Google I/O 2014 – The design sprint: from Google Ventures to Google[x]


JAKE KNAPP: Thank you guys
for coming to our talk. I know there’s a lot of
other amazing talks going on at this moment right
now at I/O. We almost didn’t come to this talk because
there’s so many good ones. So super appreciate it. We’re going to talk to you guys
today about design sprints. And you actually totally
scored by coming to this talk because it’s not just one
talk, but two, two in one. And first we’re going to
talk about design sprints at Google Ventures
with startups, and then we’re going
to hear about designs sprints at big Google. So first of all
Daniel and I are going to talk to you about what
we do at Google Ventures. So Daniel and I are design
partners at Google Ventures. And we’re going to tell you
how to prototype and test pretty much any product
in just five days. So Daniel take it away. DANIEL BURKA:
Well, before I even get started I want
to talk to you guys about a problem I’ve got. Unfortunately it’s a
really big problem. In fact, it’s a Super
Mario scale problem, a really big Super
Mario scale problem. I love Mario. I have since it first came out. And so you can imagine, this is
going to date me a little bit, but I was super excited
when they announced the DS and announced that they were
making the new Super Mario Brothers where you could
become giant Mario. And so at the time I was
living in eastern Canada, way off on the edge
of the continent, in a little place called Prince
Edward Island where I grew up. And I didn’t have a whole lot of
disposable income at the time, but it was just after
Christmas, and I had gotten some cash
gifts from some family. And I was thinking,
aw, you know, I could totally get
one of those things, and it would be awesome. I already had a little
sneaking suspicion that I had maybe a bit of an
addiction problem with Mario, but I decided, you know what? Screw it. I’m going to go get one anyway. And so I drove up
to this place called the Future Shop
in Charlotte Town. I swear to God it’s
called the Future Shop. It’s like Best Buy. And it was no less
inviting in January then as this photo, which
I grabbed off Google Maps, is now. This is the actual shop,
a real photo of it. And so I went in there, and
I threw down my hard earned money. And it was awesome. It was just everything
I imagined it would be. So I played it. And I played it a lot. I played it every single day
for about three months straight. I beat every single level. And then I beat every single
level plus the secret levels, then every single level
plus the secret levels with all the coins
on every level. And then on the DS you could
have three different lives per game. And so I beat it once. I beat it again. And then I beat it
again even faster. And it was at the end of March. And I was just about to reset
the device so I could do it three more times, and I
was like, oh, fuck this. And I put in a box, and
I mailed it off my sister in Los Angeles. And I never saw the
damn thin again. But unfortunately I
was out about $170 that I didn’t
really have to burn and three months of my life. I swear to God, I played
hours of this game. JAKE KNAPP: Well, to throw into
sharp contrast what an idiot Daniel is I want to tell
you a story of my own. And this also involves Nintendo. This is from a simpler time. So if you remember when
Nintendo looked like this. Raise your hand if you remember. Oh, OK. Awesome. All right, great. I’m glad that some
of you are also old. [LAUGHTER] JAKE KNAPP: So my story
takes place in the year 1986. And it’s actually a
story about my wife who was nine years old in 1986. As was I, although I didn’t
know her at the time. And like all nine year olds
in the United States in 1986, I don’t know about Canada, but
everyone wanted a Nintendo. And this is what the
Nintendo looked like. It was a big box. And everybody was
so excited about it. But it costs a lot of money. So it cost $199, which if you
adjust that for inflation it’s over $400. It’s a lot of money at any
time if your nine years old. Actually, Daniel, I don’t
know for you Canadians. Hopefully this will help. DANIEL BURKA: I have to
deal with this all the time. Thanks. JAKE KNAPP: This is
actually how much maple syrup you can buy for $200
in case you guys were curious. And leaders is spelled
the Canadian way. Craftsmanship here. So anyway, my wife
is very industrious. And she scrimped and saved. She did chores around
the neighborhood, and she saved up her allowance. And finally, finally
she had $200. She was ready to
make the purchase. And then right at the last
minute she got cold feet. She was like, oh my gosh,
I saved up all this money, I don’t know if I should do
this, I’m only nine years old. I don’t know if she thought
that, but she wasn’t sure. And so she made this really
unusual arrangement that for $4 she would rent a Nintendo
from her neighbor. And she’d have it all day
Saturday, all day Sunday. She could play it as much as
she wanted, all the games, you know, the laser gun, everything. And I think that what
my wife envisioned was something like this. So this is a photo from
Nintendo on the box at the time. And this is kind
of like captures the scene of excitement that
everyone had about Nintendo. If you look closely
at these brothers you’ll notice that
they don’t even seem to notice Super
Mario Brothers is actually a one player game. But that’s what it was like. I mean, it was so great. And so she pictured
this, and she pictured the family
gathered, and everybody, you know, watching her play. And the reality
was more like this. And this is not a photo. This is an artist’s
interpretation that my wife was really
excited about when she saw it. But this is what it was like. She’s up till 3:00, 4:00 AM,
you know, eyes bloodshot, barely sleeping, playing
Nintendo the entire time. And by the end of the weekend
when she gave the Nintendo back she realized I cannot
handle owning a Nintendo. And she came to this
realization after just $4 and 48 hours of her time. This idea of renting before
you buy we think also applies to product development. And that’s the
essence of what we’re going to talk to
you about today. DANIEL BURKA: So the
way that we typically see design and development
done at startups is that you come up with
an idea, a hypothesis, something that might be
really great for your product. And then you build the
lightest weight version of it that you can, you
know a simple V1. You really boil it down. Launch it into the wild,
measure the results, learn from those results, and
then iterate around the circle. Also from our experience this
is actually not a great way to operate as a startup. And what really happens
in the real world is you frequently are
starting with a bad idea. And that’s fine. I mean, that’s the whole
point of a hypothesis is something you’re not sure
of that you want to test. But then you spend a lot of
