How data will transform business
-
0:01 - 0:03I'm going to talk a little bit about strategy
-
0:03 - 0:07and its relationship with technology.
-
0:07 - 0:10We tend to think of business strategy
-
0:10 - 0:12as being a rather abstract body
-
0:12 - 0:13of essentially economic thought,
-
0:13 - 0:15perhaps rather timeless.
-
0:15 - 0:17I'm going to argue that, in fact,
-
0:17 - 0:20business strategy has always been premised
-
0:20 - 0:22on assumptions about technology,
-
0:22 - 0:24that those assumptions are changing,
-
0:24 - 0:27and, in fact, changing quite dramatically,
-
0:27 - 0:30and that therefore what that will drive us to
-
0:30 - 0:33is a different concept of what we mean
-
0:33 - 0:35by business strategy.
-
0:35 - 0:37Let me start, if I may,
-
0:37 - 0:39with a little bit of history.
-
0:39 - 0:41The idea of strategy in business
-
0:41 - 0:44owes its origins to two intellectual giants:
-
0:44 - 0:46Bruce Henderson, the founder of BCG,
-
0:46 - 0:50and Michael Porter, professor
at the Harvard Business School. -
0:50 - 0:53Henderson's central idea was what you might call
-
0:53 - 0:56the Napoleonic idea of concentrating mass
-
0:56 - 0:59against weakness, of overwhelming the enemy.
-
0:59 - 1:01What Henderson recognized was that,
-
1:01 - 1:02in the business world,
-
1:02 - 1:05there are many phenomena which are characterized
-
1:05 - 1:07by what economists would call increasing returns --
-
1:07 - 1:08scale, experience.
-
1:08 - 1:10The more you do of something,
-
1:10 - 1:12disproportionately the better you get.
-
1:12 - 1:15And therefore he found a logic for investing
-
1:15 - 1:18in such kinds of overwhelming mass
-
1:18 - 1:21in order to achieve competitive advantage.
-
1:21 - 1:23And that was the first introduction
-
1:23 - 1:25of essentially a military concept of strategy
-
1:25 - 1:28into the business world.
-
1:28 - 1:31Porter agreed with that premise,
-
1:31 - 1:33but he qualified it.
-
1:33 - 1:36He pointed out, correctly, that that's all very well,
-
1:36 - 1:40but businesses actually have multiple steps to them.
-
1:40 - 1:41They have different components,
-
1:41 - 1:44and each of those components might be driven
-
1:44 - 1:45by a different kind of strategy.
-
1:45 - 1:48A company or a business
might actually be advantaged -
1:48 - 1:51in some activities but disadvantaged in others.
-
1:51 - 1:53He formed the concept of the value chain,
-
1:53 - 1:56essentially the sequence of steps with which
-
1:56 - 2:00a, shall we say, raw material, becomes a component,
-
2:00 - 2:01becomes assembled into a finished product,
-
2:01 - 2:04and then is distributed, for example,
-
2:04 - 2:06and he argued that advantage accrued
-
2:06 - 2:08to each of those components,
-
2:08 - 2:09and that the advantage of the whole
-
2:09 - 2:11was in some sense the sum or the average
-
2:11 - 2:14of that of its parts.
-
2:14 - 2:16And this idea of the value chain was predicated
-
2:16 - 2:19on the recognition that
-
2:19 - 2:23what holds a business together is transaction costs,
-
2:23 - 2:25that in essence you need to coordinate,
-
2:25 - 2:27organizations are more efficient at coordination
-
2:27 - 2:29than markets, very often,
-
2:29 - 2:32and therefore the nature and role and boundaries
-
2:32 - 2:36of the cooperation are defined by transaction costs.
-
2:36 - 2:39It was on those two ideas,
-
2:39 - 2:42Henderson's idea of increasing returns
-
2:42 - 2:43to scale and experience,
-
2:43 - 2:46and Porter's idea of the value chain,
-
2:46 - 2:48encompassing heterogenous elements,
-
2:48 - 2:51that the whole edifice of business strategy
-
2:51 - 2:54was subsequently erected.
-
2:54 - 2:56Now what I'm going to argue is
-
2:56 - 3:02that those premises are, in fact, being invalidated.
