00:00:01.520
um you came thank you so much I'm actually really honored to be here so thanks to Chad and Ben uh and to the
00:00:07.560
O'Reilly team for inviting me you're not going to believe this but I know other speakers dream about you know speaking
00:00:13.160
at Ted and going to Davos and whatever uh but I've been sitting by my phone for a long time being like when am I going to get invited to rails comp and I know
00:00:19.840
you think I'm kissing your ass so it's but it's really it's actually the truth um there's something I know I work with
00:00:25.279
startups I do this thing called Lean Startup which you'll get to hear about more than you ever dreamed
00:00:31.119
but rails has done more for startups than like a whole boatload of venture
00:00:36.320
capitalists rails has had an incredible impact on the startup ecosystem and I see it every day uh but I learned a tip
00:00:43.079
from Gary ve which is make sure they know who the hell you are before you start talking so I write this blog called startup Lessons Learned anybody
00:00:49.640
any readers in the audience can I see whoa hey thank you thank you very much that's exciting uh for the rest of you
00:00:55.239
how many of you are entrepreneurs are in a startup right now just quick show of hands wow okay yes thank you that's what I
00:01:01.480
figured um everywhere that I have seen rails uh peripher of course especially in Silicon Valley um startups spring up
00:01:09.159
everywhere and I just wanted to share a few thoughts about why that is the Lean Startup is now uh movement W
00:01:16.960
whyde movement as you heard I've been traveling all over the place uh talking about a different way of building
00:01:22.280
startups a different way of thinking about entrepreneurships entrepreneurship and I think it's uh very compatible with
00:01:28.040
what you guys have been talking about in The Rails community I say you guys because I'm kind of a faker uh I used to be an engineer I was CTO of several
00:01:34.720
companies most of which have failed hi I'm a professional startup expert but most of my companies have failed I know you're not supposed to start that way
00:01:41.320
but that's the truth and anyone who's willing to tell you the truth about entrepreneurship not the stories you read in the Press but the real stories
00:01:46.600
you're going to see there's a lot of failure um so you know when I actually wrote code there wasn't that much rails
00:01:52.240
uh going on and now to see it grow into such an incredible Community has really been cool see I told you you think I was
00:01:57.399
kissing your ass sorry uh let's just do a some quick ground rules the first is uh who has a mobile phone show of hands
00:02:03.719
just kidding see if you're paying attention could you take it out of your pocket and make sure you turn it on so phone's on I do not want your undivided
00:02:09.959
attention please tweet amongst yourselves as much as you like all I ask is that you use the Lean Startup hashtag
00:02:15.360
uh while you do so then you can join the worldwide communic uh conversation about Lean Startup okay fair um I have this
00:02:22.840
new book coming out and you know when you're writing a book you're pretty much obligated to go everywhere and tell
00:02:28.000
people they should buy your book it's coming out setember but don't worry you can pre-order it now at lean. St uh I
00:02:33.360
can wait while you do that just you can all take a moment it's fine I don't I really don't mind no I appreciate everyone who has pre-ordered uh it's an
00:02:39.879
incredible uh incredibly supportive thank you um but I did want to just give a quick shout out to um pivotal Labs who
00:02:46.239
built this website for me at lean. St in rails of course uh it's hosted on Heroku so thanks to pivotal and Heroku my
00:02:53.040
friends at pivotal heard I was going to be speaking here and they said listen uh if you just so happen to run into a couple of thousand rails developers
00:02:58.879
could you tell them that we're hiring in so pivotal is hiring just I thought I'd pluged that um lean. St is not just a
00:03:04.680
website to sell the book it's also a fully featured experimental platform for testing how to sell the book because
00:03:10.480
it's Lean Startup as you'll see in a minute that's what we're all about uh and so we have built it uh to make
00:03:15.840
experiments on you as easy as possible for us to run to learn the most uh and if you do pre-order at lean. St you can
00:03:21.799
actually go behind the scenes and see every single experiment that we have run all the data about uh how customers have behaved you actually watch us
00:03:27.799
experimenting on you in real time so you could do that at lean. SD and I'll stop plugging things um these are some principles from
00:03:35.040
the book that I wanted to share with you uh entrepreneurs are everywhere entrepreneurship is management this
00:03:40.640
thing called validated learning the unit of progress for startups this feedback loop we called build measure learn and then something called Innovation
00:03:46.640
accounting and at this point I should probably apologize because you came for an exciting talk about entrepreneurship which is like the coolest topic ever and
00:03:53.680
now I'm telling you we're going to talk about management and accounting which are probably the most boring topics imaginable so I apolog ol but I actually
00:04:01.239
think if we're going to get good at entrepreneurship we have to make it a lot more boring so here's what I mean
00:04:06.599
who saw the social network just quick show of hands right okay uh who saw Ghostbusters two great entrepreneurship
00:04:13.519
movies people forget that about Ghostbusters most of the plot of Ghostbusters for those that remember is
00:04:18.799
concerned with these crazy guys with this new technology get thrown out of a university and how they're going to build their fledgling business and make it be successful and like all great
00:04:25.280
entrepreneurship movies social network included uh it's a story in three acts so act Act One the Plucky protagonists
00:04:32.000