time actually building out that idea, and this
invariably takes much longer than
you think it will. You’ve now invested
the time, and so you feel like there’s no going back. Aw, we better launch this
thing and see what happens. And where you’re expecting to
have really nice statistics and learn whether or
not something works, it’s usually much muddier
than you hope it will be. And what’s that quote? There’s lies, damn
lies, and statistics. And then when you’re supposed
to be iterating again, you actually just move on
to the next shiny idea. And your products
out in the while. And it’s gumming up the works. You’ve got all this code
now in your code base that you no longer want. It’s really hard to
go back at that point. So we’re trying to
do at Google Ventures is we’re trying to
shortcut that process. Over the course of a week we
ideate and flesh out an idea, we prototype it,
and then we test it. And we call that a
five day design sprint. JAKE KNAPP: I want to tell
you just a super fast bit more about Google Ventures in case
you’re not familiar with it. You probably have
heard of Google. If you haven’t, you should
totally check it out. It’s a create search engine. But Google Ventures is actually
a small, separate company. So about 50, 60 employees. And basically what happens
is Google gives us money, and we in turn invest
it in startups. So we’ve invested in
250 startups so far. And as design partners
Daniel, and I, and the other three design
partners on the team, it’s our job to help those
companies make the best products possible so
they’ll be successful. And we’ve done this with
countless different companies. But the story that I want
to tell you about today is actually about
a coffee company. So Blue Bottle Coffee. I’m not sure how many of
you have heard about it. Somebody. So actually a coffee
company, not just a clever name for a startup. But if you’re not familiar,
they have a number of shops here in San Francisco, a
bunch in New York City. And if you were to go to the
ferry building, which is not far from here– If
you’re from out of town you should totally check it out. There’s a bunch of shops inside
the ferry building for tourists and commuters, and
there’s one in particular that always has a long line
kind of going down the hallway. And it’s not the store
that sells only mushrooms. It’s actually Blue
Bottle Coffee. And here’s the line. And in fact, if you look
closely at that line you might recognize
a familiar sort of praying mantis style figure. DANIEL BURKA: A
praying mantis who only apparently owns one shirt. Nice. JAKE KNAPP: That’s
my favorite shirt. DANIEL BURKA: It looks good. JAKE KNAPP: As I was saying
before I was so rudely interrupted, it’s
a fantastic cafe. And they have this
beautiful interior design some wonderfully friendly,
knowledgeable baristas, and great coffee. Their shops are very successful. It’s a successful business. But they have a problem. it’s with their web store. They felt like it could be a
bigger part of their business, and it wasn’t living
up to what they wanted. So after we invested in them
they came to our design team and said, could you
help us out with this. So we decided to
do a design sprint. And the first thing we
do in a design sprint is to manufacture a deadline. One reason why people like
to ship early and ship often is because shipping
creates a deadline. It helps us get things done. I don’t know about
you guys, but I’m kind of a natural
procrastinator. And the deadline that
we create in our sprints is also external facing, but
we do it with user studies. So on Monday, the first
day of the sprint, we’ll schedule five
customers outside to come in and interview. And they’re going to
look at a prototype that we haven’t designed yet. So right away we’re like,
oh my gosh, the fuse is lit. Like we have to
get something done. Another key for us
on the first day is getting all the right people
in the room for the sprint. And there are more right
people at most companies than you’d think. It makes sense to
have the people who are going to be
working on the product, and building it,
and designing it. But we’ve also found
it’s critical to have the founder, the decision
makers, the people who really understand on the ground
how the product works, and how customers are
interacting with it. We all need to work together,
because in this case understanding how you
sell coffee beans online, it’s actually not
straightforward. I didn’t know how to do it
heading into the sprint. And so another thing
that we do on day one is to look at things
out in the world and see if we can find
interesting patterns. So we look at a bunch of
successful coffee websites, and we saw this
pattern right away that coffee is organized
quite commonly by region. So you can see it here. Latin America, Africa. Here it is again,
Africa, the Americas. Again here. This is the menu at Starbucks. Starbucks obviously knows a
thing or two about marketing. You can see Guatemalan
coffee, Colombian. And actually would you
guys raise your hand if you know the difference
between Guatemala and Colombian coffee. Does anyone here
know the difference? OK. So a couple of you do
know, which is awesome. But the rest of you
should not feel bad because as it turns out
normal humans don’t get this. I mean, most of the people in
this room didn’t understand. DANIEL BURKA: Heck, we
even had a guy came in for one of the user
studies who told us that he roasted his
own beans at home. So he’s buying green beans,
he’s roasting them himself, and then brewing his own coffee. And even he sheepishly
admitted when we asked him. He’s like, ah, when I
go to a cafe I never know really what
the difference is. JAKE KNAPP: We’re like, dude,
like you roast your own beans. If you don’t know nobody knows. And don’t feel bad. So this is like a
big challenge for us. Like how should we
organize the beans and help people decide what
to have shipped to their home if people don’t
understand the regions? We asked Blue Bottle,
because they really wanted to take the
in-store experience and kind of bring it to the web. And we said, well, like how do
you guys do this in the store? If somebody comes in and
they’re looking at all these brown paper bags, and
they want to buy some beans, how do you guide them
through that decision? And they said, oh,
well, when somebody asks for a suggestion we turn
around and say, how do you make your coffee at home,
because whether you make it in a French press, or a Chemex,
or a drip machine that’s going to help us recommend
a great roast for you. And I remember when
they said that. I was looking at Daniel
and his eyes got big. And it was as if this
beam of inspiration like shown in from outside. DANIEL BURKA: Just like that. Exactly. JAKE KNAPP: It was
just like that. This is a photograph,
so you know that it’s exactly
how it happened. And that kind of insight
