uh company's uh mission statement was build cool stuff, deliver great product, have fun, make money, change the world.
It was powerful because we were gonna be the people that made the robot industry real.
Let me come back in just a moment.
iRobot launches two products that could not be more different: a roving bomb diffuser for the military and a babbling toy doll for kids.
Stay with us.
I'm Guy Roz, and you're listening to How I Built This.
Hey, welcome back to How I Built This.
I'm Guy Roz.
So it's the late 1990s, and as part of iRobots' mission to make cool stuff, Colin leans into consumer goods and starts a conversation with Hasbro.
And after a little back and forth, he agrees to a three-year deal to make robotic toys.
And did you know what kind of toy it was gonna be?
Well, they wanted to make a um baby doll.
Okay.
Was the first one.
And what would what do they want the baby doll to do?
Well, what did I wanted to do?
Because we were pitching the idea.
See, we wanted to make a baby doll that was actually alive, and we were gonna go and use artificial intelligence to make a toy that could detect how it was being played with and play along and and make it magical.
And so that if you rocked the baby, and the baby would say, Hey, I'm being rocked, I guess I should be sleepy, and I could go do go do a napping play pattern, or if you're feeding it, okay, I'm I should do a feeding play play pattern.
Um, it was really cool, is it came out in in 1999 and called My Real Baby, and we sold 150,000 of them, I think.
And they introduced us to low-cost manufacturing.
And so did we make any money out of it?
We got our costs covered for three and a half years.
But we learned how to make low-cost robots.
And you learned a hell of a lot about how to make consumer products.
Correct.
That's where we learned how to make consumer products.
It's interesting, uh uh, because I I I I looked at these My Real Babies and uh and they did well for a short period of time, but then they were discontinued.
And I I guess um the some of the reviews, and again, these are uh reviews from adult writers were that some people felt they were a little creepy, but uh, you know.
When we were doing development, uh at first my royal babies eyes look left and right as opposed to opening and closing.
Oh wow.
And that that was really creepy.
So that was that was like lesson one.
It's like, okay, you you can't do that.
And I remember there was a crisis right before we launched because the the baby when it was young, it would it would coo.
You know, and then as it got older, it would start putting words together.
And someone on the Hasbro side listened to its babbling and became convinced that the robot was saying, Baby baby, bite my butt.
And so we, you know, they're like, well, that's not we can't launch with this.
It was like it randomly babbling.
It's like, no, no, no.
It's saying baby, baby, bite my butt.
And so that we had to like have a crisis software uh uh um strategy to figure out how to provably keep my real baby from babbling in that particular pattern, which was crazy and ridiculous and wonderful and and just very real and and part of the journey of consumer products.
I wanna I want to just pivot for a sec and ask you about the business side, because we're now getting, you know, we we've covered a lot of what's happened in the nineties, and really, and I think it was only in 1998 where you took on some outside funding.
You got you finally raised some money, which is is but what I'm more interested in is you, because you were 22 when you became the CEO in 1990.
So by 2000, you're 32, right?
And you, and I mean this in the most uh ad uh with great admiration.
I mean, you are an engineer, you're a nerd, you're like focused on building things that matter.
You're not a guy who has any interest in you go to business school, you're not gonna go to do a business administration degree.
And so, how did you learn over those 10 years how to run a company?
Well, you had to make choices.
And my journey at iRobot started back when I was a wilderness guide, right?
I was taking two staff people and six campers out into the woods for six weeks and bringing them back alive with all sorts of challenges: food, canoes tipping over, you know, and and all sorts of drama.
And so that sitting in the CEO role was actually very comfortable to me.
Now, what wasn't comfortable was putting down the soldering iron.
My passion is building.
Yeah.
And along the way, I realized that building a company that could build cool stuff counts.
And uh, I think that one of the habits I had that allowed me to succeed was whatever I was doing on any given day.
I try to figure out how to give away all of it.
How can I go and take all of the responsibility I currently have and delegate as much as possible to people who are going to do a better job than I can do?
Yeah.
I mean, I I remember viscerally, I le I used to love 3D CAD.
You know, it was like one of my happy places.
Computer assisted design.
Yeah.
Uh making 3D designs of things, and and I hired this guy, and he was like a fresh out from uh college, and he sat down, and in the first day, he created something spectacular.
And I looked at it saying, wow, that's spectacular.
And my heart was saying, I should never do 3D CAD again.
