WIRED25: The Future of Work With Stacy Brown-Philpot, CEO of TaskRabbit
Released on 10/12/2018
Stacy Brown-Philpot,
the CEO of TaskRabbit,
please come on stage.
(applause)
Hello, how are you?
You know you know it is a thing
when you're from Detroit,
people ask you
what high school did you go to?
It's like, what kind of coffee do you drink?
Well I asked I asked John this morning I said,
Do you have the same part of the tricks?
I noticed that you and I are exactly the same age
and John and I are like three days apart.
And she said, Well, we actually went
to the same high school.
And I said, Oh, my God
(chuckles) that's awesome.
Um, well, let's start with some questions about TaskRabbit.
Um, I was just reading a Morgan Stanley report,
which was written up in the Wall Street Journal.
It was showing
the wages of people in the gig economy
and was mostly about the wages of people at Uber and Lyft,
which are
down 53% over the last five years.
And the reason for that, supply and demand, right?
More people become Uber drivers, and...
or the number of people who become Uber drivers
is rising faster than the number of people who want them
so the wages go down.
But TaskRabbit
has stayed constant.
It's up in fact,
slightly over 2013.
Yes.
Why is that?
How've you been able to keep wages constant,
whereas your other gig economy competitors
have seen this drop?
So, TaskRabbit is a task management network
that's built on 10 years of
relationships and community bonds.
And so because of that,
we've built a community of Taskers
who set their own hourly rates.
Back in 2008,
when it was the worst financial crisis of our time,
still,
there's probably another one coming but
I know, it feels a little currently,
like this is your tenth the worst financial
anniversary it feels a little familiar
this week (laughs) It's right,
its our 10 year anniversary, this week actually.
Uh, It was the worst financial crisis of our time.
And earlier this week, I called
one of our longest standing Taskers in Boston,
just to thank him for being with us for 10 years.
And he said, You know what?
TaskRabbit really looked out for me back then
when I lost my job,
as a contractor,
and I could set my own hourly rates.
And now here we are today
and we still are five times the federal minimum wage.
So we built this company that allows
Taskers to have the flexibility to set their wages.
And even though the economy has grown,
and he had to go because he was on a task,
(chuckles) cause yet he was painting
somebody's house.
And he's like, I gotta go, the business is booming.
So we're not in the worst financial crisis of our time,
the business is booming and yet,
he can still command a meaningful income
and a meaningful hourly rate.
So we want to build
and continue to build a trusted network
for people to get help for life,
with quality people who can provide it
and the income that they earn from it,
is what helps us do that.
But even if people are able to set their own wages,
right, so let's say I wanna do carpentry
and I see that everybody in San Francisco does carpentry
at $35 an hour.
I'm just gonna set mine at 34
or 33, because there's probably a premium
for people with experience.
Why does that not just drive the wages down?
Because no one would show up for the job.
So, if I'm deciding what I'm worth,
I bring a certain set of skills to the table,
then
I determine
what I'm worth.
So if I'm looking for help,
and the market says it's $34 or $35 or $37,
then that's what I pay.
If no one's gonna charge 20 to $20 an hour,
then you're not gonna hire someone for that rate.
And so you expect that the wages that the TaskRabbiters
get, for the next 10 15 years will continue to be stable,
to climb. You won't see what's happening to Uber and Lyft?
Well ultimately,
the Taskers who task in our network,
and we have over 140,000 now,
want to earn a meaningful income.
So we only want people who wanna earn a meaningful income
to be in the community.
When they do that,
that brings a great experience to the client
who's getting someone to get something done.
Because of that,
that flexibility
allows them to set the rate that means the most to them.
And so, I don't know how wages will change?
Wages will change depending on
why you're tasking on TaskRabbit.
Some people are part time,
and they're just trying to earn some extra income
to go on a trip.
Some people are part time because
they are stay-at-home parents
and they need some flexible income
but they have to pick up their kids.
And some people are full time,
who are putting their children through school
and paying their bills every month.
And that meaningful income for them
is really what determines the wages on our platform.
I've had all 140,000 of your TaskRabbiters
come to fix things in my apartment
that my kids have broken. (audience laughs)
Very good. (laughs)
Let's talk a little bit about the demographics
of the TaskRabbiters.
Is TaskRabbiter the proper phrase?
Tasker.
Tasker? Yes.
I kinda like TaskRabbiter.
Yeah. (audience laughs)
They, they don't. Yeah.