actually only comes to us because we have the
whole team together. We have people who understand
every part of the product. And that turns out to be true
across all kinds of companies. So on the second day we’ve got a
bunch of insights from day one. We want to come up with a
bunch of different solutions to compete with each other. So we don’t just take one idea
and run with it off the bat. And the way we do
that is with drawing. But I want to make
sure you understand we’re not doing
group brainstorming. So we found that group
brainstorming just does not yield high
quality results. Instead what we want
are individuals spending a long time, so an hour, two
hours, sketching very detailed, very well thought
through, and very divergent opposing solutions. By the end of the day we’ve got
10, 12 different designs that are all very detailed, and
they compete with each other. They can help us pick from a
number of different options. That choice though
is not so easy. And if any of you know, you
may have been in meetings like this where you’re
kind of discussing things, and
intellectualizing, and try to imagine like
which solution will work best in the real world. And, you know, we want to get
some wisdom from everybody, but we also don’t want
design by committee. And the way we kind of hack
the decision making process is something we call
weighted voting. So first we’ll give
everybody as many as they want of these
little blue stickers. And you put the stickers by the
parts of ideas that you like. So very quickly
we get a heat map on all these
drawings of the parts of the ideas that work best. Next, we talk to the company
about how they make decisions. In this case, James the
founder, Katie the COO, they’re the ones who
really need to make the call on which
prototype we build, on which design goes forward. So we give them a number
of these very big red dots, and they have a
limited quantity. In this case they
pick three designs. So first there’s this idea
of recreating the cafe. So what if we literally took
the interior design of a Blue Bottle Cafe and made a
website that looked that way? The second idea is storytelling. Over the last day
and a half we’ve heard all of this great
knowledge and expertise from talking to the
Blue Bottle folks, and we wonder what
happens if we just write a lot of that
stuff on the website. And then third,
there’s that idea of organizing beans by the way
you brew your coffee at home. Problem is we’ve got like
three different ideas, and we still have to
make a prototype, right? So we decided to
do, in this case, is what we call a Battle Royale. We’re going to pit the three
prototypes against each other, and we’re actually going
to build all three. Problem is, it’s
already Thursday, and we only have one
day, and it seems like there couldn’t
possibly be enough time. DANIEL BURKA: So
we’re going to build three prototypes
in a single day. And we’re not going to stay
up until 4:00 in the morning to do it. We’ve got a few fancy tricks
in order to get there. The first, is that when we
were doing the sketching we were not just sketching
individual screens. We were sketching little
three step or four step flows. So when we unpack the
ideas that we want to make, we’re able to lay them
out in a storyboard. And this is, running
from the top, a storyboard of the
15 or 20 screens that a user will run through. And we can just translate
those right onto the board. And then, as Jake
was saying earlier, when people were
doing the sketching they weren’t BSing it. They were taking the
time to write real copy. They were putting where
the image might be, where buttons might be, what
realistic micro copy there’s going to be. And so we’re able to just
take it from low fidelity and push it up to a
much higher fidelity. And to move to that
level of fidelity we actually choose
intelligent tools. And we’re not working
in Photoshop generally. We often work in
Keynote for instance. This is a Keynote mock from
the Blue Bottle sprint. And we find it gives us that
Goldilocks level of fidelity. It’s neither so
basic that users know that they’re in a
prototype, but it’s also not a good enough tool to do
production design where we’re spending a lot of
time polishing buttons and making everything
look exactly perfect. We just want it to be good
enough to suspend disbelief. Keynote also has the advantage
that many people on the team can contribute to it. It’s a low barrier
to entry design tool. And then we use things
like Keynotopia, which is a little toolkit
that you can buy first for Keynote to drag and drop
form elements into the mocks, so we’re not redoing things
from scratch every time. It helps us move
really, really fast. So in the end we’ve taken
these three different ideas and stitched together
15 to 20 step prototypes of each one of them. We’ve got this one under the
brand Telescope Coffee that is exploring the storytelling,
Lindon Alley Coffee, which is doing this kind of
skeumorphic version of the store, and Potting
Shed Coffee, which is doing the filtering
interface for choosing coffees. JAKE KNAPP: So it’s Friday. And all week long this
fuse has been shortening. Now the bomb’s going to go off. This is an amazing
drawing that I did in Keynote of an explosion. But it’s time for us to
find out which prototype is going to succeed. So we’re doing the research. And I want to talk
for just one second about research, because
a lot of companies, both small and large, are
reluctant to do user research. And one big barrier
is that people feel like it’s going to be
really hard and complicated. People think, ah, I need to
have a behavioral psychologist on our team to run
the interviews, and we need a special room
with like a one way mirror, and like a laser eye tracker. And the reality is you
really don’t need much. This is our very fancy set up. It’s a laptop. And on the laptop we’re
running Keynote full screen. We’ve pasted a browser
bar on the slide so it looks like you’re
running a browser. And there’s a webcam. And using GoTo
Meeting, or Apple TV, will project into another room
so that the rest of the team can watch while one
person from our team conducts the interview
with these customers who we’ve recruited. So we do five one
on one interviews. And each interview we’re
showing the customers the three prototypes. We’re showing them the
existing Blue Bottle website, so we get kin d of a baseline. And what’s great
about this is we get really good, really deep
data from just those five users. We’ll hear a lot about why
things work and don’t work. We’ll get a sense from
watching people react about whether they understand
parts of the designs. So the results of
our Battle Royale are clear at the end of Friday. And this idea of recreating
the cafe totally bombed, which is disappointing
to me personally because it’s the
design I worked on. But it’s actually
really good news if you’re Blue Bottle,