Nice.
It was like, oh yeah.
Oh crap, okay, fine.
This is it's time to move from being the guy who loved bending the sheet metal to the guy who loved creating an environment where creativity could flourish, sprinkling a little bit of, well, that's not gonna work, or have you thought of this to indirectly guide decision-making processes?
How let's talk about um defense products for a moment because I, you know, if you are a defense contractor, your your client is a is the government, right?
And so, and that can be a big business.
That's why all these defense contractors are headquartered in Washington, DC, because they want a piece of the action.
You guys had contracts with DARPA, the defense advanced research project agency.
Yep.
You did uh something called the PACBOT, which was an all-terrained robot that was designed to like detect bombs and then dispose of them.
And and these were actually used in Afghanistan and Iraq by the U.S.
military.
Absolutely.
So if you think about like a pizza box with with tracks on either side, you're you're close.
Right.
Um, and then we started being able to put arms on them and so that they could be used by a a bomb disposal team when 9-11 happened and we sent troops into Afghanistan.
We saw on the news that they were sending the airborne troops into the tunnels and the caves in Afghanistan with you know a flashlight and a pistol and a rope tied around their waist so that they could pull the soldier out in case they were shot while they were down underground.
And we said, no, no, no, this is what our robot was for.
And so we started going down to Washington with our packbot robot that we had designed and said, You gotta send this to Afghanistan.
How many total do you remember?
How many um packbots did you eventually sell to the military?
I mean, it was thousands.
Thousands, wow.
And each one, well, I guess the price eventually went down to about 45, 50 grand, but so that I mean, we might have gotten up to maybe 50 million of revenue coming in from the defense side.
And probably, I'm sure some of these pack bots uh got blown up, right?
Absolutely.
I mean, I th I probably five hundred to a thousand of the robots were destroyed.
But that was okay.
Yeah, because you're saving a human life, of course.
We would get postcards from the soldiers saying, You have saved lives today.
And it was very moving.
And the uh there was one robot that we actually got back, where on the head of the robot, someone in pencil had sort of ticked off every time that robot took out an IED.
Wow.
And we got it back because finally the bomb got the better of the robot.
And the ID technician apparently carries this back to the depot, tears in his eyes.
Can you fix it?
Like it's a fallen comrade.
Or a dog, right?
Like yeah, because of course dogs are used.
Yeah.
It's amazing.
I I know I was familiar with these because I back in the day I was a war reporter and spent two and a half years in and out of Iraq and saw saw these in action.
And uh uh they were remarkable.
I mean, iRobot still was a relatively small company, and you had all of these concurrent projects.
You had, you know, you you've got the the these robots, right, detecting these pack bots for the military.
You've got you know, working on S.
C.
Johnson project, you're working on eventually with Hasbro.
But let's dive into Roomba.
I mean, you up until this point, okay, you made the these dolls for Hasbro, and and so there were some consumer products.
So you had learned about mass production and um and consumer uh marketing.
But what's the origin of Roomba?
How do you start talking about, hey, wait a minute, let's make this.
You know, I I I told the story about how people would come and say, Hey Colin, uh, when are you gonna clean my floors?
Yeah.
And so we knew this was a good idea.
And one of the early employees, a guy named Joe Jones, had actually built while he was at uh working at MIT, uh a little round robot that picked up some, it couldn't actually pick up dust, but it if you crumpled up little the right size little bits of ball, it could suck it up.
It could suck it up.
You know, and then 12 years pass.
And the toy contract was coming to an end.
And Joe came up to me and said, Hey, Colin, I think it's time to do the vacuum.
Can I can we work on it two weeks?
Give us 15 grand in parts, and I'll have a prototype.
So they went and they um they made it round, and it was round so that it would never drive into anywhere that it couldn't turn around in its own diameter and and and and escape.
And it dragged an electrostatic cloth, like a Swiffer cloth behind it.
It wasn't a vacuum.
Okay.
And it kind of worked.
But it didn't have no eyes.
Now it's different.
No camera for navigation, no.
No, that came much later.
Okay.
Gotcha.
Um, so we kept developing this idea.
And we knew we needed a better navigation system.
But at the time there was no navigation system good enough to give you an actual position.
Right, because if it was just randomly bumping, you couldn't guarantee that it was gonna clean the entire floor of the room.
Correct.
You know, no one's gonna die if you didn't.
Uh but it had to do it had to do a good enough job.