(audience laughs) And it's the 140,000
of them.
(audience laughs)
One of you. (laughs) We're going with Tasker.
Sir Tasker,
or Man Tasker. (laughing)
Okay, so the demographics are...
I was looking at them in the same Morgan Stanley report,
and they were,
kind of what you might expect, at first glance,
right? There more sort of, I think between 30 and 40%
then it declines 40 to 50 declines 50 to 60
declines 60 to 70.
But one of the hypotheses, that,
one of our writers at Wired has put out, is that
older people will actually start making up
a much larger proportion of the gig economy
cause they are more flexible, right? They are,
they don't have the same salary requirements,
they can move around more.
Do you anticipate that the over 65 demographic
will soon start to make up
a larger percentage of TaskRabbit?
Or do you think that the deciles we see now
are about where they'll stay?
Yeah, so over the last 10 years
that our task management network really has grown, I mean,
the numbers obviously have grown.
Most people are still part time,
and there's a lot of people who are full time,
but most people are still part time.
And it's not because
TaskRabbit isn't meaningful enough for them,
it's because they've got another part time job
and they have to have three part time jobs to equal
the full time income that they wanna earn.
And they're between the ages of 30 and 40
35 and 45 years old.
So that's been a trend that's pretty much stayed the same.
These people are really interested
in doing something that matters.
They're at a point in the stage of growth
in their life, in their family or income needs.
And I expect that to stay the same.
We do see a lot of retired people
decide to task on TaskRabbit.
And it's because
the connections that you make,
I mean, really the relationships and the bonds
that get formed from going out and completing a task
is something that everyone seeks to have
over the course of their lifetime.
And we find that retirees come out of retirement
and find a lot of meaning on TaskRabbit,
not just because of the money that they earn,
but because of the connections and the relationships
that they build and the bonds that they make
in their own community.
I would expect that one reason why older demographics
have a lower participation rate on TaskRabbit
is also partly just unfamiliarity with
booking tasks on your phones, or sort of,
fear of that or modern
I can't imagine my mother, you know,
doing that.
But as we get older, we'll also
we'll be past that,
right? We'll be Yeah.
I think everyone will be totally comfortable
with everything on their phoness.
Well, we're in the first digitally native generation now.
And even in that generation, only 4%
of people have gone online to find help
for something around the home, which is really small.
What we've invested in is,
making our app more easy to use.
So if you're a Tasker, there's a different app
than the app that you use if you're a client.
And we've invested our technology in making that app
very easy to use, very simple, very easy to understand,
so that anyone of any age can task in the community.
So let's talk about your matching algorithm.
So you have an algorithm that when I enter a task,
it then says, These are the people who are qualified.
And obviously, there's a huge premium for you
in finding someone who can do the task,
who can like repair the blind, technically has those skills,
even though it's a hard skill to describe,
who I'll like and whose scheduled matches my schedule.
So explain the key elements in your matching algorithm
and the ones that have improved the most.
So, what what we've done, so a test management network is
it's about the information that we collect.
The information about the Taskers,
the information about the clients.
Over the years, we've evolved our business model
to actually collect more information about
our Taskers in advance.
We know, who they are, what skills they have,
they create a profile, what their hourly rates are,
they tell us when they're available in their schedule,
and we give them demand maps on,
based on their availability in the areas
that they're interested in working in,
where demand is gonna be, that day.
So that's evolved over the last several years for us.
For our clients, we collect information from them on,
what do you need to get done, um,
you know, what category is it? A minor home repair?
Is it furniture assembly?
Is it cleaning? And then a little bit of details from you
about how big that job might be.
And so our matching algorithm,
that our data science team built,
takes all that data and creates the best match.
So we focused on improving the match.
We focused on understanding
how Taskers perform on prior tasks
that are similar to yours.
We focused on understanding that
the text that you enter about a task
might mean something more than you're saying,
because you often don't know what kind of wall you have
for that TV that needs to be mounted.
Right.
You think you know, Right.
but you might not actually know.
So we can predict how's, if when you're specific enough
about the kind of wall, if you don't tell us
you probably don't know.
So we make sure we have a task
which is flexible enough to kinda
put a hole in a bridge That's a very interesting
So what are,
like, what are some of the interesting advances,
particularly in AI, that have made your algorithm
work better.
There are ways of parsing language that you didn't have
a couple of years ago?
There a way of parsing images
that make this matching algorithm work?