-
3:02 - 3:05First of all, let's think about transaction costs.
-
3:05 - 3:07There are really two components
to transaction costs. -
3:07 - 3:10One is about processing information,
and the other is about communication. -
3:10 - 3:13These are the economics of
processing and communicating -
3:13 - 3:16as they have evolved over a long period of time.
-
3:16 - 3:19As we all know from so many contexts,
-
3:19 - 3:21they have been radically transformed
-
3:21 - 3:23since the days when Porter and Henderson
-
3:23 - 3:25first formulated their theories.
-
3:25 - 3:27In particular, since the mid-'90s,
-
3:27 - 3:29communications costs have actually been falling
-
3:29 - 3:31even faster than transaction costs,
-
3:31 - 3:34which is why communication, the Internet,
-
3:34 - 3:38has exploded in such a dramatic fashion.
-
3:38 - 3:41Now, those falling transaction costs
-
3:41 - 3:43have profound consequences,
-
3:43 - 3:44because if transaction costs are the glue
-
3:44 - 3:47that hold value chains together, and they are falling,
-
3:47 - 3:48there is less to economize on.
-
3:48 - 3:51There is less need for vertically
integrated organization, -
3:51 - 3:54and value chains at least can break up.
-
3:54 - 3:57They needn't necessarily, but they can.
-
3:57 - 3:59In particular, it then becomes possible for
-
3:59 - 4:01a competitor in one business
-
4:01 - 4:04to use their position in one step of the value chain
-
4:04 - 4:06in order to penetrate or attack
-
4:06 - 4:09or disintermediate the competitor in another.
-
4:09 - 4:12That is not just an abstract proposition.
-
4:12 - 4:13There are many very specific stories
-
4:13 - 4:15of how that actually happened.
-
4:15 - 4:19A poster child example was
the encyclopedia business. -
4:19 - 4:20The encyclopedia business
-
4:20 - 4:22in the days of leatherbound books
-
4:22 - 4:24was basically a distribution business.
-
4:24 - 4:26Most of the cost was the
commission to the salesmen. -
4:26 - 4:29The CD-ROM and then the Internet came along,
-
4:29 - 4:32new technologies made the distribution of knowledge
-
4:32 - 4:35many orders of magnitude cheaper,
-
4:35 - 4:37and the encyclopedia industry collapsed.
-
4:37 - 4:40It's now, of course, a very familiar story.
-
4:40 - 4:42This, in fact, more generally was the story
-
4:42 - 4:45of the first generation of the Internet economy.
-
4:45 - 4:47It was about falling transaction costs
-
4:47 - 4:48breaking up value chains
-
4:48 - 4:51and therefore allowing disintermediation,
-
4:51 - 4:53or what we call deconstruction.
-
4:53 - 4:56One of the questions I was occasionally asked was,
-
4:56 - 4:58well, what's going to replace the encyclopedia
-
4:58 - 5:01when Britannica no longer has a business model?
-
5:01 - 5:03And it was a while before
the answer became manifest. -
5:03 - 5:06Now, of course, we know
what it is: it's the Wikipedia. -
5:06 - 5:09Now what's special about the
Wikipedia is not its distribution. -
5:09 - 5:11What's special about the Wikipedia
is the way it's produced. -
5:11 - 5:14The Wikipedia, of course, is an encyclopedia
-
5:14 - 5:16created by its users.
-
5:16 - 5:18And this, in fact, defines what you might call
-
5:18 - 5:21the second decade of the Internet economy,
-
5:21 - 5:24the decade in which the Internet as a noun
-
5:24 - 5:26became the Internet as a verb.
-
5:26 - 5:28It became a set of conversations,
-
5:28 - 5:32the era in which user-generated
content and social networks -
5:32 - 5:34became the dominant phenomenon.
-
5:34 - 5:36Now what that really meant
-
5:36 - 5:40in terms of the Porter-Henderson framework
-
5:40 - 5:43was the collapse of certain
kinds of economies of scale. -
5:43 - 5:45It turned out that tens of thousands
-
5:45 - 5:48of autonomous individuals writing an encyclopedia
-
5:48 - 5:50could do just as good a job,
-
5:50 - 5:52and certainly a much cheaper job,
-
5:52 - 5:55than professionals in a hierarchical organization.