And then we had done research with customers around the product.
One of the early problems with Roomba was if it was only dragging an electrostatic cloth, it was not valued by the consumer.
Right.
It's just that's just a dress, it's just dragging a piece of cloth on your floor.
Right.
I can do that.
Right.
And so that they would say, Well, how much would you pay?
And it's like, well, we pay 40-50 bucks for that.
It's like the battery costs 40 bucks.
I mean, I mean, come on, yeah.
The um uh, you know, this isn't gonna work.
It's like, well, what if it was a vacuum?
It's like, oh, well, if it was a vacuum, then I'd pay hundreds of dollars for it.
And it's like, uh, okay, uh, time to make a vacuum.
And then we had to go invent the world's most efficient vacuum cleaner because we we couldn't afford a lot of battery power.
And so we um we said, well, what if we break this whole vacuuming problem down into two steps, where we have these counter-rotating brushes, just spinning brushes that pick up all the big debris, and then have what we called it a squeegee vac.
It's it it uh between two little rubber blades is a very narrow hole.
And because you're pulling the air through that narrow gap, it moved faster.
It's kind of like putting your thumb on the end of a of a hose to make it squirt faster.
And the velocity of the air actually is what picks up the dust.
And so that by breaking vacuuming into two pieces, one for doing large debris and one for doing uh small debris, we had invented a new way of vacuuming that was incredibly efficient.
And so that, you know, it's uh why didn't Roomba happen earlier?
Well, you just needed to have a small, desperate company that happened to be expert in mine hunting toys and industrial cleaning.
Cleaning, right to come along, and then it's then it's kind of obvious that what you should do.
And and uh, you know, the thing the some of our biggest challenges had to do with how do we sell this thing, because we had far less experience on on that front.
This is in 2002 when you finally had a product to launch.
Finally had a product to launch.
And we thought Sharper Image was gonna do it, um, and then they backed out, and I called up Brookstone.
Yeah, I was gonna say sharper image, no, then you call Brookstone.
Right.
And and the Brookstone story was amazing because the only person who answered the phone, because we dialed everybody, was this woman who was in charge of the hard-to-find tool catalog.
And it was her first day on the job.
So she hadn't yet been trained that she shouldn't talk to people like us.
And so she heard our pitch and said, okay, well, well, I guess you can come in.
And so we brought, you know, we we got in the car, we went down to New York City or whatever it was, and and we showed her the Roomba.
And she was like, Oh my God, this is this is amazing.
Let me let me sh let me show it to my boss.
And so she brought in her boss, and her boss said, Oh, my God, that's that looks like it actually works.
Um, let me let me show you to my boss.
And you could demo this.
You could say, Here you go.
Yeah, yeah.
We we we we had working demo.
Would you bring like Cheerios or something to like put on the ground?
So my secret was to, I I would like a vacuum cleaner salesman, I would have a little ziploc bag of Cheerios in my back pocket.
You literally had cheer, I just said Cheerios because I think of kids and Cheerios.
No, no, it's Cheerios was the best.
Because what I would do, uh you know, we'd be in it be in their conference room, and I'd they would say, Oh, that's neat.
Does it actually work?
And I said, Well, of course it works.
It's like, really prove it.
And then so I would take these Cheerios out of my back pocket, I would sprinkle them on their on their floor, and then I would in great with great drama, I would I would stomp and crush and try to grind the Cheerios into the rug, at which point I had their complete attention because I just messed up their conference room, and then put Roomba down.
Roombo is great at Cheerios, great.
The dust, the the big chunks, everything.
And it was, you know, Cheerios are bright, so on a dark carpet they show up really well.
And and it's like, oh my god, it it works.
And we went to like from the lowest to the highest level at at at uh Brookstone in about 40 minutes, and then the head of product or merchandising came in and we showed a market oh that's pretty cool but we couldn't possibly sell this for any price uh above 200 and we had planned on selling it for $199.
Amazing.
And so that we launched with Brookstone and about two weeks later I get this call from the woman managing her account Pam Camelstein I remember her name and she's like hey Colin, how's it going?
It's like, well how's it going with you?
And it's like well well pretty good.
You know we'd like to we'd like to order some more product and it's like really well well how much and she said well how much can you build wow how did that how did that happen?
I mean Brookstone of course we I remember we did an episode on the temper Pedic mattress and it started with a pillow and they launched that at Brookstone.