You know, the interesting thing we've been able to do
is recommend the right category for you.
Aha.
So often you have something that needs to get done,
and you write in, Oh, I need my gutters cleaned.
Right.
And you don't know if that's Yard Work.
If it's home repairs, like where does that actually fit?
And because we are better at
well, who are the taxpayers who do yard work or home repairs
or in fact who just who just does gutter cleaning?
You can tell us in your to-do list.
Just create a to-do list and
put in, just go to Cleaning
and we'll recommend the right category for you.
Once that's selected, you have a much higher chance
than your own human understanding of
what a gutter cleaning is, Yeah.
to actually get the right Tasker to do the job.
And you have a satisfaction percentage
that you've seen climb over time based on
whether you've recommended the right person
or not recommended the right person?
That is all predicated on the reviews. So,
taking that to do list, making recommendations for you,
the customer then says, Yes, I'm satisfied,
I'm not satisfied.
It's all based on the quality of that interaction
of which part of it and most of it is,
did we find you the right person
with the right skills to do the job at the right time?
And what have you learned about the way we live
from the kinds of tasks that are requested?
(chuckles) Like, I would imagine,
there's like, way more TV installation cause we like
stare at our TVs all the time, (laughs)
right? And eventually, they'll be less like,
you must have a pretty good insight
into the American home life Yes. (chuckles)
from what's requested. So
tell me something I don't know
about how we live.
So we, um...
we care so much about life at home
that we know that everybody has this storm cloud
that's just like looming over them.
And the way we live is we
let that storm cloud sort of build up.
And so something, an event will happen.
So now there are five pieces of IKEA furniture
sitting on your floor (audience laughs)
that need to get put together.
And one day, someone in your house who's tired of it,
decides that we're just gonna hire somebody
from TaskRabbit to do it for you.
And when that happens,
the doors open up. There's that hole in the wall
over there, that's been nagging
that everyone's been ignoring.
Or you kind of put a box, the Amazon box,
in front of it so no one sees it,
and you actually fix that.
Or the Tasker comes in and says, You know what,
I can fix this creaky door for you,
which is irritating in the middle of the night,
cause it might wake up the baby.
I mean, there's all sorts of things around us,
that creates small irritations that
have this whole storm cloud and then,
all of a sudden, TaskRabbit is the stress-free way
to just
take care of everything that you need
and be your trusted network for you.
So the, the progress in society,
for many of us,
has been that we are doing so much
that the small things often get forgotten,
and sometimes those small things
are the interactions we wanna have with our children.
And sometimes the small things are the annoying things
that we just always forget to do. But they create stress.
And so the way we live,
and the way TaskRabbit helps
is we now can be your go to network
for taking those little stressful things out of your life.
But you don't get that door opened
until like those five, it has to be five for some people.
(chuckles) some people is three
and some people is five.
But eventually you let go
and you let someone come in to your life
like from TaskRabbit and help you get those things done.
Just find a hero to map my tasks
and go through my task history,
you'd have like a map of a life, right?
It's like painting your walls
when you care about your apartment. And then it's like,
you know, installing the crib
and then it's, you know, Yes.
removing the crib. And then it's like
fixing broken furniture.
Yes. (laughs) (audience laughs)
So what are the tasks that you think that
robots will be able to do? What are the tasks where
you would advise TaskRabbits to, you know,
diversify and find some other skills.
Yeah, I think Nobody needs to sweep
their apartment anymore, right?
You can get a Roomba to do that.
So, as robots become more sophisticated,
what will they do that TaskRabbit do now?
Yeah, I think um
there's always this question of like
what are robots going to replace?
And I just believe that robots will never replace
the empathy and judgment that humans can provide.
And so what I do think will happen is that
our Taskers will get more sophisticated.
They won't necessarily have to interact with you,
because your refrigerator will tell them
that the filter needs to be replaced
then someone will just show up and do it.
Or they'll have a lot of better clarity on
what needs to be done,
how big is that room in your house?
Um, can we really get this item through the door
(audience agrees) that you want to get
the item through
because you think Most certainly not
it can fit through the door. But there'll be enough AI
that says okay, if we want to get this item into the house,
here's how we do it.
Can this couch really go down those stairs?
Because when we get there to help you move,
like, what is it gonna take?
So the sophistication of how we execute on a task
will make a lot more efficient, a lot easier for our Taskers
to succeed in helping you
and then we can apply the judgment in the areas where,
you know, you really want the TV over there
but optimally speaking, you know,
this distance is the best way to watch football,
for example.
