13:58
TED@BCG San Francisco

Philip Evans: How data will transform business

Filmed:

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.

- Consultant
BCG's Philip Evans has a bold prediction for the future of business strategy -- and it starts with Big Data. Full bio

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

▲Back to top

About the Speaker:

Philip Evans - Consultant
BCG's Philip Evans has a bold prediction for the future of business strategy -- and it starts with Big Data.

Why you should listen

Since the 1970s, business strategy has been dominated by two major theories: Bruce Henderson's idea of increasing returns to scale and experience and Michael Porter's value chain. But now decades later, in the wake of web 2.0, Philip Evans argues that a new force will rule business strategy in the future -- the massive amount of data shared by competing groups.

Evans, a senior partner and managing director at the Boston Consulting Group, is the co-author of Blown to Bits, about how the information economy is bringing the trade-off between "richness and reach" to the forefront of business. Evans is based in Boston.

More profile about the speaker
Philip Evans | Speaker | TED.com