because they didn’t have to build, and
launch, and then wait to get those
results from the wild. And you remember that this
was an idea that we really liked before we tested it. So it’s great. We dodged a bullet. And it turns out that both
this idea of storytelling and the idea of sorting
by how you brew at home were very successful. And that’s also kind of cool
because both of those ideas were a bit risky. Putting a lot of
text on a website isn’t what you suspect
will normally work. And nobody else was
organizing coffee in that way, so if they just
had one shot at it they might not have
taken those risks. Now they had the confidence
that these ideas worked, and they went ahead and did the
ship early, ship often thing. And so they designed
the full site. And here’s what it looks
like in the wild today. So you can see you organize the
coffee by how you brew at home. And you might recognize
this from the sketches. A lot of that copy
is still there. You see this really
long block of text. And conventional wisdom, don’t
put a lot of copy on the web. People aren’t going to read it. Turns out, we know that
it builds confidence that these guys
are legit and they know a lot about the coffee. They launched this website,
and it did quite well. So they doubled the
time spent on site, and in turn they doubled their
sales growth, which is great. That’s exactly what
they were hoping for from this part
of their business. But you may be thinking
like this is just a website. DANIEL BURKA: So we’ve
done these types of sprints with many different companies. Part of the reason
we chose Blue Bottle is because it’s a pretty
simple story to tell, but it’s also a pretty simple
app when you look at it. You know, everybody’s designed
an e-commerce site before. We’ve done lots and lots of
mobile prototypes as well. This is an example
from a sprint we did with a company
called Cluster. So you can see we’re
using a similar method to sketch it out,
similar ideation. Here we’re designing in
Keynote, piecing it altogether, and then we’re dragging it into
an app called Flinto, which is an excellent mobile
prototyping tool, which means that we can
get it onto device. And we’ve developed two
comps here in a single day. As you can see,
they look and feel a lot like a real mobile
app, even though they’re just static images stitched together. They’ve got buttons
and transitions. The title bar stays still. It’s good enough to
suspend disbelief so we can get really,
really valuable feedback from the users. And the second brand,
same kind of idea, all put together
in a single day. And the set up is really
similar for user studies too. Just a document camera
over top of the device. And we have the person actually
interacting on a phone. We’ve done iPad prototypes. This is a company called FitStar
that does a mobile fitness app. And here we were
actually prototyping– I’m not sure this
is going to go. We were actually
prototyping motion and sound on this prototype. And so there’s
actually video playing, and we were recording audio
during the prototyping phase. So we were actually
prototyping an audio interface where you’re stepped back three
or four feet from the iPad, and it’s giving you
instructions out loud. And that was a really
interesting study that we did. See if we can get to
the next slide here. Maybe if I press it harder. JAKE KNAPP: This
is pretty exciting. I’m going to go try to
physically press the button. DANIEL BURKA: All right. I don’t know. My technique might be off. JAKE KNAPP: Guys just talk
amongst yourselves for a moment here. I’m sure we’ll be
ready to go in no time. DANIEL BURKA: Oh boy. Jake’s the funny one. JAKE KNAPP: Yeah, so a
skeleton walks into a bar. And he says– DANIEL BURKA: Oh
here we go Jake. It’s moving. JAKE KNAPP: I’d like
a beer and a mop. DANIEL BURKA: Oh geesh. JAKE KNAPP: Thank you. Thank you for
laughing at that joke. DANIEL BURKA: So you can see
a bunch of video and audio as well. And then the examples
I just gave you, everything I’ve shown
you thus far has all been consumer stuff. And so it’s easy to recruit
users for consumer things, because they’re similar to
us, and there’s lots of people all over the world who
fit that demographic. But we’ve actually done work
with many different startups in many different areas. We’ve recruited people like
geneticists and oncologists, woodworkers, truckers even
for one of the studies that we did, and 80 other
companies across a wide variety of industries, everything
from small startups to big enterprises. JAKE KNAPP: So we’re
investing in these companies, and we want them
to be successful. We think this is an
excellent process for them to use to build
confidence quickly. And if you guys are interested
in running a sprint like this, there’s kind of three
key ideas that I think you ought to remember. One of them is
creating time pressure. It turns out that you can
manufacture a deadline in a lot of ways. The second is getting
into a prototype mindset. And we try to figure out how to
build something high fidelity as quickly as possible, rather
than building something real. Just creating a
veneer of reality. And finally, getting in this
mode of doing quick research, which really doesn’t have
to be very complicated. We hear this phrase a lot
around Silicon Valley, and I don’t think it’s
necessarily like a bad idea, but we just want to edit it
a little bit, because what everybody really
ought to be doing is learning early
and learning often. And when you do
it that way you’ll find that you can really build
great products with a lot more confidence, or as my wife
would say, rent before you buy. If you guys are interested
in running sprints we’ve written a series
of blog posts about how to run your own
sprint, everything from the kind of pins you
should buy to all the way through to how to
run the research. If you go to gv.com/designsprint
you’ll find the whole series. We’ve heard from a lot of
people who never actually talked to us, and successfully runned
sprints using that DIY guide. And thank you very much. That is the end
of the first talk. Now, you may have one lingering
question about design sprints, which is like if I
work at a big company, like this is great
at the startup, but how do I make this work? And I actually created
the design sprint process when I worked at Google. And I ran a bunch
of different sprints with a bunch of
teams across Google. But I wasn’t sure
if it could really stick in a large
organization, which Google is. Nadya Direkova is a staff
designer and design evangelist at Google. And she’s taken the design
process, made some hacks to it, made some hacks to
Google’s culture, and been able to do over
80 sprints on some really awesome products. I’m very excited to
introduce you to Nadya. And she’s going to tell you
how it’s done at Google. So without further