So you know Brookstone, but still Brookstone is not Target.
It's not Walmart.
How did it do so well?
Was it uh was it just because it was so weird and unusual it got press coverage?
It was two things.
First, Brookstone discovered that if they put it in front of the store and just threw down dirt and Cheerios and stuff and what and showed it working.
Yes.
It would draw traffic into the store.
In the mall.
In the mall.
Right in the shopping mall, yeah.
So it it demoed really well.
But combine that with the press, because we had no money to advertise.
We had up $5,000 a month for PR, I think was was our budget.
But that whole shtick where I go into the conference room and throw Cheerios on the on the on the ground.
No one believed this thing worked.
And then when they saw it working, then if you were press, you had a secret.
And nothing is better than taking a skeptic and flipping them into being a supporter.
And we had a hundred and fifty articles written about the Roomba immediately after launch.
Everything from the Wall Street Journal to USA Today to Newsweek to it was amazing.
We sold 70,000 Roombas in the first three months.
That's incredible.
So you guys were probably popping champagne bottles thinking, this is it.
We we we finally struck struck gold here.
Absolutely.
It was validation after twelve years of beating our heads against the wall.
Again, 12 years of wondering whether we would ever find that robot that everyone could get.
It's amazing.
I read that that by the end of that year, I mean, what a turnaround.
14 almost 15 million dollars in revenue.
Yes.
For iRobot, which is incredible given that you the year started out not good.
Um I read that on the strength of that amazing those three months in 2002, you used very well very understandably ordered a ton of robots to to m to be made for 2003 because you sell 70,000 in three months, you're gonna need hundreds of thousands for 2003.
Right.
So uh that euphoria certainly uh turned around fast.
I mean it it you know 70,000 in a quarter we should sell three hundred thousand in in two thousand and and three and um it almost didn't turn out that way in fact we had a commercial that we made which completely failed I describe it as a commercial made by engineers for engineers.
It was just a practical commercial showing what the thing did yeah and and it didn't resonate.
I haven't seen that commercial was it just like a literally like hey our floor is messy why I got this like was it it didn't show any people uh it just showed the robot driving around the ground picking up stuff going under things.
Oh my god you should have put a cat on it.
We did have you know one of our most epic slogans in that first commercial which was if it's down there, we'll get it.
Uh, but you know, that didn't work either.
The uh no no.
But uh, you know, so we ended in 2003, it was after Black Friday, after Cyber Monday, and we had sold a grand total of fifty thousand robots from that uh year, and we're sitting on two hundred and fifty thousand robots.
That's not good.
I mean, that's a lot of inventory and yeah, that would be existential crisis time.
Well, we come back in just a moment.
Roomba gets a marketing makeover from a totally unexpected source, a Pepsi ad starring Dave Chappelle.
Stay with us, I'm Guy Roz, and you're listening to How I Built This.
I'm Guy Roz.
So it's the fall of 2003, and after an early boom in sales, Roomba is starting to flatline, and iRobot is sitting on a lot of inventory.
But suddenly, and for a reason Colin never could have predicted, things turn around.
We used to uh bring the whole company together for for stand up for a couple minutes at the beginning of every day, and just sort of, okay, how's it going?
How's it going?
So here we are, it's like you know, Wednesday after Cyber Monday, and the guy running the the uh uh e-commerce store said, Hey, why did sales triple yesterday?
And nobody knew there was no new press, there was no new anything.
And then uh I think it was an intern who raised his hand and said, you know, it's funny.
I I saw a commercial, and sure enough, Pepsi had made a commercial with Dave Chappelle, the comedian, where you know, Dave walks into this beautiful house and he looks over, and there's a kitchen table with uh Pepsi's and bowls of snacks and chips and so forth, and he goes over and grabs his Pepsi and grabs a potato chip, and a Roomba comes out, and he looks down at the Roomba and he says, A vacuum.
It's like, God, you know, the whole world didn't know Roomba was a vacuum, but he knew it was a vacuum, and he throws the potato chip on the ground, and the Roomba goes and cleans up the potato chip.
And then the Roomba locks onto the Pepsi and starts chasing Dave to try to get the Pepsi.
And there's a cut screen, and now he's on his back, his pants are ripped off.
He stands up, a beautiful woman appears, and he says, Your vacuum cleaner ate my pants.
There was nothing I could do.
And it's nothing chases thirst like a Pepsi.
And we sold 250,000 Roombas in six weeks.