-
5:55 - 5:57So basically what was happening was that one layer
-
5:57 - 6:00of this value chain was becoming fragmented,
-
6:00 - 6:02as individuals could take over
-
6:02 - 6:05where organizations were no longer needed.
-
6:05 - 6:08But there's another question
that obviously this graph poses, -
6:08 - 6:10which is, okay, we've
gone through two decades -- -
6:10 - 6:13does anything distinguish the third?
-
6:13 - 6:15And what I'm going to argue is that indeed
-
6:15 - 6:16something does distinguish the third,
-
6:16 - 6:19and it maps exactly on to the kind of
-
6:19 - 6:21Porter-Henderson logic that
we've been talking about. -
6:21 - 6:24And that is, about data.
-
6:24 - 6:26If we go back to around 2000,
-
6:26 - 6:28a lot of people were talking
about the information revolution, -
6:28 - 6:30and it was indeed true that the world's stock of data
-
6:30 - 6:33was growing, indeed growing quite fast.
-
6:33 - 6:35but it was still at that point overwhelmingly analog.
-
6:35 - 6:37We go forward to 2007,
-
6:37 - 6:40not only had the world's stock of data exploded,
-
6:40 - 6:43but there'd been this massive substitution
-
6:43 - 6:45of digital for analog.
-
6:45 - 6:47And more important even than that,
-
6:47 - 6:49if you look more carefully at this graph,
-
6:49 - 6:51what you will observe is that about a half
-
6:51 - 6:52of that digital data
-
6:52 - 6:55is information that has an I.P. address.
-
6:55 - 6:57It's on a server or it's on a P.C.
-
6:57 - 6:59But having an I.P. address means that it
-
6:59 - 7:01can be connected to any other data
-
7:01 - 7:03that has an I.P. address.
-
7:03 - 7:05It means it becomes possible
-
7:05 - 7:08to put together half of the world's knowledge
-
7:08 - 7:10in order to see patterns,
-
7:10 - 7:12an entirely new thing.
-
7:12 - 7:14If we run the numbers forward to today,
-
7:14 - 7:15it probably looks something like this.
-
7:15 - 7:17We're not really sure.
-
7:17 - 7:19If we run the numbers forward to 2020,
-
7:19 - 7:21we of course have an exact number, courtesy of IDC.
-
7:21 - 7:26It's curious that the future is so much
more predictable than the present. -
7:26 - 7:30And what it implies is a hundredfold multiplication
-
7:30 - 7:33in the stock of information that is connected
-
7:33 - 7:35via an I.P. address.
-
7:35 - 7:39Now, if the number of connections that we can make
-
7:39 - 7:42is proportional to the number of pairs of data points,
-
7:42 - 7:44a hundredfold multiplication in the quantity of data
-
7:44 - 7:47is a ten-thousandfold multiplication
-
7:47 - 7:48in the number of patterns
-
7:48 - 7:50that we can see in that data,
-
7:50 - 7:53this just in the last 10 or 11 years.
-
7:53 - 7:56This, I would submit, is a sea change,
-
7:56 - 7:58a profound change in the economics
-
7:58 - 8:00of the world that we live in.
-
8:00 - 8:01The first human genome,
-
8:01 - 8:02that of James Watson,
-
8:02 - 8:06was mapped as the culmination of the
Human Genome Project in the year 2000, -
8:06 - 8:09and it took about 200 million dollars
-
8:09 - 8:11and about 10 years of work to map
-
8:11 - 8:13just one person's genomic makeup.
-
8:13 - 8:16Since then, the costs of mapping
the genome have come down. -
8:16 - 8:17In fact, they've come down in recent years
-
8:17 - 8:19very dramatically indeed,
-
8:19 - 8:22to the point where the cost
is now below 1,000 dollars, -
8:22 - 8:25and it's confidently predicted that by the year 2015
-
8:25 - 8:26it will be below 100 dollars --
-
8:26 - 8:29a five or six order of magnitude drop
-
8:29 - 8:31in the cost of genomic mapping
-
8:31 - 8:34in just a 15-year period,
-
8:34 - 8:36an extraordinary phenomenon.
-
8:36 - 8:41Now, in the days when mapping a genome
-
8:41 - 8:44cost millions, or even tens of thousands,
-
8:44 - 8:46it was basically a research enterprise.
-
8:46 - 8:48Scientists would gather some representative people,
-
8:48 - 8:49and they would see patterns, and they would try
-
8:49 - 8:52and make generalizations about
human nature and disease -
8:52 - 8:54from the abstract patterns they find
-
8:54 - 8:57from these particular selected individuals.
-
8:57 - 9:00But when the genome can
be mapped for 100 bucks, -
9:00 - 9:0299 dollars while you wait,
-
9:02 - 9:05then what happens is, it becomes retail.
-
9:05 - 9:07It becomes above all clinical.
-
9:07 - 9:08You go the doctor with a cold,
-
9:08 - 9:10and if he or she hasn't done it already,
-
9:10 - 9:12the first thing they do is map your genome,
-
9:12 - 9:14at which point what they're now doing
-
9:14 - 9:18is not starting from some abstract
knowledge of genomic medicine -
9:18 - 9:20and trying to work out how it applies to you,
-
9:20 - 9:23but they're starting from your particular genome.
-
9:23 - 9:24Now think of the power of that.
-
9:24 - 9:26Think of where that takes us
-
9:26 - 9:29when we can combine genomic data
-
9:29 - 9:30with clinical data
-
9:30 - 9:32with data about drug interactions
-
9:32 - 9:35with the kind of ambient data that devices
-
9:35 - 9:36like our phone and medical sensors
-
9:36 - 9:38will increasingly be collecting.
-
9:38 - 9:41Think what happens when we collect all of that data
-
9:41 - 9:42and we can put it together
-
9:42 - 9:45in order to find patterns we wouldn't see before.
-
9:45 - 9:48This, I would suggest, perhaps it will take a while,
-
9:48 - 9:50but this will drive a revolution in medicine.
-
9:50 - 9:52Fabulous, lots of people talk about this.
-
9:52 - 9:55But there's one thing that
doesn't get much attention. -
9:55 - 9:58How is that model of colossal sharing
-
9:58 - 10:01across all of those kinds of databases
-
10:01 - 10:03compatible with the business models
-
10:03 - 10:06of institutions and organizations and corporations
-
10:06 - 10:08that are involved in this business today?
-
10:08 - 10:11If your business is based on proprietary data,
-
10:11 - 10:14if your competitive advantage
is defined by your data, -
10:14 - 10:17how on Earth is that company or is that society
-
10:17 - 10:20in fact going to achieve the value
-
10:20 - 10:23that's implicit in the technology? They can't.
-
10:23 - 10:25So essentially what's happening here,
-
10:25 - 10:27and genomics is merely one example of this,
-
10:27 - 10:30is that technology is driving
-
10:30 - 10:32the natural scaling of the activity
-
10:32 - 10:35beyond the institutional boundaries within which
-
10:35 - 10:37we have been used to thinking about it,
-
10:37 - 10:39and in particular beyond the institutional boundaries
-
10:39 - 10:42in terms of which business strategy
-
10:42 - 10:45as a discipline is formulated.
-
10:45 - 10:49The basic story here is that what used to be
-
10:49 - 10:52vertically integrated, oligopolistic competition
-
10:52 - 10:55among essentially similar kinds of competitors
-
10:55 - 10:57is evolving, by one means or another,
-
10:57 - 11:01from a vertical structure to a horizontal one.
-
11:01 - 11:03Why is that happening?
-
11:03 - 11:05It's happening because
transaction costs are plummeting -
11:05 - 11:07and because scale is polarizing.
-
11:07 - 11:09The plummeting of transaction costs
-
11:09 - 11:11weakens the glue that holds value chains together,
-
11:11 - 11:13and allows them to separate.
-
11:13 - 11:15The polarization of scale economies
-
11:15 - 11:18towards the very small -- small is beautiful --
-
11:18 - 11:21allows for scalable communities
-
11:21 - 11:24to substitute for conventional corporate production.
-
11:24 - 11:26The scaling in the opposite direction,
-
11:26 - 11:28towards things like big data,
-
11:28 - 11:29drive the structure of business
-
11:29 - 11:32towards the creation of new kinds of institutions
-
11:32 - 11:34that can achieve that scale.
-
11:34 - 11:37But either way, the typically vertical structure
-
11:37 - 11:40gets driven to becoming more horizontal.
-
11:40 - 11:42The logic isn't just about big data.
-
11:42 - 11:46If we were to look, for example,
at the telecommunications industry, -
11:46 - 11:48you can tell the same story about fiber optics.
-
11:48 - 11:50If we look at the pharmaceutical industry,
-
11:50 - 11:52or, for that matter, university research,
-
11:52 - 11:54you can say exactly the same story
-
11:54 - 11:55about so-called "big science."
-
11:55 - 11:57And in the opposite direction,
-
11:57 - 11:59if we look, say, at the energy sector,
-
11:59 - 12:02where all the talk is about how households
-
12:02 - 12:06will be efficient producers of green energy
-
12:06 - 12:08and efficient conservers of energy,
-
12:08 - 12:10that is, in fact, the reverse phenomenon.
-
12:10 - 12:12That is the fragmentation of scale
-
12:12 - 12:14because the very small can substitute
-
12:14 - 12:17for the traditional corporate scale.
-
12:17 - 12:19Either way, what we are driven to
-
12:19 - 12:22is this horizontalization of the structure of industries,
-
12:22 - 12:25and that implies fundamental changes
-
12:25 - 12:27in how we think about strategy.
-
12:27 - 12:29It means, for example, that we need to think
-
12:29 - 12:31about strategy as the curation
-
12:31 - 12:34of these kinds of horizontal structure,
-
12:34 - 12:36where things like business definition
-
12:36 - 12:37and even industry definition
-
12:37 - 12:40are actually the outcomes of strategy,
-
12:40 - 12:43not something that the strategy presupposes.
-
12:43 - 12:46It means, for example, we need to work out
-
12:46 - 12:49how to accommodate collaboration
-
12:49 - 12:50and competition simultaneously.
-
12:50 - 12:52Think about the genome.
-
12:52 - 12:53We need to accommodate the very large
-
12:53 - 12:55and the very small simultaneously.
-
12:55 - 12:57And we need industry structures
-
12:57 - 13:00that will accommodate very,
very different motivations, -
13:00 - 13:03from the amateur motivations
of people in communities -
13:03 - 13:05to maybe the social motivations
-
13:05 - 13:07of infrastructure built by governments,
-
13:07 - 13:10or, for that matter, cooperative institutions
-
13:10 - 13:12built by companies that are otherwise competing,
-
13:12 - 13:15because that is the only way
that they can get to scale. -
13:15 - 13:17These kinds of transformations
-
13:17 - 13:21render the traditional premises
of business strategy obsolete. -
13:21 - 13:24They drive us into a completely new world.
-
13:24 - 13:25They require us, whether we are
-
13:25 - 13:28in the public sector or the private sector,
-
13:28 - 13:30to think very fundamentally differently
-
13:30 - 13:32about the structure of business,
-
13:32 - 13:35and, at last, it makes strategy interesting again.
-
13:35 - 13:38Thank you.
-
13:38 - 13:41(Applause)
- Title:
- How data will transform business
- Speaker:
- Philip Evans
- Description:
-
What does the future of business look like? In an informative talk, Philip Evans gives a quick primer on two long-standing theories in strategy — and explains why he thinks they are essentially invalid.
- Video Language:
- English
- Team:
- closed TED
- Project:
- TEDTalks
- Duration:
- 17:45
Jenny Zurawell edited English subtitles for How data will transform business | ||
Jenny Zurawell edited English subtitles for How data will transform business | ||
Madeleine Aronson edited English subtitles for How data will transform business | ||
Morton Bast edited English subtitles for How data will transform business | ||
Morton Bast approved English subtitles for How data will transform business | ||
Morton Bast edited English subtitles for How data will transform business | ||
Morton Bast edited English subtitles for How data will transform business | ||
Morton Bast edited English subtitles for How data will transform business |