ado, here’s Nadya. NADYA DIREKOVA: Thank you. Thank you, Daniel. Thank you, Jake. I’m really excited
to be here today. I’ve been looking forward
to this because very rarely do we get a chance to talk about
how we do design at Google. And with the design
sprints we have something that’s
really cool and unique. Is it cool and unique because
design matters for all of us. And speed matters. If you’re a startup up,
you’re running out of money as soon as you raise them. And if you’re a
large company you might have a team
and a goal that is so large that by the
time that you meet it it becomes obsolete. By combining speed and design
we are creating design sprints to help a number of products
advance their goals. That helps us avoid
wasting time, money, and precious ideas. It is true what Jake said. I have done 80 design sprints. And perhaps I should
warn you, it’s an addictive way of working. Once you start you don’t know
where you’re going to end up. I have not done these
sprints by myself. In every case there’s
been a talented team that worked with me
to compile their goals and to decide how and
where they want to end up. So today I want to share with
you some stories about sprint at Google X, sprints
at Google at scale, and I want to give you
five hacks about how you might be able to
do your own sprints. Let’s start with
Google X. As you know, X is the new product laboratory,
and we developed things like the self driving cars. That was also the first
large sprint that I did. The team came to
me, and they said, we want to create a shared
vision of how this product can be not just an amazing
scientific discovery but also an amazing product for people. As I started talking
to them I realized that an interesting thing is
that by the nature of their job each of them are focused
on different time frames. Some people are focused
on the short term, and some people are
focused on the long term. In addition, what
was going to be available for the team in terms
of technology and capacity changes over time. So what’s available today,
tomorrow, and in four years from now really expands. The technology frontier for
anyone expands over time. So how do you create the
shared vision for something that changes so much,
and where people already perceive it in different ways? What we did is create
two teams in parallel. In five days we were able
to create even more output by sending one team
with requirements for working two years
from now, and another team working with requirements
for four years from now. They were able to create
really exciting vision videos, and show the team kind of
like a postcard of what might be happening in the future
and help make better choices. That was one week out of many,
but a very interesting week. One day I walked into
campus and I saw this. And I thought what is this? This is one of the
balloons for Project Loon. The team is sending
these balloons in the air in order to help
distribute internet for parts of the world
that don’t really get it. It’s an amazing project. It gives me a sense of wonder. At the same time what
was really important for them to create
great antennas so the signal that’s
sent from the air can be received on the ground. So the talented design team at
the design kitchen at Google X ran a design sprint. They came together. They explored different ideas. And they made prototypes,
like hardware prototypes, like various different kinds
of antennas that they can test. What’s interesting about it
is that if you put an antenna on a conference room table, you
cannot tell if it’s too large, too small, or if
it’s easy to install. It doesn’t give you
a lot of information. So you actually
have to try putting the antennae on different
pieces of something that kind of looks like a house. And, in fact, they built
a house mock up in order to be able to do that like that. How cool is that? If you go to your
team and say, let’s build a house so we
can test our ideas, they might think you’re crazy. But that might also
be the right thing to do if you want to
test your idea if it’s related to something like that. Glass is another team
that uses sprints internally as one
of the toolkits. But they also pioneered
the idea of design sprints for developers, for you guys. So the Glass development
team invites developers to come and learn
about the platform. What are the design
requirements? How do you make
something good with this? They had a chance
to try the hardware, to come up with different
ideas, and quickly put the ideas from a Post-it
into something that’s like a functional prototype. It all happens in
about four hours. The sprint is not very long. And they logged their ideas. And now they can see
what they’ve made. And they can get feedback too. Quick learning. Exciting experience. So you see how this works
pretty cool for a product lab like Google X. But
does it work at scale? If you have a huge team and a
huge goal how does it happen? Indeed, we have
found that sprints have become very popular with
teams across the company. It’s become one of the ways
in which the teams work, from Google Fiber to Hangouts. In fact, if you have used
Hangouts to call your mom please thank Jake over here. The guy created the
idea of design sprints at Google working with
team, and he used the sprint to create an early version
of what became Hangouts as a product. The team that’s
working on Hangouts is also using design sprints
to redesign and launch new UIs. So if that’s not a design
feat, I don’t know what is. The sprints here helped us
work with teams of four people, and up to teams of 175. I’m not kidding. We literally had
that many sprinters in a recent sprint
one month ago. I want to share a video
with you that gives you a sense of what it’s like to
be a part of this large sprint, where teams are running
in parallel in order to meet goals in a very
shortened time frame. [VIDEO PLAYBACK] -Great design happens
within constraints. And those constraints
are good because it helps narrow your focus. There’s something about
pressure, and in many cases conflict, that
creates good things. -So in our context a sprint is
a very focused, intense period where we have a team of
people coming together from different
disciplines and focusing on solving a specific
design problem. -Too often there’s a process
where one group works on something, and then they
hand it to the next group, and then that group hands
it to the next group. And something gets
lost in translation. So this is a mechanism where
we can all get together on the same page, and you get
much more tangible results much more quickly. -The ability to kind of leapfrog
the traditional design process is incredible. -I think a lot of people’s
idea of creativity is you have this
freedom, you know, leaning back in your chair
and just imagining things. And it’s really not
like that anymore. -Watching what was
going on yesterday, there was a lot of arguing. There was a lot of