You said something crazy in there.
So you said
the refrigerator will call TaskRabbit
when it needs to be cleaned?
What?
When the filter needs to be replaced.
When the filter needs to be replaced?
So my appliances are gonna be talking to TaskRabbit
and ordering Taskers?
And then are those Taskers gonna have
keys to my apartment? Is gonna be like Amazon?
(audience laughs) Maybe.
Do you have appliances in your house
that call Taskers (audience laughs)
and get fixed, without you seeing the Tasker?
Not yet, but
isn't that the robot you really want?
Right? Isn't that the robot you want is the
your refrigerator that's like
We need more fruit, for a snack time tomorrow.
And the filter
needs to be replaced As long as
it shares my values
Which is why the water (audience laughs)
is dripping, (laughing)
and no carbs, because we don't eat carbs in our house.
Right? That's, you know, that's,
that's really what you want, right? And by the way,
we can get TaskRabbit over here
to fix just So are you integrating
with Amazon. Are you integrating
into Amazon dash buttons and stuff?
Not right now.
Will you?
Maybe.
Tell me more. (laughing)
I mean, this is amazing. (laughing)
Tell me
Imagine a future where all the technological problems
you wanna solve are solved.
Things advance, people become infinitely comfortable
with TaskRabbit.
What is the maximalist version of what TaskRabbit
will do in my life in 10 years?
It's gonna create human connections for you.
I mean, that's really what we're all about.
What we do is help people create human connections.
Our task management network is trusted,
it's your go-to team, you have this storm cloud
and there's somebody there
who's going to give you help for life,
whatever that is for you.
But it's not just human connections,
it's not just human and human connections.
It's appliance to human connections.
(audience laughs) Yeah, but you're
But that's okay with the appliance.
That's, that's just I mean, I think it's awesome
You don't have to to enter it in,
we just take care of it for you.
Does that make you feel better?
(audience laughs) If you can do it right,
and you can actually identify
if my dishwasher can identify
when the hinges are coming loose
and when the hinges need to be tightened,
or whatever it is that, is broken right now,
and you can get a Tasker there, I will be thrilled.
Right. That's what we want.
Okay. For you.
And so how are you
how exactly are you building that?
Are you and what are the steps you need to take
to accomplish that. Clearly, you need to
build up the trust network,
make the matching algorithm work,
but you also need to integrate with the companies
that make the appliances. What are the other steps
that we need to do
to get to that Yeah. And that's
part of what we're doing. Is we're
we see ourselves and leading the future together.
And it's the partnerships that we will have
with companies that are already building those technologies
and thinking about those technologies. So when
we think about technology investment at TaskRabbit,
it's obviously the operation
that makes it really easy for you to find a great person.
But it's the technology in the background
that can integrate with partners that are building these
these machines.
Okay. So then, one last question. And it's a,
it's a big question, which is,
there's a lot of talk about bias in algorithms, right?
And you've read all that,
you've spoken eloquently about it.
How do you prevent bias from integrate
How do you prevent bias from entering TaskRabbit algorithms,
so that Taskers of different genders or races
are selected more? How do you prevent that
and have you seen any of that?
We've focused on diversity and inclusion at TaskRabbit
from the very beginning. So the people in our company,
we seek that for those people to reflect the population
that we serve.
So we're proud to have a very diverse company
that is building the products
and that is the core of how you prevent these things.
Who's building the products?
Because they can understand what the technology is doing?
Is there bias in it or not?
And address it right away.
Okay. Wonderful.
Well, we have no TaskRabbits cleaning up after this event
(laughing) but in the future we will.
Stacy thank you so much for coming on stage.
That was a fantastic conversation
Pregnancy Doctor Answers Pregnancy Questions
Jen Easterly On The Future of Cybersecurity and Her Agency's Survival
Psychologist Answers Couples Therapy Questions
Building The Biggest Telescope You Can Buy
Football Historian Answers Football Questions
Bill Gates on His Early Years & The Inspiration Behind 'Source Code: My Beginnings'
Paleontologist Answers Extinction Questions
Five Things to Watch for in Mark Zuckerberg's Congressional Testimony
Taking on Amazon, Google & Silicon Valley: EU Competition Chief Talks Tech with WIRED's Nicholas Thompson
How Humans Get Hacked: Yuval Noah Harari & Tristan Harris Talk with WIRED