and their uh character strengths and defects which will be very important in just a moment how they had the great
00:04:38.199
idea and maybe how they got their first customer and then we shipped immediately into act two which I call the photo
00:04:43.440
montage usually lasts about two minutes and it goes you know you on the social network it's like them pounding on keyboards they chug some beers in
00:04:49.680
Ghostbusters they bust a ghost or two and then all of a sudden they're on the cover of magazines uh and everything's
00:04:54.880
really successful and then we go into act three the really interesting part how to divide up the spoils who sues who and what happens to the Stay Puff
00:05:00.759
Marshmallow Man and that is an exciting story those movies are great movies for a reason and
00:05:05.800
when you read stories about entrepreneurs in magazines everywhere you see those great stories it's really fun and so I have this question which is
00:05:12.400
how come the photo montage is so short and it has no dialogue when in my opinion every
00:05:19.120
important decision that determines the outcome of a success or failure of a startup happens during the photo montage
00:05:25.000
and the answer is because all that work is too boring for the movie
00:05:30.319
I mean who's ever sat through a product prioritization meeting anybody right does that seem like movie Material to
00:05:35.919
you no of course not and yet product prioritization decisions what features to build and not build which bugs to fix
00:05:41.080
which customers to listen to and who to ignore those are the really important but really boring topics of
00:05:46.520
Entrepreneurship and as an industry we have to get better at not just doing that stuff but also talking about it
00:05:52.560
because the exciting stuff how do you get your great idea you know how to divide up the spoils I know those of you who are entrepreneurs already have you
00:05:58.240
know worked it out in great detail what are you going to do once you get very rich and these are very important questions they're very exciting they're like the 5% of Entrepreneurship and yet
00:06:05.400
the 95% is the boring day-to-day mundane details and that's what we're going to talk about today so sorry to disappoint
00:06:12.199
if you were looking for something exciting um entrepreneurs are everywhere what's amazing about asking that
00:06:17.280
question hey who's an entrepreneur people ask people to raise their hands I've done that in countless cities and countries around the world I've never
00:06:22.440
once had someone say excuse me I'm not sure whether to raise my hand or not everyone thinks they know what
00:06:27.479
entrepreneurship is you know if I'm doing that crazy thing like in Ghostbusters then I'm an entrepreneur and if I have a regular job at a normal
00:06:33.800
office you know then no you know no poll to slide down then not and yet I think
00:06:39.720
we need to do better than that if we're going to start making our practice of Entrepreneurship more rigorous so here's my definition of a startup a Human
00:06:46.440
Institution designed to create something new under conditions of extreme uncertainty what I think is most
00:06:52.840
important about this definition is what it omits nothing to do with how big your company is what sector of the economy
00:06:58.280
you're in what industry you're in if you face the conditions of extreme uncertainty which are the soil in which
00:07:03.479
startups Thrive then you are an entrepreneur whether it says that on your business card or not which is all just a fancy way of saying that a
00:07:09.240
startup is an experiment but unlike in past eras today our experiments are not
00:07:15.319
experiments in can something be built right we're not trying to figure out hey can we make the flux capacitor work the
00:07:21.160
question is should it be built can we build a sustainable set of products and services around this idea that is the
00:07:29.080
question of of Entrepreneurship and since uh we're faced with that question more and more especially as our technology gets better and better we can
00:07:35.639
pretty much build anything that we can imagine so that means that our future
00:07:40.960
Prosperity unemployment crisis GDP everything will depend on the quality and effectiveness of our Collective
00:07:47.159
imaginations which is kind of a strange place to be and the problem is that we are wasting people's time on an
00:07:53.520
industrial scale because most of the startups that we build are a complete and total waste of time and are doomed
00:07:58.759
from the beginning so those of you who raised your hand who are entrepreneurs thank you so much you are changing the world most of you and yet most of you
00:08:04.440
are going to fail and I think that is something we can do something about but just in case you weren't don't believe
00:08:10.240
me about the whole failure thing I I maybe you've heard Entre entrepreneur are a little bit optimistic I brought a
00:08:15.280
demonstration so we know that most tups fail anyone here remember Web 2.0 remember when Web 2.0 was really cool
00:08:21.280
and those of us in the valley were busy hyping the hell out of these companies and pouring way too much Venture Capital into them uh this is a graphic that was
00:08:28.199
made in 2006 by designer who was really excited about the potential of Web 2.0 and then in 2009 just 3 years later a
00:08:35.800
different graphic designer was feeling a little bit different when they put together this graphic here's our
00:08:40.880
three-year report card in Web 2.0 uh and it's pretty sad most of those companies
00:08:46.200
have failed and I don't think they failed for good reasons and I think we have to start to ask ourselves why are we building so many companies that fail
00:08:53.040
you know it's not like everyone understand entrepreneurship is risky but it's not like we're trying to build quantum teleportation here we're
00:08:59.079
building things especially in the software industry especially in Web Web 2.0 everything we we can imagine we can
00:09:05.000
build so why are we choosing to build things that nobody wants that's what we're going to talk about so it has to