dissatisfaction, and they had to figure
their way through that. And so today they’re actually
testing their prototypes and changing tack a little bit. And that’s kind of
the magic of it. When you see these transitions
that these teams go through. -A lot of what comes
out of sprint week is this intangible
camaraderie that helps us work better together. But the second part
are the projects. -What we’re trying to do here
is create very high quality deliverables that turn
into real products. -An actual artifact at
the end that you can then take to engineering to build,
or you can take to customers to get more feedback on. We hope that there will
be a lot of the projects that people are working
on that will get picked up in some form or another. But there’s really no losers. When you participate in a sprint
you either win or you learn. [END VIDEO PLAYBACK] NADYA DIREKOVA: All right. Win or learn. That’s the spirit of the sprint. You win when you get to
make something awesome and get totally shocked
at how fast you can run. And then you learn if
you create something that advances your goals. You know how to do
it better next time. So you might have
your design process and your own way of working. I want to share five
hacks with you that will help you
compress your design process from a large piece
to something that’s short, a week or less. We’re hoping that these
hacks will help you. First is to be aware that the
sprint is something to design. It’s worth putting
the effort into it, and here’s the workflow
for a typical sprint. There’s work before,
during, and after. Before the sprint
you want to make sure that you’re solving
the right problem. It doesn’t matter if
you create something that no one wants
to use or it’s not useful in your organization. So you need to focus
on a challenge. You want to create a team. Bring the right people together. Schedule the sprint so that it
makes sense and every minute is planned. During the sprint it’s about
letting the teams experience the right set of design
methods and research methods. And it’s about working
as fast as you can, helping resolve the
conflicts, looking for that deeper insight. After the sprint it’s
all about making sure that the results are
pointing towards launch. You want to be successful. I want to draw
your attention here to how much effort is
involved before and after. Many people skip that
part, and they only focus on the middle
of the sprint. It’s so exciting, let’s just
run and make amazing things. You want to make sure
that you’re prepared, and that you’re pointing towards
launch before and after so that you can be successful
in order to create. Now, here’s another hack. Because the design sprint is
a unique way of working we have also created a new role. We call that the sprint master. This is the person who’s going
to be the CEO of the project and make sure for the duration
of the sprint everybody’s working at an amazing pace. Now, this is not a
duck with a snorkel. It’s a ninja. These people are awesome. So here’s Jin. He’s the designer and
the sprint master. Here’s Marty. He’s a designer and
the sprint master. Ellen, Lauren, and Dave. What they have in common is
that they’re awesome designers, but they also are prepared
to be sprint masters, and they can take
their team and make them totally shocked in how
much they can achieve together. We found that by measuring
the satisfaction of sprints that had and didn’t
have trained sprint masters that the
second kind was better. People are more satisfied
with participating in a sprint that’s well crafted. So the reason also that
sprint masters matter is that if you have a team
that’s only sprinters, the most predictable
outcome is that they end up in different directions. You need the sprint
master in order to create the sense of a
common goal, of a shared goal, and a shared dream,
and let them run fast. And the sprint
master ensures that. Designing the team is another
and very interesting part of it. Now, you know from Jake that
working with designers, PMs, engineers, everyone
is really crucial. What happens if you’re
missing someone? Say you need to add the
researcher to your team because you don’t have one. The sprint is an opportunity to
redefine who is on your team. You can invite the researcher
from your larger organization, or find someone
from the industry. No one wakes up
hoping that they’re going to have a boring day in
which they do not contribute decidedly anything
towards mankind. So when you invite
someone to a sprint they have a chance
to do something cool, and generally people say yes. I’ve been surprised on how many
people I’ve been able to invite and enrich the team. Another hack is
about accountability. I put these brackets
here because I call these
accountability brackets. It’s about starting
and ending something. Just like in software you
don’t start and forget to end something because it
doesn’t compile if you don’t have the right braces, these
accountability brackets help remind us about what
to do at the beginning and at the end. No matter what your
design processes is we recommend that you start
and end with user research. At the beginning you
want to open the bracket and talk to users,
learn about their needs. It is always amazing the
sense of empathy and discovery that you can get from that. Like the time Daniel kind of
got this shining light over him because he observed
what people are like when they go to
the store to buy coffee. And you want to end
with user research. You just created something. You finished, closed
Photoshop, or you closed your
prototype and coding. And now you can bring
that to people and ask, is that meeting your goals? There’s always interesting
things to learn. The other hack
around accountability is how to optimize
the executive time. Now, we saw that
at Google Ventures the executives are often a
part of the entire experience. What happens if your
executives are not available, or not available
for the whole time? You can invite people
and your executives to visit you at the
beginning of the sprint. You can interview them
for about 20 minutes and learn about their
goals so that your team can be aligned with where
they want to go. As a bonus, you can bring
them in in the middle to get a check in. Are we going the right way? And then afterwards
you want to make sure that resources are committed
so that your sprint is going and continuing way
beyond the [INAUDIBLE] that you created it for. Now, if you’re a
designer it might feel intimidating to
go to an executive and invite them to a sprint. I understand. But think about it from the
executive’s perspective. They don’t wake
up in the morning hoping to have a boring day. By going to a sprint,
even in 20 minutes, they get a chance to
participate in something special, energizing, impactful. So it might be worth
it for them too. So starting and ending