After that ad.
After that ad hit.
And Pepsi basically made an ad for you without even telling you.
With a celebrity endorsement, with you know, and it was funny.
And we kind of realized we knew nothing about marketing and started hiring people into the company that could help tell our story better.
And so it's a huge epiphany for me.
It was 2004.
Suddenly we we tripled our sales from 2003 and were starting to feel like we knew what we were doing, other than the fact that uh we had another crisis, which was the fact that we had designed Roomba to um uh last 150 hours, which was the European standard for upright vacuum cleaners.
To last before it needed to be recharged.
To last before it died.
Oh, oh, before it died, right?
Oh, but upright vacuum cleaners don't get used very often.
So 150 hours is a really long time for an upright vacuum cleaner, but it's a really short time for a robot that might be used half hour at a time, half hour at a time, an hour a day kind of thing.
And so that we were having a uh a huge crisis where people were having their roombas break very early after six months of use, and so that we had to make a decision as a company to uh just ship them a new robot.
You had to uh just a free replacement.
Free replacement.
That that must have cost you a fortune.
It did, but this was a moment to preserve the brand.
And you know, another wonderful truth is if a company stands behind its product, you generate more customer loyalty than if that customer never had a problem with their product at all.
By investing in the brand, iRobot became a trusted and respected and beloved brand.
And and presumably you were making, I mean you had learned about mass production while you had worked with Hasbro a few years earlier.
So I'm assuming you were you were producing these in China, right?
I mean that that was the only way to make them affordable.
Yes.
So they were they were manufactured in China for many years until much later in the game we moved it to Malay to Malaysia.
Malaysia, okay.
Yep.
So within two years, I think the iRobots doing a hundred million dollars, almost a hundred million dollars in revenue.
You still have the PacBot, which is still an important product, but in 2005, you guys go public, you become a publicly traded um company on the Nasdaq.
And help me understand.
When you filed for for an IPO in 2005, because in in 1990 it was three people.
Do you remember at that by that point how many employees iRobot had?
Probably 250 employees by that time.
But you know, I think it was 2008-2009, we plateaued um for a little while, and part of it was the economy, and part of it was the fact that we were running out of early adopters.
And it exposed a pretty significant fallacy in our technical direction, right?
Because if you ask me at the time, what Colin, what's the what's the perfect Roomba?
I would tell you, well, the perfect Roomba is the Roomba you never see, you never touch, and you come home and your house is clean.
Right.
I mean, it's that right, that sounds pretty compelling.
Unless you put that in the context of actually hiring someone to clean your home, right?
Because if you remove Roomba and you insert cleaning person, and you say, imagine a cleaning person coming into interview for a job, and and the cleaning person says, You can't talk to me.
You can't really give me any guidance.
You just have to trust me that I'm gonna do a great job.
Am I hired?
Well, that sounds a little creepy.
And the innovator and the early adopter are willing to believe that this technological solution is gonna be cool, but the early majority, which we needed to continue growth, absolutely did not.
And that's the customer you need to actually scale.
Right.
So that early majority is where companies go from being interesting to being valuable.
Okay, you start to plateau, yep, and you've got to convince people that this is something they need.
What do you do?
I mean, does it is that the technology is getting better so you can you as the consumer can actually have more agency over the machine?
We needed to change how we talked about the roomba.
We needed to change the technology, and that's where navigation became important.
We need systematic navigation, we need object recognition, we need to be able to go and clean around objects that are specified.
And then we need to communicate it to the customer in a way that allows them to feel control, right?
Because that was what that's what was missing.
The Roomba works for you, not the other way around.
The customer wanted confidence that they could control where the Roomba cleaned and the Roomba would be safe and what wasn't going to go and knock over the precious vase or what have you.
So that it was, you know, efficacy, control, and safety, and it opened up the marketplace to the next phase of growth when uh iRobot and robot vacuum cleaners as a category really started eating vacuum cleaner sales.
All right.
So I mean, I think eventually, by by I don't know, maybe 2016 or uh 2014, I think you'd already sold 10 million Roombases, but I have to imagine that in your mind at least, and certainly in the mind of others, if the company is really going to grow, you had to grow the consumer business side, but you couldn't just have the Roomba as your sort of hero product, right?
Like I imagine you're starting to think, well, we gotta come up with other things that we can sell.
Right.
You know, we needed to diversify.
And, you know, the obvious second act to us was wet floor care or mopping.
Right.
So we developed a robot called Scuba, which was a wet mopping robot.
We also looked at, we thought lawn mowing was going to be our big second act, if not mopping robots, and so that we invested tremendous amount in a robot called Terra.
You know, should we be working on a dishwashing robot?
Should we work on a laundry folding robot or a bed changing robot?
And you know, the answer is no.
But what we did do is started realizing that Roomba was an amazing smart home device, because we could figure out where your rooms were.
We could figure out whether your other smart devices were.
That idea that if I could only have a Roomba figured out where things were and what should happen and have a voice interface like Alexa, suddenly you have a new solution for the smart home.
And really that was what uh led me to saying, hey, uh iRobot and Amazon might have some pretty interesting synergies.
And because Scuba hadn't succeeded, because Terra hadn't succeeded, then when Amazon started to be interested in iRobot, uh, it felt like a good idea.
Yeah, because by 2020, you sold 30 million Roomba robots, mainly Roombas around the world, and we're entering COVID, and um, you know, there's a lot of challenges, but I think 2021 was the year where your revenue peaked.
You hit over uh one and a half billion dollars in revenue, but uh but also from what I understand, you really start to see sales drop.
I mean, you can see what what's where things are headed in part because there are uh cheap knockoffs that are coming out, right?
And there are uh there are competitors that are making robot vacuum cleaners uh that may not have been as good, but that's cutting into your market share.
Right.
So there was a period of of a real disruption starting in 2020 with the first round of tariffs, and then you have COVID, which shut everything down, so that was boom boom.
That was that was really rugged.
Then people stuck at home started thinking, hey, uh, I'm stuck at home, I should buy a Roomba.
So that sort of started to make things better.
But we weren't able to fully enjoy that upside because at the same time shipping costs went through the roof, and components started becoming unavailable because there was such demand.
And so that you know, we were spending $22 on a half a cent resistor so that we could actually build a Roomba.
And foreign competition, uh, particularly the Chinese fast follower business model, finally started to put product on the marketplace, which customers liked.
And we saw our market share starting to drop.
There was a perfect storm of just challenges.
I and this is going to lead to the to the Amazon acquisition uh offer and also why it made sense.
Uh it certainly did make sense because you could see, I imagine you as the CEO of the company could see that Amazon coming in and acquiring iRobot could be a great result, right?
And they were probably going to, they were excited about integrating this with the smart home and also building other products, maybe, or developing other products.
Absolutely.
It was like a win-win.
Our our our investors were going to win, the consumer wins, uh, and the employees of of iRobot get to be innovative again.
Amazon 2022 makes a bid to acquire iRobot for 1.7 billion dollars.
Amazing.
What an incredible, you know, close to uh at that point, you know, 32 years of iRobot, right?
I mean, what a cool result.
And so now, of course, an acquisition like this has to go through the regulatory agencies, both in the United States and in Europe, because Amazon's a global company.
Uh there's the FTC in the U.S., the European Trade Commission, I think it was what I can't remember what it was called.
EC or Europe European Commission, right.
And as the FTC was had conducted an antitrust investigation, essentially saying, or their argument what or was seemed like their argument was uh Amazon this would be a monopoly, that Amazon already had um a bunch of different robotics companies.
This was going to be a monopoly.
Amazon acquiring another robotics company is a problem.
I mean that doesn't make sense, but yes, that was that was the argument as you describe.
And so I some point in 2024, iRobot and Amazon announced that you guys would abandon the acquisition plans, but I from what I understand the FTC had not made a decision yet.
So why why did you guys make that decision?
We made the decision when it was clear that the European Commission and the FTC were blocking this deal.
And the FTC blocked this deal, knowing that they were effectively putting iRobot in a box and handing it to somebody else.
In other words, they you're saying they knew that if they blocked this deal, you guys had to sell at to some point because your business wouldn't survive, and you'd be forced to do that.
They just didn't want Amazon buying it.
That was their position.
Now we're into the land of speculation, but it does seem that way.
And also it's like you guys were offered 1.7 billion dollars.
I mean, that's that's a lot of money and a lot of lives that were gonna be changed, and they can say no, you can't do that, you can't sell it.
I mean, it doesn't make sense, right?
I I believe the FTC is there to prevent abuse, and it serves a very valuable role, but it's been taken too far.
In fact, I had to testify was deposed as part of this process.
And as walking, you know, between sessions, walking the facility, and on the doors of the agents at the FTC, the employees, like badges of honor on the doors were deals blocked.
Like people would put on their doors like uh blocked this merger.
This merger blocked.
This merger.
I mean, and I mean the biggest loser was the consumer in this deal, right?
Certainly, I certainly lost.
A robot lost, Amazon lost.
But the biggest loser was the hundreds of products, innovative products that could have uh come out as a result of this merger.
Yeah, Colin, you shortly after this unravels, you step down.
34 years at the company as CEO, and eventually within a year, I think it it was sold to a Chinese company, which now owns all the technology and all of the products that you guys built.
Yes.
We used to past tense lead the consumer robotic floor care industry.
We don't anymore.
China does.
And we chose to give it to them.
It was a choice.
We didn't have to.
And to me, that's frustrating.
Yeah, it's uh it's wild.
I mean, especially when you think about sort of the the geopolitical uh context here, right?
The the reality is that if there's a US company that spends 35 years developing technology in the US, and then that technology goes to another country, does that make sense?
In other words, if if that happened in Germany or if it happened in China and it was the the opposite ended ended the opposite way, I think it would also be people in China would be sort of asking the same questions, right?
Like hey how how is it that we spent 30 years developing this technology and now it's in the US or now it's in Germany?
So it it is um yeah, I mean it's it is it is what it is, I guess.
Um when you step down, um were you uh you you went off on to to create a new company, and I know you're still kind of in in um stealth mode.
Stealth mode, yes, right.
Stealth mode.
So you you you can't really talk about what you're doing, but it is, we can sort of assume broadly it's in robotics.
Absolutely.
It it is uh in robotics, it it's in consumer robots, it's taking advantage of this amazing new toolkit, generative AI, and using these generative tools to control robots.
Because I'm still who I am.
I'm still this builder.
And and so the when I left iRobot after, you know, uh the last decade was a little tough.
I felt like I wasn't building, I was reacting, I was protecting.
So uh, you know, all we can do is move forward.
Uh I mean I think every time I write an email, I sign it onward because the things that drive my passion still drive my passions.
And the when I go in now to talk to a venture capitalist or to talk to a factory, it's a much easier conversation than when I did it before I had done iRobots.
Sure.
You know, it's it's you know, it's a journey.
Uh and uh the journey's not done yet.
When you think about you know, this journey, I mean it's an amazing journey.
And of course, there, you know, it didn't end in the uh in the way that you wanted it to end, but but you still had incredible success with this product and a brand.
Um how much of where where you got it to and what happened with your career do you attribute to the work and the grind?
And how much do you think had to do with being lucky, the right place, the right time with some of these products?
You know, the the journey you know had some amazing serendipitous moments, but they were enabled by sitting back and actually doing the work.
Uh I had one of my early mentors when I presented in one of my first board meetings a 12-year plan where the year the revenue in year 12th was was presented to the second decimal point.
He said, Colin, never show me a chart like this again.
Try to make your next quarter, try to make your next two quarters, and then let's talk.
And and uh we don't have to have it fully planned out in order to embark on the journey of entrepreneurship.
We just need to have a north star we're heading towards.
And um, I would say that uh, you know, here we are 35 years in.
Uh, I've got one big disappointment and thousands of bright shining moments of innovation and steps forward, and uh I like those odds.
That's Colin Engel, the co-founder of iRobot.
You know the robot I want is the one who drags my ass out of bed in the morning and makes me go down and work out.
That's the robot I want.
Uh I might have something that that uh I want that robot.
Go work out now.
I know it's warm in bed.
I know you're comfy, but get out of bed.
Uh, I will I will uh hold you to that and we'll see what the future brings.
Hey, thanks so much for listening to the show this week.
Please make sure to click the follow button on your podcast app so you never miss a new episode of the show.
And if you're interested in insights, ideas, and lessons from some of the world's greatest entrepreneurs, sign up for my newsletter at guyros.com or on Substack.
This episode was produced by Catherine Cypher with music composed by Ramtina Bluey.
It was edited by Neva Grant with research help from Nor Gill.
Our engineers are Patrick Murray and Quasey Lee.
Our production staff also include Sam Paulson, Chris Massini, John Isabella, Alex Chung, Carrie Thompson, Casey Herman, Ramel Wood, and Elaine Coates.
I'm Guy Roz, and you've been listening to how I built this.