00:09:10.959
be somebody's fault uh you know most of my startups have all these startups are failing obviously can't be my fault
00:09:16.640
preferably it' be somebody who's dead uh so he can't argue so I blame Frederick Winslow Taylor uh this is the father of
00:09:22.959
so's called Scientific Management what we today call management uh and Taylor is famous for a number of ideas most of
00:09:28.880
which we can considered to be too obvious to imagine them ever being invented like that managers should
00:09:34.959
divide up work into tasks and delegate those tasks to functional Specialists or that my favorite what he called the task
00:09:40.240
plus bonus system which we just call tasks that if someone does their task especially well they should be paid a
00:09:46.160
bonus not penalized see in the 1910s and 1890s they would penalize you if you did a task better than expected because that
00:09:52.279
meant you were sandbagging you're a person of low moral character that you've been doing it the slow way all this time you can imagine the kind of
00:09:58.279
workplaces that that gendered And yet when you the more you read about craft production about the kind of production
00:10:03.320
that existed before Fred Taylor the more it sounds suspiciously like the way we manage programmers today and I think we
00:10:10.519
are basically living in the craft production the pre-scientific age of Entrepreneurship and that is on the
00:10:16.560
brink of changing and you guys are part of that which we'll talk about uh so what I to I would talk about is that
00:10:22.720
entrepreneurship is management I know it sounds strange but it's not the general management of Fred Taylor it is a new
00:10:28.040
kind of entrepreneurial management management a management discipline specifically for that context of extreme uncertainty I mentioned before uh and
00:10:35.079
the most important entrepreneurial management concept is the pivot I'm sorry for those of you who are sick and tired of hearing about pivots already
00:10:42.000
it's kind of gotten out of control I saw this in the New Yorker magazine the other day I'm not leaving you I'm
00:10:47.120
pivoting to another man I'm sorry I never intended I didn't
00:10:52.440
imagine you know none of us in the Lean Startup movement thought this would become jargon but the reason this such an important concept is that this is
00:10:58.480
what true story stories of Entrepreneurship looks like remember that in Ghostbusters you know they're practically on their last dollar they're
00:11:05.480
about to go bankrupt when their first customer finally comes in and you think about it if Zuul decides to show up in New York a year early or a year later
00:11:12.200
there's no Ghostbusters they're out of business right that's actually what usually happens in entrepreneurship and we need to structure our company so that
00:11:18.720
that wouldn't be so devastating for the city of Manhattan if you look at the real story you'll discover this weird zigzaggy path
00:11:25.680
between the initial idea and the successful idea you ask yourself did success uccessful entrepreneurs have better ideas than unsuccessful
00:11:31.680
entrepreneurs I say the answer is no they're equally crazy it's just that successful entrepreneurs when they run
00:11:37.160
into difficulty they don't just give up and go home but neither do they persevere the plane straight into the
00:11:42.519
ground they do this thing called the pivot they change some elements of the business while keeping others constant
00:11:47.600
they keep maybe a new strategy for achieving the same vision so you pivot the strategy you stay true to the vision
00:11:53.839
and the premise of The Lean Startup is this if we can reduce the time between pivots we can increase increase our odds
00:11:59.760
of success before we run out of money because what matters in a startup Runway is not how many months of burn you have
00:12:05.680
left it's how many pivots do you have left and if we can learn to Pivot faster we magically increase our Runway without
00:12:11.519
having to raise more money or maybe without having to raise any money at all so let's talk about validated
00:12:17.320
learning uh when I was trained as an engineer in Silicon Valley I was taught the waterfall methodology of software
00:12:22.560
development this may be familiar to some of you this is straight out of Fred Taylor I was taught it as the
00:12:27.600
manufacturing metaphor of software development you can imagine how pissed off I was when I found out that they don't use this in manufacturing anymore
00:12:34.519
so it's not clear to me why we're using it in software so I hope none of you are doing this but maybe you have a friend and the problem with traditional product
00:12:40.839
development waterfall style is this it allows us to achieve failure to
00:12:46.000
successfully execute a bad plan see if if we really believe that most startups are failing because they're
00:12:51.320
fundamentally building something that nobody wants then why are we so proud of having done that on time and on budget
00:12:57.000
you go to Startup board meetings what are we talking we're talking about milestones and plans and forecasts but plans and forecasts are tools that only
00:13:03.360
work if you have a long and stable operating history anyone here feel like the world is getting more stable every
00:13:08.839
day no tools and forecast plans and forecasts are no longer useful guides
00:13:13.920
especially for startups so what we need instead is something a little bit more adaptable now most of you are probably
00:13:19.880
thinking this this is uh agile product development I was a a devote of of extreme programming and I think one of
00:13:25.680
the reasons why rails has been so successful is that as a community uh you guys have have adopted a lot of
00:13:31.199
techniques out of this Playbook which are much better than the old waterfall thing and you've changed the unit of progress see in in uh in waterfall the
00:13:39.040
unit of progress is advancing the plan to the next stage that's what makes achieving failure possible you feel like you're making progress even though
00:13:44.920
you're busy building something nobody wants uh we've changed in agile the unit of progress to a line of working code
00:13:52.120
but when I was at IMVU that's the last company I founded we built this 3D avatar instant messaging add-on and the
00:13:58.160
idea was we would the technology virtual world and bring it to IM and customers would love it and it would spread virally throughout the IM networks it
00:14:05.120
had to be an add-on because everyone knows uh IM has Network effects and therefore you can't just bring a new
00:14:10.399
instant messaging Network to Market because the everyone's already on an instant messaging Network and so to switch to be high switching cost that to
00:14:15.680
bring their friends uh you know that's kind of standard strategic analysis and so we said all right the brilliant way to get around that is to build an add-on
00:14:22.000
and to make a long story short uh we shipped this product it was super buggy super low quality I I was really
00:14:27.480
embarrassed even to put it out there but we really didn't want to have make the mistake of building something nobody
00:14:32.720
wanted so we actually were charging people money for this crappy product from day one I was really nervous we finally shipped it and luckily nobody
00:14:39.040
used it so at least nobody found out how bad it was and we weren't embarrassed but then I started to have this really
00:14:45.560
nasty thought which is wait a second uh why what you know why did we do that why
00:14:50.680
were we arguing every day for months about which bugs we absolutely had to fix when customers wouldn't even discover that the product could crash
00:14:56.399
their computer uh and to make a long story short we had to Pivot the business away from the insta messaging add-on
00:15:01.519
towards a standalone Network and um my code that I had written got thrown away
00:15:07.320
thousands and thousands of lines so maybe you can sympathize with me for a moment okay because I had written all
00:15:14.079
the software and I got thrown away yet I had done it all agile I had good test coverage it was well factored if I do say so myself Etc and yet the good code
00:15:20.440
and the Bad Code all got thrown away equally and so then I was like gosh did the company need me at all like why was
00:15:26.759
I here well couldn't I have just been on a beach you know on vacation ding sipping a nice drink while the company went ahead and did this and I said no no
00:15:32.959
no uh if I hadn't been there if we hadn't built this thing and shipped it we wouldn't have learned this very important thing about customers namely
00:15:38.880
that they thought the instant messaging add-on concept was stupid and refused to use it so that was good to know I mean
00:15:44.720
that excellent I'm glad we found that out but I started to have this really nasty thought which is like wait a minute if our goal the last six months
00:15:51.000
was to learn this important thing about customers why did it take six months I mean we had supported like I don't know
00:15:56.440
12 different insta messaging networks for interoperability my question was like would the learning have been the same if we' supported only
00:16:02.040
six with half the code yeah only three with one quarter the code right what if if we' only supported one network would
00:16:07.920
we still have discovered that customers don't want to download our software of course and then I had this thought this really freaked me out I was like hold on
00:16:14.120
what if we hadn't built any software at all we just created one single web page that said here's the screenshot of this
00:16:20.399
product we think we're going to build would you like to download it would we even have had to create page two where
00:16:25.720
we apologize the product's not available or would a 404 have been efficient 404
00:16:31.040
would have been fine why because they didn't want to download the software that means they wouldn't click the download button and I was really
00:16:36.600
depressed I mean I was just wait hold on how is it possible my my business card says Chief technology officer and how is
00:16:42.440
it possible that my work for the last six months could have the same value as a three-hour web page that a designer
00:16:48.240
whips up that doesn't seem right to me uh and that was the beginning of a journey that led me to Lean Startup into
00:16:53.440
this concept called validated learning that what really matters in a startup is not how much stuff we build but if we
00:16:59.720
can learn how to build a sustainable business uh here is customer development and Agile development put together into
00:17:05.600
this companywide feedback loop of learning and Discovery simplified it looks like this a software company is
00:17:12.679
nothing more than a catalyst that turns ideas into code when customers interact with that
00:17:18.480
code then we can collect some data qualitative and quantitative about what they think about what they do about how
00:17:24.799
it works and then we can choose to learn impacting our next set of ideas it's pretty simple three stage feedback loop build measure learn we call it and yet
00:17:32.200
as technologists if we're going to do entrepreneurship we have to get serious about the simple heris that our goal
00:17:38.240
should be to minimize the total time through this feedback loop and this is where I come to my point about why rails
00:17:44.200
makes startups because what matters in a technology platform for entrepreneurship
00:17:49.480
specifically is not how scalable it is it's not how much fun it is to write code in it although I hear Ruby is very
00:17:55.799
fun it's not uh you know how many many different people you can have concurrently working on it it's not does
00:18:01.600
it have a really beautiful syntax it's not even is it very writable or is it very readable actually all that matters
00:18:07.280
is how flexible is the platform and again not for just making stuff building to the specification how flexible is it
00:18:13.320
for our ability to learn from customers learn what is working and what isn't and