with someone’s time, someone who’ll be approving
your work, is very important. So with that we learned
about stories and sprints from Google X, across Google,
and we learned some hacks. That’s great. But something’s missing here. I promise. There’s something that
you haven’t heard yet. You cannot learn about
design by listening to me. Talking and thinking about
it is only so helpful. I would like to invite
you to some sprints. I’m so glad you’re excited. Tomorrow at 9 o’clock there’ll
be a sprint with Glassware. The Glass team is putting
a sprint for you guys. And at 11:00 the legendary
team at Google Ventures will be running a sprint
for you guys as well. These will not be
five days experiences. I promise. AUDIENCE: Yay! NADYA DIREKOVA: There’ll
be two hours in which you have a chance to learn about
the basics of the methodology and to test and learn
something quickly. Later this summer we
would like to welcome you to additional sprints for
Glassware Material Design that got announced today,
and for Android Wear. You can sign up. And if you’re interested
in this way of working, let us know at this form. designsprints with an S 2014. Let us know and bring your team. I want to leave you
with this message. Make great things. And you might as
well do it fast. The world needs you. [APPLAUSE] NADYA DIREKOVA: OK. JAKE KNAPP: Do we have time for? NADYA DIREKOVA: Let’s check. JAKE KNAPP: Oh, we do. We have four minutes. NADYA DIREKOVA: We have four
minutes for your questions. JAKE KNAPP: Fast questions. NADYA DIREKOVA: Take it. JAKE KNAPP: Right there. Yeah. First hand up. NADYA DIREKOVA: There’s
two mics here and here. So you can line up and
give us your questions. JAKE KNAPP: Oh sorry. DANIEL BURKA: Or just yell
it out, and I’ll repeat it. AUDIENCE: How do you go about
selecting users for your user testing at the
end of the sprint? JAKE KNAPP: Great question. So how do we go about
selecting users for our studies at the end of the sprint? The method that we
use at Google Ventures is actually mostly Craigslist. We’ll post an ad on Craigslist. We’ll get 200, 250
responses in a major city. And what’s really important is
that we have a screener survey. So we’ll use a
Google survey to ask them to answer a bunch of
questions that doesn’t reveal exactly what we’re
looking for, and then we get a spreadsheet
we can go through. The way you structure
that survey, the way you write that ad is
really important so that you don’t get, you know– You
want quality customers that match exactly what
you’re looking for. And there’s some posts on there,
if you follow that same link, gv.com/designsprint, you can
find your way to some posts about how to do that. For some customers, just
to add one more thing, if you need a certain
kind of expert user we’ll usually rely
on the company to use their connections. AUDIENCE: OK. My question will be what are the
general questions that you will be asking the users on
the interview session after you have the prototype? I mean, you must have
a template questions to facilitate your
design to the users. What are those templates
and general questions? Thanks. JAKE KNAPP: The best source
for those questions is also gv.com/designsprint. But to give you a
little tantalizing hints of what it’s like
the key is actually to structure the interview
guide right to the prototype. So you want to get people
reacting rather than giving you feedback. You want them to be moving
through the prototype as though it was real. And the best way to do that
is actually a little nuanced. So I’ll totally direct you
to check out the blog posts for all the detail about
writing those questions. AUDIENCE: Cool. You guys spoke about two
pretty different scenarios. You spoke about
sprinting as individuals, whereas you talked
about bigger groups and sort of teams that sprint. My question is when you have
groups that sprint together, do you interfere at all with
the dynamics of that group? Because sometimes you have,
you know, really strong voices, and really shy voices,
and stuff like that. Do you at all interfere with
the dynamics of the group, or do you just kind
of let it happen? NADYA DIREKOVA: The question
is whether in large groups trying to interfere or not? That’s the role of
the sprint master. I don’t call it interference. I call it leading. JAKE KNAPP: We also
want, when there’s a dynamic about how decisions
are made in the company, we want to understand
it and expose it. We want to ask
them how it works, and make sure that the
way we make decisions mirrors that so
that we don’t create a false sense of collegiality. DANIEL BURKA: A lot of the
decision making stuff Jake was showing with
the stickers, you know, we’re not standing
around arguing with each other, because then the person with
the loudest voice can win. We’re stopping, doing a
lot of individual thinking, and then voting individually. So people are much more
independent of each other, and not susceptible to
so much group think. AUDIENCE: Hi. What are some of your tips
for enforcing deadlines when you need to wait
for significant data to make intelligent iterations? JAKE KNAPP: Well, I mean that’s
a pretty complex question. I think that what significant
data is defined as is like– I don’t
know what you mean. We think that you can
get really good data from a lot of different ways. And one of them is these
kinds of user research. There are other
ways to get it fast. And I understand sometimes there
are certain kinds of decisions that you can only get
from a real world launch. But our framework,
our lens, is what is the fastest way we can test
this hypothesis that we have. There’s something we don’t know,
and how can we get to an answer as quickly as possible? We need to test it somehow. And if we think of it as
a prototype and a test it opens up a new
world of possibilities. When you think the
only way you can get data is to
launch and measure a live product, or a fully
functional live product, it limits you. So without detail
it’s hard for me to not give that kind
of hand wavy answer, but I think the key
for us is always thinking what’s the fastest
way to answer the unknown. NADYA DIREKOVA: We’ll
take one more question. AUDIENCE: What tools
or modifications can you suggest for working
with a distributed design team? So if they can’t physically
be there to put stickies do you have like
tools they you use? JAKE KNAPP: The best
tool is a plane. If you can just get them there
that would work the best. Nadya, do you have
any experience with this that’s worked well? NADYA DIREKOVA: You
can do Hangouts. I mean, it’s best
if you’re together because you really
feel that bonding. But you cam work in a virtual
way using like virtual tools. You just need to plan
it, or think about it. AUDIENCE: Thank you. NADYA DIREKOVA: You’re welcome. JAKE KNAPP: Thanks
you guys so much. You can chalk us down. Thanks a lot. [APPLAUSE]

