00:00welcome to the a 16z podcast today we
00:03have one of our hallway style
00:04conversations between Benedict Evans in
00:06Steven Sinofsky and they'll be
00:07discussing big companies in their
00:08platforms Google Apple Facebook and
00:11Amazon what makes them work how their
00:14tactics and strategies differ and how
00:15they've adapted to change over time good
00:18afternoon this is Steven Sinofsky and
00:20Benedict here we were having this
00:21hallway conversation about what is it
00:23that makes a big company unique and
00:25interesting to think about from an org
00:27perspective the thing about big
00:29companies that we wanted to start with
00:30is of course is that they work like
00:33they're not dysfunctional as a very
00:35nature and from the outside everything
00:37might look mixed up or people might be
00:40complaining but for the most part like
00:42they got to be big yeah so I thought the
00:45thing I thought was specifically
00:46interesting just kind of looking at the
00:48tech world today at Google Apple
00:50Facebook Amazon as being together about
00:5210 times the size in terms of the number
00:55of people that they were 10 12 years ago
00:57and how is it that they've ballooned in
01:00this size and have 10 times the people
01:02that they used to have not that long ago
01:03and yet continue for all intents and
01:06purposes to work well continue to create
01:08great products more or less how do those
01:09companies run and when you talk to
01:12people there what you realized very
01:13quickly is that they're actually running
01:14radically different ways from each other
01:16from each other and inside like even
01:19across them inside they're quite
01:20different yeah and that's partly related
01:22to the product that they have but it's
01:25also just kind of reflects kind of
01:26complete difference in philosophy the
01:28interesting way to think about it is
01:29that obviously we think of Google Apple
01:32Facebook hammers and as being platform
01:34companies but when we say that we're
01:35really thinking about the software and
01:36the network effect and the people and
01:38the chips and so on actually just as
01:40important is the platform of the people
01:42and the organization of those people and
01:44how that can be deployed right and of
01:45course very old saying in technology and
01:48a lot of studies have gone into this
01:49notion that when companies make products
01:51the product is a reflection of the
01:54organization and you know the flipside
01:56is we don't ship the org chart but the
01:59truth is is that the the org is as much
02:02a product of a company as the product we
02:04should caveat this whole thing by saying
02:05that you know whatever we say about any
02:08of the given companies is both right and
02:10probably runs counter to any one
02:13for me the other thing that's super
02:15important is that even in the the most
02:18well-run large organization it's
02:21actually never like one company it's not
02:23homogeneous it's not self consistent
02:26it's not a monolithic thing no matter
02:28what people say from the outside it's
02:31fun to point to and it's fun to try to
02:34draw an arrow from a product to that org
02:36chart or assign causal relationships but
02:39it's just it's not really really like
02:41that yeah there's always that complexity
02:43you have to think about like the
02:45internal dynamics of kind of chaos and
02:46entropy and you know personal politics
02:48and so on as well as like the overall
02:50strategy but there's also like another
02:52important distinction that each of these
02:55companies tend to have kind of a
02:56fundamental strategy but they also have
02:58a bunch of things going on inside that
02:59are more like tactics that support that
03:00strategy and the fun thing about looking
03:02at specifically Google Apple Facebook
03:04Amazon is that generally the stuff that
03:07you know any of the stuff that's a
03:09strategy for one company as tactics for
03:10other companies so yeah Apple strategies
03:13have an operating system and for Google
03:15that's just a tactic Google strategy is
03:18to do search for Apple that's just
03:19attaching and right on and you know and
03:21Amazon has searched and rule and uses
03:23Google ads but views it as a tactic to
03:25implement their store what's tricky it
03:27from the outside is that there are
03:29people whose pay really close attention
03:30to only one element or one part of that
03:32strategy and sometimes they think a big
03:35company is missing the boat because
03:36they're not all in on on doing battle in
03:39that space you know like like say Apple
03:42or I mean say Google and hardware like
03:44people are like are they gonna do it not
03:45do it and it's a tactic for them it's
03:47not a core core part of the strategy and
03:49that doesn't make it that mean that
03:51they're making a mistake doing it it
03:52means that they've made a decision about
03:54how they want to go about doing it but
03:55with classic thing that drives me crazy
03:57people talking about Apple thinking that
03:59the App Store is like a primary revenue
04:01driver for Apple when the whole purpose
04:02of the App Store is to drive sales of a
04:04$700 iPhone every two years at a 50%
04:06gross margin right well if you think
04:07about like just just your iCloud
04:09subscription like if Apple really really
04:12thought that that was the most important
04:14thing would they bury it like five
04:16levels deep in the UI to upgrade know if
04:19it was the most important thing every
04:20time you unyu remove your phone from
04:22sleep it would warn you that you need
04:24more storage the way that using a
04:26George is their primary business might
04:28but they also there's also within that
04:30actually you can see the kind of there
04:31is a tactic and the strategy because the
04:33grand strategy might say well why are
04:34you charging at all you should give it
04:35away for free but then the tactic you
04:37see well we can make for money day
04:38rights quite often the things you read
04:41about that say companies are messed up
04:42come from people who aren't happy with
04:45what's going on they might not agree
04:46with the strategy they might more likely
04:49just not agree with the implementation
04:51choices the choice of product the choice
04:53of code the order leadership choice and
04:56that causes them to sort of tell a story
04:58and you also get kind of its what it's
05:00not it's the opposite of survivor bus
05:01you get the UM you have the ex post
05:03facto description well why it was
05:05obvious that this company was going to
05:06fail is obvious who thought it was going
05:08to fail and actually if you go and look
05:10at now the kind of the descriptions of
05:12the genesis of the iPhone now like if
05:13the iPhone had failed you could tell all
05:14of the same stories to show why it
05:16wasn't going to work well you would have
05:17said like oh my god they they had the
05:19working thing with the iPod Group they
05:21didn't pick the iPod code they really
05:22should have gone with the thing that was
05:24already in market and already working
05:25and they made some big science project
05:27bet on port animal and of course Apple
05:29was going to collapse just look at the
05:30political conflicts that they had five
05:31years later well obviously that meant
05:33that the iPhone was going to go down in
05:34flames well yeah you can always come up
05:36with these things in hindsight the
05:38flipside is when when you read this
05:39glowing positive story about how a
05:41company worked they either when the
05:43negative stories make it look like a
05:44company is Fight Club or like twister in
05:46how it works and then the positives make
05:48it look like this orchestrated ballet or
05:50symphony when you know oh we had a small
05:53disagreement so we had a meeting we
05:54resolved it by evaluating the
05:56alternatives and everybody agreed and
05:58then we had tea and it was wonderful and
05:59then we rushed out the hub of business
06:00movies case study yes
06:01the one other thing I would say too is
06:02is often when people look at these
06:05things they were like looking for the
06:06one thing that they should do when they
06:08run their their company their org and
06:11they look to these successes the
06:12survivor bias to big companies and it's
06:14really important to keep in mind that
06:15you you can't really take one thing that
06:18a company does and then extract from
06:20that like oh this is how we're gonna
06:21operate and it's very much I always I
06:23wanted to do in the gratuitous Star Trek
06:24analogy and it's from the episode of the
06:26original series squire of gothos the
06:28episode is everybody knows that the
06:30squire of gothos is out there and he's
06:32got some fancy radar sonar thing and he
06:34can see what earth is like hundreds of
06:35years ago and he replicates it for the
06:38for the landing crew but
06:40all of a sudden they taste some of the
06:41food and it doesn't taste right and the
06:44reason doesn't taste right is because
06:45his sensors don't pick up taste so he
06:48could fake what it looks like but he he
06:51doesn't know what it really should be
06:53and that's a lot about like you know you
06:55will well you know Amazon is famous for
06:57three page memos that doesn't mean
06:59everybody should go right three page
07:00memos and then you'll be a half a
07:01trillion dollar company
07:02it doesn't quite follow this Apple is
07:04famous for secrecy it doesn't mean you
07:05don't tell anybody anything and then
07:07you'll become and the world's biggest
07:08consumer electronics company exactly so
07:10what are some of the things that comes
07:12to your head when you want to try to
07:13understand what's really important to a
07:15company for me like you first want to
07:17know like are they focused on api's or
07:19you know like features and experience
07:21yes I mean are they creating I'm a
07:23waterfall this is the Marcus Aurelius
07:24thing what is it that what is the thing
07:26in itself again you can get you kind of
07:28get misled by this so you know people
07:30look at Amazon Facebook and Google and
07:32think they're advertising companies I
07:33mean they're not am of Google is a
07:36company that's around handling vast
07:38amount amounts of data and everything
07:40that Google then does is sort of an
07:41expression of that Facebook is a product
07:44around connecting people and everything
07:45that it does has to be around that
07:47whereas Apple is a company around
07:49creating a very specific kind of
07:51hardware company that has like a five
07:53year product cycle between like three
07:55years of planning and then two years of
07:56it being in the market and so they need
07:58an organization structure that can
07:59handle that and deliver that kind of
08:01thing which is very different from
08:02Facebook be needing to be able to decide
08:04while this thing isn't working so we'll
08:05kill it and replace it with something
08:07else tomorrow and then Amazon I think
08:09perhaps is kind of most unique of these
08:11four companies in the deceased sort of
08:12radically decentralized almost atomized
08:14organization that you have this
08:16underlying platform of logistics in
08:17e-commerce but then everything else is
08:19like the one Pizza team all the tea
08:20Pete's a team so it's like makeup in
08:22France is three people shoes in Germany
08:23history in fact geography is interesting
08:25overlay on how they even approach that
08:27and for me like those descriptions are
08:29perfectly encapsulate this notion that
08:31the way that you're a five hundred
08:33billion dollar company is you have like
08:35a monstrous platform and and that used
08:38to be a really big word that Alfred
08:40Sloan created in thinking about Detroit
08:42and cars and the 50s you have a platform
08:45you have a car platform and then you put
08:46different bodies on brands yeah and you
08:48make an Oldsmobile in a Buick out of the
08:50same platform you look at that platform
08:52then you start to ask yourself questions
08:54like you know how did they decide are
08:56they focused on revenue are they focused
08:58on usage are they focused on enterprise
09:00customers or consumer companies you know
09:03there's a lot that goes into using the
09:05platform that's based on internal
09:06notions like are there requirements that
09:09you do things with the platform
09:10internally or or not and that's what
09:12drives the organization itself you know
09:15like Google is well known and we'll
09:17start with Google like they're well
09:18known for having this incredible
09:20environment that once you're when you're
09:22an engineer at Google you show up and
09:23you you like with in a heartbeat can tap
09:25in to like this massive build process
09:28this massive experimentation platform
09:30and all of this data Google is a
09:33platform for handling handling hard
09:35problems of scale for handling massive
09:37amounts of data at scale and it's
09:38relatively agnostic about exactly what
09:41those problems are it's quite relaxed
09:43about having like two or three different
09:44teams trying to solve that problem in
09:46different parts of the company and then
09:47things within that then become tactics
09:49the real point is the data and the
09:51solving that problem for customers right
09:53and fundamentally from a platform of
09:56people point of view Google and this is
09:58where this the subtlety and nuance
09:59really needs to be figured into this
10:01conversation which is Google is is not
10:03one organization it's one type of
10:06culture around billions of scale scale
10:09on the billions super-hard problems but
10:12they really are like an enterprise
10:14company focused on search you know like
10:16that's built in there and they have a
10:17product cycle that looks like an
10:18enterprise company a salesforce that
10:20looks like an enterprise company they
10:22treat changes to the product the way
10:24that an enterprise company would treat
10:25changes to the product yeah and I
10:27mentioned the fallacy earlier of people
10:29thinking that the revenue from the app
10:30store matters to a play there's an
10:31equivalent one which is people say
10:33Google doesn't make any money from
10:34Android well if there's not the point
10:35that's not like that's not what in any
10:36sense why and what exists right it's a
10:38tactic to support those underlying
10:39platforms and and the companies will ebb
10:41and flow from a management perspective
10:43about how much experimentation and in
10:46that regard they they can financially
10:47support or maybe they need more
10:50resources in another place and the fact
10:52that those connections are loose
10:53actually make those decisions easier
10:55because they don't you know so it's a
10:57smart management strategy to not goal
10:59everybody on everything because the
11:01minute you do that then your your your
11:03linear equations are too hard to move
11:06having flexibility is both good for the
11:08employees and good for the for the
11:10company and it's so it is how you end up
11:12with like something that looks
11:13strategically not so smart like multiple
11:16messaging clients and but from a tactic
11:18point of view that's that's not a big
11:20deal that's like having different
11:21marketing messages in France versus
11:24like it's okay I think that's an
11:25interesting way on messaging to go into
11:27Facebook yeah yeah there's quite a lot
11:29of similarity between Facebook and
11:30Google in variability and their cloud
11:31companies and um but also that they have
11:34on one hand the enterprise Salesforce of
11:36course which was built by the same
11:38yeah so it's again an enterprise company
11:40and you have feit that the Facebook core
11:42product kind of changes at the velocity
11:44of an enterprise product more than you
11:46think of as as like a fast pace like
11:48constantly change that you've got the
11:50enterprise Salesforce and then you've
11:51got the kadhi you've got the data thing
11:52with a much narrower focus right it's
11:54not you know understand everything in
11:56any way to anybody it's work out the
11:58right ways of connecting people and I've
12:00always kind of thought about Facebook as
12:01surfing user behavior that it's very
12:03very good at thinking about how people's
12:05behavior changing and changing the
12:06product and so Facebook has been like
12:08five different things over the length of
12:10its life you know there was a time when
12:11it was all about filling in information
12:12on your home profile yeah who goes to a
12:14Facebook profile now right and that's in
12:16a sense been directed by Facebook right
12:18so the willingness kind of the key there
12:19is is to to sense which way the wind is
12:22blowing and to to move well users are
12:24going away user behavior was going on
12:26top of that like remember when when all
12:28of a sudden photos became really
12:29important and the idea was get people to
12:31post photos and all and then all of a
12:33sudden there was this kind of a like oh
12:35wait in order to get something to
12:37surface in the feed it had ahead of a
12:38photo and then they decided video was
12:40important but all within this same sort
12:43of very deliberate very sort of mature
12:45view of pace of change and these go
12:48counter to this notion of like that was
12:50the very early origin of move fast and
12:52break things and then then the other
12:54part of course entrant from an
12:55organizational perspective is that it
12:57very much like a very large-scale mature
13:00company they have other product lines
13:02you know between whatsapp and Instagram
13:04oculus and so on that are are run as
13:07very distinct entities and they don't
13:10overlay some grand strategy about what's
13:13gonna go where and who's gonna do what
13:15when and tie all the pieces together
13:16yeah I mean this there's a route for all
13:18analogy I think you're looking at
13:20like Microsoft Office there's stories in
13:22facebook and the stories in Instagram
13:24and the stories in whatsapp just as like
13:25there was a charting tool in and tables
13:27because of course you'd be able to do
13:28tables in a slide and of course you
13:29should be able to do tables in in Word
13:31should I there be exactly the same table
13:33tool well maybe at some point we'll get
13:35around to integrating them is that our
13:36absolutely highest priority right now no
13:38not really the key point is that that
13:40application for looser needs of its
13:41users and shapes what that particular
13:43product needs to be these are these
13:44things that if you over think from the
13:46outside and overanalyze you get all you
13:49bent out of shape that this doesn't all
13:50fit together and some neat puzzle
13:52because the products aren't clear when
13:54do I use whatsapp versus this and I
13:56guess but it doesn't who cares and like
13:58no person is making that choice they're
14:01just using the one that they're gonna
14:02use because of whatever network effects
14:04or whatever reason and that's where like
14:07it's fascinating for me because Facebook
14:08is the usage model is aimed at consumers
14:11they're not put under the sort of the
14:13industry analyst Enterprise Gartner
14:15report over like is this a charting tool
14:18or a numbers tool or a presentation
14:20graphics tool and then it can't have
14:22overlap and it needs to all fit
14:23strategically together and that's shows
14:26a real strength I think it's a huge
14:28organizational strength and maturity
14:29that they not just permit that to happen
14:32but their org is designed around making
14:35that an acceptable way to move forward
14:37it's at its the same thing it's like
14:39what's Marc Marc Andreessen phrases you
14:41know strong opinions look weakly held
14:43right back to what we were saying on you
14:45about Google that okay there is a core
14:46capability that we're going to do but
14:48we're kind of relaxed about having four
14:49different groups working on that in
14:50different ways in different places
14:51because you know each of those will
14:53solve things in a different way and it
14:54doesn't kind of matter that it doesn't
14:55all kind of tie up neatly in an org
14:57chart the external view is that that it
15:00needs to all fit together and the
15:01internal view just it just isn't a
15:03problem that's a tactic that's not very
15:05and and which is actually interesting
15:08because the internal external view
15:09definitely leads us to Apple where for
15:12Apple has this this very clear mode of
15:15working which is they have a strategy
15:17and they have a plan and they're an
15:19execution monster and they're just not
15:22gonna share it with anybody and that
15:23just drives people crazy yes like like
15:26this notion of how can you know all of
15:28this and then just not have billboards
15:30on 101 telling us what the iPhone 11 is
15:33like if we took cool stain I know well s
15:35and what's interesting to me about it is
15:36that it gets back to our earlier
15:38conversations on Twitter about leaks and
15:41people think that Apple is like it's
15:43oppressive from the outside it must be
15:45oppressive and you can't leak and it's
15:46like no actually if you talk to anybody
15:48about Apple they're pretty chill about
15:50the whole thing because they actually
15:52know they have a plan and they know they
15:54have a strategy and so they're not
15:56worried that it's all wrong they have
15:58confidence in their management structure
16:00the manage structure is appropriate ly
16:03translucent I like to say throughout the
16:05org and so people don't run around like
16:07leaking their their things and you know
16:09and when there is a leak often the
16:11people feel really really bad as a team
16:13they don't look to blame the person the
16:16outside world wants to find the culprit
16:18but internally they're quite mature
16:20about the whole thing but at this point
16:23that's like there's a deterministic
16:24element here which is you know Google
16:26and Facebook and producing stuff that
16:27gets deployed every hour so that creates
16:29a certain kind of structure Apple is
16:31producing something you know they the
16:32iPhone 8 for the sake of Arc for the
16:34whatever the name is is going to be
16:35announced in a month or so but that was
16:37being planned three years ago yeah and
16:39it's been baked in for basically at
16:40least a year now they pretty much
16:42decided what the next one is going to be
16:43and they're well into the one after that
16:44and so that produces a requirement for
16:47an organization that looks very
16:49different that's much more kind of
16:50systematic and methodical and structured
16:51because you know stuff has to be decided
16:53now in 50 million of them were going to
16:55go off in the ship in a year and a half
16:56yeah and also we should point out that
16:58like there are projects going on at
17:01Google or Facebook to reinvent big parts
17:04of what we see and those are gonna take
17:07years to deploy if Google wants to
17:09replace MapReduce with some new thing or
17:11if they're gonna roll out like the new
17:13way of doing machine learning internally
17:16they'll work on that for the same amount
17:18of time that Apple works on a phone or a
17:19file system or a file system it's just
17:22that they're not gonna release it little
17:23bits at a time and they're it's not
17:25gonna do press releases you know at
17:27every milestone we talked about the
17:29split inside Facebook and Google between
17:31like the enterprise sales very crudely
17:32the enterprise sales business and the
17:34stuff all around that that supports that
17:36and there's the enterprise sales product
17:37and then there's a consumer product and
17:39then there's the stuff that sits all
17:40around this kind of pushing it driving
17:42it the thing at Apple is there is like
17:45the as it might be the Gantt
17:47Tippett use of specific products that
17:49are going to drop in a certain period of
17:50time but there's also the platform there
17:53of the people and the operating of both
17:56the operating system and the chip but
17:57the the software people and the chip
18:00people and the capability to point all
18:02of those people and say we're going to
18:03solve that thing now yeah and so just as
18:06we're Google could say like machine
18:08learning is a thing now let's point out
18:09at photos and soul photos in a fairly
18:13short word yeah and probably it was it
18:14was literally less than a year I think
18:16that that whole project I've spent
18:18equally you know Apple has taken that
18:20team and pointed it at the echo and said
18:23make an echo that's athlete again this
18:25is this part of a platform so they have
18:27the platform of the people then know how
18:30to work their platform of code and
18:33product and what Apple is able to do is
18:35the bet that they've made is that we
18:37have a platform that's made up of a set
18:40of API is a UNIX operating system kernel
18:42these arm chips that have all of these
18:44sensors and capabilities and it turns
18:45out the thing that they do uniquely is
18:47they can take that platform and then
18:50take the 80% of it that they need for a
18:52watch or a speaker or a television set
18:54or whatever and then what's unique about
18:56how they approach it is that they take
18:58the 20% that might not be viewed as
19:00super hard or super difficult to get
19:02done and they actually spend the massive
19:05amount of time and energy to do the
19:07whole part of what it takes to be a
19:10watch and I think that that's an
19:11interesting contrast say with Android
19:13wear they also had the platform but they
19:16had a different approach which was like
19:18well that's not the necessarily the hard
19:19part and the ecosystem will round it out
19:21and people will try different things and
19:23so you get a different kind of product
19:25out of that and that is what people
19:27think of as the end-to-end engineering
19:30of Apple but it's also an org commitment
19:32that they make like they tell the team
19:34and they get the team to go finish the
19:36whole of the product which is a unique
19:39way of doing it well it's a specific
19:40culture but it's also the capability
19:42yeah you have the chip people and you
19:44have the designers and you have union
19:47right down to the stores you have an
19:49ability to create a certain kind of
19:50product and you can take that ability
19:52and point it at things yeah you can
19:54point it at a watch you can point it at
19:55home speaker you can point it at TV
19:58there's stuff that you pointed out there
20:01where you kind of try and work out what
20:02it would be and you know clearly will
20:03the Apple right now is kind of trying to
20:05work out what it would mean to do a car
20:06right and that's what's so much about
20:08that is is how you know that capability
20:11isn't universal and so a huge part of
20:13the culture of a company is making sure
20:15that they choose the right thing to
20:17point it out what you know back in the
20:1960s when when Ford had the Mustang you
20:21know every Detroit company had a
20:23platform of four wheels and a
20:25high-performance engine to make a car
20:28but GM just could not figure out how to
20:31respond to the Mustang and they were in
20:34a panic and and so figuring out once you
20:37have the underpinnings it's also a huge
20:38management task and that's the
20:40discipline that I think that Tim Cook
20:42talks about is figuring out what to say
20:43no to what to say yes to I wrote a piece
20:45a couple of years ago kind of comparing
20:47out Google to the story about shot how a
20:49shark will bite something to see if it's
20:51a seal and if it's not it'll spit it out
20:53it's our Google bytes thing to see if
20:54they fit Google and so like they tried
20:56healthcare and it just didn't work and
20:57they would have had to have changed what
20:58Google was to much in order to make
21:00healthcare works which is super
21:01disciplined and they tried radio like
21:02ten year over ten years ago they bought
21:04a radio advertising company they worked
21:05out that they could not make radio work
21:06without changing what Google was so they
21:08didn't do it and that's exactly the same
21:10as Apple saying no to a product because
21:12it would have to make them not Apple in
21:14order to make that work each of these
21:16platforms have kind of there were places
21:18where they were extremely good at giving
21:19that kind of product and maybe the other
21:22companies in this set might be very bad
21:23at doing that but you know the stuff
21:24that Apple would be really bad at making
21:26that thing and Google would be really
21:28bad at making that thing and Facebook
21:29would be really bad at making that thing
21:30and part of the strength isn't actually
21:32knowing that and pulling back and saying
21:33no we're gonna stay in our canal and we
21:35have Amazon but seriously as they might
21:37not worry too much if they're really
21:38good at it or not really good they seem
21:40to be very brave about like let's make
21:43phones let's make readers let's make
21:44speakers yes voice recognition I think
21:48I'd see I'd see it as sort of the
21:49structural nice to kind of at a right
21:50angle to that so I think I'd kind of my
21:52sort of elevator description of Amazon
21:54is that you have two fundamental
21:56platforms you have the physical
21:57logistics platform in the e-commerce
21:59platform you could argue now they're
22:00adding a third in the form of crime
22:01which is slightly different and then
22:04everything else sits on top of that and
22:05is radically decentralized so everything
22:07is a team as a team of like three people
22:09so shoes in Germany of three people and
22:11makeup in France is two peers for people
22:13music players will pick a category is a
22:16small number of people and it's
22:17completely atomized and everyone has
22:19internal transparency as to their
22:20metrics and the point of that is that
22:22you can kind of scale that almost
22:24indefinitely if you want to add more
22:26categories you don't need to add another
22:27direct report jeff bezos it's not it's
22:29not a tree structure right and
22:31importantly neither is Amazon itself
22:33like especially in fact search has made
22:36this such an interesting thing because
22:37if you find everything by search and
22:39this was the difference between Google
22:40and Yahoo with Yahoo was a hierarchy and
22:42you have to enter Yahoo from the top and
22:44then sort of navigate and once you do
22:45search you don't need the top of the
22:47Internet and Amazon doesn't need the top
22:50like you they can sell you anything
22:52because you're just gonna have to look
22:54for and they have to get good at doing
22:55that and so they can really scale what
22:58they offer like fairly infinitely and
23:00what so what's so fascinating about that
23:03is there they also nothing has to be a
23:07like it it they can decide to do
23:09something or not do something and work
23:11on it for a while and then decide if
23:13that's delivering the customer
23:15satisfaction that they want or the ROI
23:17that they want or the connection to
23:18other products that they want and and
23:21yeah and that means on the one hand that
23:25they can scale almost indefinitely by
23:27country and buy into new product
23:29categories you know it raises a question
23:32okay are there product categories where
23:34that don't fit that commodity more well
23:35and that's whatever he's asking about
23:36Whole Foods and groceries yeah you know
23:38I can also write well they put the
23:40question right now is fashion can you
23:42say fashion and the challenge is not for
23:44so to speak can you create a nice web
23:46experience that would sell fashion
23:47because clearly there are companies that
23:49are doing that it's how would Amazon
23:50adjust that operating model such it
23:52because right now there's like sweet
23:53thousand teams in there who are selling
23:55different products so can now teams can
23:57start asking for their own UI ROH okay
23:59how would that work that would be any
24:00kind of an interesting organizational
24:02challenge right if the fashion people
24:04want a pioneer like a virtual reality
24:05try things on or or even logistically
24:08they want to you know that you you know
24:10they deliver it and then you have five
24:12hours to try it on and then the car will
24:14become might pick it up and take it if
24:15it doesn't fit like they might want to
24:16do that but that's a whole new logistics
24:19offering for them yes oh it needs to be
24:21packed differently and and that's where
24:23like this notion of all all four of
24:25these companies Google Facebook Apple
24:27the management is is all about finding
24:29the products that really fit that they
24:31can make well and of course the the risk
24:33is either that they pick something that
24:35they can't make well and they make it a
24:36really big bet or they they fail to be
24:40able to change that underlying platform
24:41when it really needs to change and you
24:44know one of the more brilliant things
24:45that Facebook and Google both did was
24:47move to mobile yes you know and that's
24:49like that's a huge platform shift and
24:51then now moving to machine money yeah in
24:52a sense they're rebuilding from a
24:54consumer and this is from a consumer
24:56point of view almost don't notice it
24:58right though they rebuild all of their
25:00stuff around machine learning and now
25:01you get like specific tangible things I
25:03now you can search inside photos or not
25:04but very often let's just like you okay
25:06you're now your google search you're at
25:07the same Google search you did five
25:08years ago it's just now suddenly it's
25:1015% better because of machine learning
25:11well and you see that like Amazon you
25:14know it always used to be people who are
25:15interested in this also bought this and
25:17most of that was sort of based on very
25:19straightforward query results of their
25:21purchase history yeah but now they're
25:23they are gonna have the technology using
25:25machine learning to start offering
25:26recommendations that are based on a lot
25:28deeper insight based on on these
25:30learning techniques this was kind of the
25:32challenge of mobile was it completely
25:34changed operating environment and so for
25:36Google and for Facebook there was his
25:38real question like actually our whole
25:39platform needs to be pointed on a
25:41completely different thing and I think
25:43maybe there's like another strand here
25:45three of these four companies are still
25:46one effectively run by the founder each
25:49of them has seen the last 30 years of
25:51tech and say they saw what happened to
25:55IBM they saw what happened to Microsoft
25:57they saw what happened to to Yahoo and
25:59to AOL and to MySpace and say they've
26:01sort of seen all of the lessons of what
26:03worked and what didn't work and what can
26:05happen to you and they were all you know
26:07still super aggressive and quite young
26:09and like looking at that and a very
26:13determined to kind of take this platform
26:14and keep moving it I can't really
26:16overstate that kind of observation
26:19because that was one of the things that
26:21I thought was most fascinating very
26:23early on with not just the YouTube
26:26acquisition but the integration strategy
26:28of YouTube because historically in the
26:31software industry acquisitions were were
26:33done and then assimilated as quickly as
26:36and I remember watching YouTube and
26:38going when are they gonna be able to
26:39sign on with a gmail account
26:41they took them three or four years
26:42before you could do that and then you
26:44watch Facebook and you watch how that
26:46they've dealt with whatsapp and
26:47Instagram and they've kept them at this
26:49artists liked it and that's like a very
26:51clear lesson from how to deal with what
26:54we today know our disruptive
26:56technologies a book that was just you
26:58know written in 1998 you know I mean you
27:00could argue that the program one of the
27:02problems with disruption is it doesn't
27:03work if everyone's read the book right
27:05on the other hand what we're really
27:06saying is just like in the finance world
27:08like there will be something an event
27:10that happens it's just that it won't be
27:11like they won't get disrupted by some
27:13technology paradigm shift that'll be
27:15some other challenge that they face
27:16because I think that's something that
27:18you'd really need to internalize about
27:19these four companies is that they have
27:21built platforms strategies and
27:24organizations that understand how to
27:26deal with changes in technology in a way
27:29that past generation companies just had
27:31challenges with the lesson is how
27:34flexible these companies are it's sort
27:37of pointing themselves at a different
27:38problem it's not actually the specific
27:41product that you're holding in your hand
27:42or the chip or the you know the
27:43particular piece of software it's the
27:45ability of that company to kind of pick
27:47itself up and point it somewhere else I
27:49mean we have this kind of this storage
27:50from the past if you know the Bill Gates
27:51memo about the internet and the kind of
27:53Microsoft picked it up and pointed
27:54itself at something else and this is
27:55like this big case study of oh my god
27:57this is an amazing thing that happened
27:58like how many times has Google and how
27:59Google and Facebook done that in the
28:01last ten years yeah in a sense how many
28:02times even as Apple done that but save
28:04maneuvered around you know we'll see
28:05they have a longer product cycle how
28:06many times as Amazon reoriented or
28:08itself around different things it's like
28:09actually a substructure all change of
28:11these companies being able to react to
28:13change differently I'm her early early
28:14on in 1995 you know Jeff Bezos liked
28:17doing books and telling people locally
28:20in Seattle like it small dinners and
28:22stuff like well we're definitely only
28:23doing books because he had this very
28:25special understanding of book
28:26distributors and how they would you know
28:28not have warehouses like we're not gonna
28:30do warehouse and we're definitely not
28:31going to do CDs and DVDs and movies and
28:34then you know a year later they're the
28:36largest DVD seller they have they're
28:38starting to build giant warehouses
28:39they're stocking the top twenty-five
28:41books and he reinvented the company that
28:44he had and that was a startup pivot but
28:46then you think about how like we are
28:48never gonna deliver groceries and now
28:50here we are what we tried to do with
28:52this podcast was just wrap up a
28:54we've had about just how these big
28:56companies work and we see lots of things
28:59that are pretty interesting from an org
29:01perspective and a lot of lessons for how
29:03founders should think about about how
29:04they organize adapt to change and also
29:06work with these companies so this has
29:08been Steven Sinofsky been at heavens on