00:18:19.720
it's not just the technology itself it's the technology and the community together that make a platform what it is
00:18:27.400
and if you have a community that Embraces not just making cool technology not just having the kind of libraries
00:18:33.360
that allow you to test new ideas very quickly not just ideas from Agile development that allow you to build your
00:18:38.760
software with higher quality with better organization and better factoring but that Embraces the entire project of very
00:18:45.440
quick prototyping learning testing reacting you can get through this
00:18:51.039
feedback loop faster than anybody else and the reason I believe that startups are using rails is not because it's better technically but because it's
00:18:57.280
better at this and as technologist this is the question we have to answer for entrepreneurship
00:19:03.559
or like I said anybody who finds themselves in a in a position of facing extreme uncertainty how can we become as
00:19:09.280
adaptable learn as quickly as possible and so if you talk to enough entrepreneurs if you read enough
00:19:14.320
entrepreneural magazines you know you just you'll start to see this really weird pattern startup advice is
00:19:19.600
ridiculously contradictory it's like listen everybody knows that you know really good design is critical to the
00:19:26.080
success in entrepreneurs Steve Jobs would never ship a product that didn't have great design except if you're eBay or Craigslist or Myspace uh when they
00:19:33.159
didn't have really good design because actually it's more important that you get your thing out there real quick and have customers use it but make sure it's
00:19:38.679
really scalable otherwise you'll be the next frster uh unless scalability actually doesn't matter that much in which case don't worry about it because
00:19:44.679
Facebook was able to scale up later and you know make sure that you listen to your customers unless you're doing
00:19:49.760
Newsfeed and customers don't know what they want which case definitely don't listen to your customers so you should definitely have good design but not good
00:19:55.000
design be scalable but not too scalable make sure you listen to your customers and also ignore your customers good luck
00:20:00.760
uh and as an entrepreneur that kind of stuff used to drive me crazy because it's like well which which of you should I listen to We should have a board meeting with advisers and they would all
00:20:07.000
everyone give us their opinion and sometimes our you know our CEO would be like okay CTO you heard all that feedback now go implement it I'm like
00:20:14.039
but but that was like four four ways contradictory he's like that's your problem not my problem uh and welcome to
00:20:19.880
being an entrepreneur right if if people give you contradictory feedback that's your problem you don't get a gold star for listening to customers you don't get
00:20:26.039
a gold star for Building Product you don't get a good star for making money necessarily my belief is you only get a
00:20:32.760
good star for results and the results that matter are not hey can you make a lot of money today but can you figure
00:20:39.280
out how to build a sustainable business for the long term now there's a lot more to Lean Startup there's all these
00:20:45.000
techniques you can read about uh probably the most controversial in our Pantheon is something called continuous deployment uh at my last company mvw and
00:20:52.080
a lot of other companies now we're putting software into production like 50 times a day on average the lights are
00:20:57.559
too bright so I can't see I know there's always a few Engineers that start shaking their head at this point they're like that does not seem like a good idea
00:21:02.799
you know right think of all things that could go wrong just indulge me for a second you know like uh someone some random engineer could just take the site
00:21:08.320
down we used to let like QA people designers and marketing people anyone who wanted could have their own sandbox and Chip code to production anytime they
00:21:14.200
wanted and our point of view was listen if it's so easy to take our site down that even a designer even a marketing person can do it then shame on us but I
00:21:21.159
hear that doesn't seem like a good idea you know what if what if an engineer some random person regresses an old bug by accident or my favorite takes the
00:21:28.080
checkout button out of the e-commerce flow so now your uh your website you know your business has become a
00:21:34.000
hobby not actually that fun that kind of thing can happen but continuous deployment and all of these techniques are discipline methods for avoiding
00:21:41.159
those kinds of problems so that then your team can operate with maximum speed and maximum courage because you all know
00:21:47.400
there's tons of features out there that take longer to prioritize than they do to build you know what I'm talking about
00:21:53.840
right and so for those features instead of arguing all day about whether to build them or not let's just do an
00:21:59.000
experiment to get ourselves through this feedback loop quickly and discover if we're right or not and then we have
00:22:04.640
facts then we can actually do science not astrology it's pretty exciting uh and just to close my example about
00:22:10.120
continuous deployment you know sure it's pretty easy to have functional tests that you know detect if the button is missing something like that just try
00:22:16.240
this on for size instead of as our little like April Fool's joke instead of taking the button out of the e-commerce
00:22:21.440
flow we'll just change it to be white on a white background anyone ever had that bug you know the button still there all
00:22:26.640
a functional test pass it works fine just no human beings can see it so business equals hobby uh at mvw if you
00:22:33.120
went back in time to my desk and we actually did that prank check it into my sandbox uh in the next 20 minutes we
00:22:39.559
would get an email back it' be like dear Eric this is your cluster speaking thank you so much for trying to check in change
00:22:45.720
1,421 uh turns out that that's a really bad idea so we've automatically reverted that change we've automatically notified
00:22:52.679
the team that this happens so so much for my prank and also we've shut down the line This is an idea right from
00:22:57.760
Toyota production system you know the famous andon cord we made it impossible for anybody else to check in or deploy till a human being gets to the root
00:23:04.480
cause of what went wrong so it's not a very good prank what it means is that we're building products
00:23:10.080
that have an immune system that can automatically detect the most boneheaded and stupid mistakes so that we can clear
00:23:16.360
out our fear and act faster now in regular engineering most people are like ah who cares yeah okay so that might be
00:23:23.080
marginally better but you know it's more efficient for me as an engineer I can build faster if I don't have to waste
00:23:28.600
time and getting all this feedback from customers Engineering in stealth mode is the most fun we all know that engineering is the most fun if it wasn't
00:23:34.799
for those damn pesky customers always getting in the way right so if we can clear out all that building and measuring and learn if we just just
00:23:40.840
build we can be more efficient in the same way that uh if we want to drive our car as quickly as possible the fastest way is just to close your eyes and floor
00:23:47.760
the accelerator uh if you want to get the car in motion fastest that is so I want to mention one last thing which is
00:23:54.159
innovation accounting I'm not going to go into this in great detail but um we can talk about it in the Q&A if you guys are
00:24:00.200
interested instead of doing product milestones we need to use learning milestones and so instead of saying hey
00:24:06.520
here's a specification let's go build it out as quickly as possible we need to work backwards from what are we trying to learn and what is the minimum viable
00:24:13.559
product to learn that thing that way we can establish real Baseline data about the I was about to say crap but really
00:24:20.360
the crap it's in our business plan right we all have a business plan it has a spreadsheet in the back and it says you
00:24:25.520
know in some tiny little cell in twoo font it says 10 % of customers who see our sign up page will actually sign up
00:24:31.799
that's a leap of faith assumption that should be like printed out in giant bold red letters in the business plan like by the way if 10% of customers won't sign
00:24:38.320
up then we have no business but we we like to hide it and buried in the spreadsheet so let's make those assumptions really explicit and then
00:24:44.880
let's go find out what's the truth of that right now how many customers right now will sign up for what we're building
00:24:51.120
and if it's 0% if it's 5% if it's 50% it doesn't matter getting the truth is more important than having good news and then
00:24:57.440
once we have that base established then we can try to tune the engine to make changes to our product and actually find
00:25:03.559
out if they're making the product Better or Worse most of the startups that I have worked with most of the product development teams I know most of the
00:25:10.279
features that they add to their products actually make it worse not better so we can stop doing that and of course the
00:25:17.159
most important thing if it's not working if we're getting diminishing returns on those efforts then we can decide that we
00:25:22.279
either need to Pivot or persevere you can schedule that meeting in advance we can say instead of saying well we'll
00:25:27.360
meet about it we'll decide to think about a pivot only after we fail or if we fail right if we fail how about we
00:25:33.600
say look we're probably going to get some things wrong so let's decide 12 weeks from now let's have a meeting at that meeting let's talk about whether
00:25:40.039
we're going to Pivot or whether we're going to persevere so I've left a number of questions unanswered like how do we
00:25:45.080
know exactly when to Pivot the relationship between Vision strategy and product what exactly should we measure
00:25:51.200
uh how is it that products grow how do we know if we're creating value or just building a great Ponzi scheme what
00:25:56.360
specifically should go in the minimum viable product and of course as we scale can we go faster instead of suffering
00:26:01.640
the inevitable fate of bureaucracy um I've left those questions unanswered because I hear there's this really great book coming out uh really soon of course
00:26:08.360
you should all pre-order it as soon as possible thank you all very very much appreciate it thank
00:26:19.679
you I have the world's most overqualified Mike Handler been in chat I've volunteered to do that so please I
00:26:25.559
I can't see because of the lights but I if there's questions please feel free to come up and ask them while they're doing that I'll just mention we have a new
00:26:31.240
Lean Startup website up at the lean startup.com and of course we're doing startup lessons learning the conference uh this coming Monday May 23rd if any of
00:26:38.039
you are in San Francisco you're cordially invited we'll also be simal casting uh in more than 100 cities around the world so wherever you are
00:26:44.000
it's very likely that you can attend for free so do we have any questions one hey
00:26:49.679
Eric um my question to you is this it seems that um people have taken your idea and sort of perverted it in in a
00:26:56.600
way that's so it seems like there is no design uh there is no upfront uh planning uh there is no business plan
00:27:03.279
it's sort of like kind of we we're just going to hack away and you know pivot 15 times and guess what there's there's VCS
00:27:10.480
who are willing to fund that uh right now uh what do you say to that yeah thank you for that question first of all
00:27:15.679
there's VC's ready to fund anything right now let's just be honest okay uh and just you know those of you who who
00:27:21.120
like um who like uh George R or Martin will know what I mean when I say that winter is coming so so be ready
00:27:29.760
um but yeah this kind of perversion look every idea uh gets misunderstood I I think I even have this
00:27:35.679
this graph I I get this question a lot so I have a graph ready for everyone know the Gartner hype cycle uh you know
00:27:41.399
when something happens we reach the peak of inflated expectations uh and then we hit the trough of disillusionment when
00:27:46.440
it find when we find out that it won't like wash my car and make me live forever and give me quantum teleportation so uh we're kind of in
00:27:51.720
this Peak at the moment around Lean Startup and uh I mentioned that because from my point of view being
00:27:57.320
misunderstood is a huge step up from being ignored so I really appreciate that they were achieved that level but I
00:28:02.360
think we can of course do better and this idea that uh you know that it doesn't matter if you have design it
00:28:08.600
doesn't matter having a business plan I just think that's really wrong I hope from this talk you can get a sense of
00:28:13.880
why if your plan is to ship it and see what happens then you are guaranteed to succeed at seeing what happens but then
00:28:22.399
what right as soon as you have three customers you already have five opinions about what to do next uh and the worst
00:28:28.279
thing that can happen to you actually if you ship it and like everything blows up and nobody uses it like what would happen to me at inie that actually is a
00:28:33.519
blessing the worst possible fate is you ship it and you have a little bit of success you have some customers and you
00:28:40.399
have a 4% sign up rate and it's like well is 4% good or bad and then you work on it some more and you get the sign up
00:28:45.440
rate from 4% to 4.1% you're like Hey we're making progress and it goes up to 4.15% and then 4.17% I meet teams like
00:28:52.519
that that are really proud about every little uh increment and yet the problem is they have no idea if they're actually
00:28:59.200
making progress because maybe to make a successful business it needs to be 10% and getting from 4.1 to 4.11 is not
00:29:05.480
progress that actually means you're doomed or maybe 4% is great so you don't know how to evaluate what you do anyway
00:29:11.399
ship Andy what happens is not science right if if people did chemistry the way that we do entrepreneurship most
00:29:17.000
chemistry Labs would explode just like throwing stuff around to see what happens no Sciences make a prediction
00:29:22.240
about what's supposed to happen and then if it doesn't happen that sets you up to learn something really interesting so uh
00:29:29.000
I think the role of design business strategy um those topics are really important for helping us understand what
00:29:35.600
questions need answering what do we need to learn about as quickly as possible that's that's what I think we got time
00:29:42.399
for one more yeah right over here in the graphic you showed with the uh all the Web 2.0 companies that failed I was just
00:29:48.720
wondering do you know how many of those companies failed because they were just giving everything away from for free and
00:29:54.679
hoping for ad revenue and they just didn't actually you know they couldn't actually stand at that way or because they had a product that nobody wanted
00:30:01.320
yeah I mean I don't know the answer to your question specifically uh you know I haven't conducted in-depth survey about
00:30:07.600
it my sense is that it's not that nobody used their product I mean it's actually very rare to get exactly zero people the
00:30:14.240
real issue is that uh the model that is supposed to support your product doesn't work so for example people look at these
00:30:21.080
successful companies and they draw I think the wrong conclusion so you you look at something like Facebook which didn't charge people money from day one
00:30:27.360
you say ah that you know therefore don't charge money from day one but the idea that you just get big fast and kind of
00:30:32.840
hope for the best is kind of ridiculous and yet sometimes it works and there's a question well how come it worked for
00:30:38.200
Facebook but it didn't work for all these other companies and I think the answer is that uh there's a concept in the book I talk about called engines of
00:30:43.840
growth that a product which is designed to grow virally is really different than one that grows through paid advertising
00:30:49.600
is really different from one that grows through engagement or sticky what we call compounding uh interest and
00:30:55.039
depending on which engine of growth you're using the metrics that are critically important uh are different
00:31:00.559
and so you need to know what you're trying to accomplish in order to tell whether you're making progress I mean a lot of dotcom failures a lot of Web 2.0
00:31:06.639
failures tons of venture back failures that are quote unquote getting big fast are basically paying money to acquire
00:31:12.440
customers attention that they then plan to resell to advertisers so just glorified middlemen but that doesn't
00:31:18.279
make any sense if you're paying money for advertising you better be creating value for customers in a way that they're going to pay you more than that
00:31:24.880
and if you make more money per customer then you actually and then it cost you to acquire them then you can actually
00:31:30.039
grow otherwise you're doing that old grocery store joke of losing money per item but making it up in volume uh it's
00:31:35.720
ridiculous I think that's the problem we've been having uh certainly in Silicon Valley certainly during summer when money is Flowing you know people
00:31:41.840
feel like they can uh they can invest in anything and it'll work out you know we all know what's coming uh that's why we
00:31:47.480
need Innovation counting that's why we need a me a mechanism for telling whether we're on the brink of success or actually on the brink of
00:31:55.559
failure hi I was uh wondering if you had any tips on continuous deployment uh
00:32:01.760
when it comes to bootstrapping it and not necessarily having enough users to do accurate split
00:32:08.559
tests yeah um having to having a pathetically small number of customers
00:32:14.440
is awesome they startups are you always in a rush to get tons and tons and tons of people in attention it's like no that's actually not an advantage having
00:32:21.120
a small number of customers means you can get to know each customer extremely well so you can actually learn a lot
00:32:27.159
more about them then those of us who have zillions of customers have to rely on all these summaries and charts and metrics that they're all designed to
00:32:33.320
hide the customer from us look so we can look at spreadsheets and I have to teach you know people who have that problem we have to do the saying in Lean Startup
00:32:39.120
that metrics are people too so like remember that there's human beings behind those numbers and and act accordingly if you have a pathetically
00:32:45.399
small number of customers the imperative for continuous deployment is much smaller because the consequences of failure are practically non-existent who
00:32:51.480
are you going to piss off you just said you had a pathetically small number of customers so instead of worrying about split testing and continu deployment all
00:32:56.559
this stuff focus on as Steve blank says getting out of the building and getting to know those customers extremely
00:33:05.360
well hello there you go yeah thanks for uh for your blog and everything it's been I've been an Avid Reader and in
00:33:12.039
fact I implemented a lot of the continuous deployment stuff in previous companies but the question I have is um
00:33:18.159
what advice would you give you're obviously speaking to a bunch of Engineers um and what advice would you
00:33:23.279
give to a bunch of Engineers who probably a lot of us have aspirations to be entrepreneurs and to have businesses or run some to you know to the pivots
00:33:32.000
are they always you know is it people is it DNA is it sometimes I mean you can get a bunch of engineers in the room and
00:33:37.360
you maybe we're pivoting on you know how to refactor code perhaps and what kind of advice would you give to surrounding
00:33:43.399
yourself maybe with a more balanced team or a group and what advice would you give to an engineer who's seeking that
00:33:49.639
now thank you for that question look this most people don't want to tell you this but balance is crap seriously I
00:33:56.240
know everyone's like make sure you have a business co-founder make sure you have exactly the right DNA in your team and they have the right astrological signs
00:34:01.960
and your balance and all that stuff and the reason we're so focused on that is that all we have today the only data we have is we look around at the companies
00:34:08.280
that are successful and we see that balance but for every example of a balance team I can find you an unbalance
00:34:14.000
team that did just as well so I think as Engineers we shouldn't be ashamed of the fact that we have a certain mindset you
00:34:19.879
guys know what I'm talking about other people have names for it that are not as nice some of them are clinical but you know we are who we are and as an
00:34:26.960
engineer I always felt like the thing that got me into entrepreneurship and into this the thing that really pissed me off was I kept Building Products that
00:34:34.119
nobody used okay I me just kept happening to me right I kept building Pro working on teams where we just had the smartest most talented people just
00:34:40.560
spending tons and tons of time and energy on stuff that didn't matter and as an engineer that offended my sense of
00:34:45.720
efficiency and just my sense of justice that that isn't right so what I would
00:34:50.800
say is you gotta start you got really think about whatever your job title when
00:34:56.280
you become an entrepreneur I don't care what it says on your business card really your job title is entrepreneur
00:35:01.359
okay and the definition of entrepreneur is person who does whatever it takes to be successful that's it so like whether you used to write code for a living or
00:35:07.480
do marketing for a living or whatever now your job is to do whatever it takes to be successful and if you can get really serious about saying how do I
00:35:14.560
know that what I'm working on right now actually matters to anybody right now I
00:35:20.560
hope it matters I think maybe it will then you can start to take those engineering defects and turn them into strengths right because we are
00:35:26.320
maniacally focused on things like efficiency some say a little bit too much but we're just used to thinking
00:35:31.359
about local efficiency like how good a coder am I because that's what we've all been promoted on that's what we get our raises for underwaterfall you can win an
00:35:37.920
award and you can be promoted for developing the amazing architecture for a product that never ships that's
00:35:43.760
ridiculous so in entrepreneurship we can't do that we have to get serious about how do I know in a scientific
00:35:49.400
sense how do I know that what I'm working on matters and if you can really get serious about that question then
00:35:55.280
Engineers actually are have a tremendous advantage over people who aren't used to thinking in a rigorous or disciplined
00:36:01.079
way not to say that other functions you know you guys know what I mean um I'll just say one last thing if you look at
00:36:07.359
the history of management every past management Revolution was led by Engineers okay the famous Alfred Sloan
00:36:14.640
who ran General Motors he was trained as a mechanical engineer he ran a ball bearing Factory which was bought by General Motors Fred Taylor who I showed
00:36:21.319
you before he was an engineer Henry Ford engineer taono who invented the Toyota production system engineer there's a
00:36:28.000
reason why management has always been revolutionized by engineers and that's because management fundamentally is
00:36:33.079
human systems engineering and that's what we're really good at so thank you all very much I really appreciate it
00:36:39.000
thanks for coming out and uh being in touch thank you