12 comments

  • ratan mojo

    Great talk, indeed inspiring.

    Reply
  • Bruno Garcia

    Great talk 😉

    Reply
  • Miroslav Košík

    I would LOVE to be in a Bay Area for these sprints :-/ Have a gr8 time guys 🙂

    Reply
  • sproket343

    Been working this way for the last 10 years. It works great. Some interesting subtleties here, but in general fail early fail often and definitely fail before you build is good advice. 

    Reply
  • ZeikJT

    Interesting that at dollar value per day both Daniel and Jake's wife paid almost the same amount. Daniel paid about $2.15 per day ($179 divided by 79 days according to the chart) and Jake's wife paid $2 per day ($4 for 2 days). Good thing Daniel enjoyed his time with it 🙂

    Reply
  • SHOYWEB

    Thank you for the help!

    Reply
  • David Arnoux

    Isn't it risky to kill a design based on 5 data points only? You're leaving a lot up to chance… you could simply have been unlucky and fallen on 5 people who don't like it whereas the next 15 would have loved it. I love the design sprint concept but I'm concerned about the statistical significance of the data on day 5…

    To put it more simply.. if you flip a coin 5 times and get tails 5 times, does it really mean you have 100% chance of always falling on tails?

    Reply
  • Professor YinYang

    17×10≠250

    Reply
  • Martino Liu

    What does a Hack means in this context? Minute 33:47

    Reply
  • Akbar Ali

    Who else is here from T Blake!!

    Reply
  • Challis Hodge

    "Some of you are also old?" Exactly what do you call old???

    Reply
  • Karan R

    Daniel you sound like Mr. Gates! 😉

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *