These slides contrast two processes by which new technologies become economically feasible. Some technologies become economically feasible as advances in science facilitate the creation of new concepts and improvements in the resulting technologies. Other technologies become economically feasible as improvements in electronic components (e.g., Moore's Law), smart phones, and the Internet experience improvements.
How and When do New Technologies Become Economically Feasible
1. A/Prof Jeffrey Funk
Division of Engineering and Technology Management
National University of Singapore
For information on other technologies: see http://www.slideshare.net/Funk98/presentations or
Exponential Change: What drives it? What does it tell us about the future? http://www.amazon.com/Exponential-Change-
drives-about-future-ebook/dp/B00HPSAYEM/ref=sr_1_1?s=digital-text&ie=UTF8&qid=1399871060&sr=1-1&keywords=exponential+change
2. Some new technologies destroy both an existing economic
system and create a new one (Schumpeter, 1942)
These technologies
provide significantly higher economic value than do old ones
enable dramatic improvements in economic productivity and thus
living standards (e.g., Solow (1957)
create winners and losers at individual, firm, and country level
have a large impact on our ecological and social environment
Over last 20 years, Apple, Google, Amazon, and Microsoft have
enabled creative destruction
What is the long-term evolutionary process by which
opportunities become economically feasible?
Many Terms: Technological Discontinuities, Radical
or Disruptive Innovations, Creative Destruction
3. Diffusion (and Obsolescence) are the Last Stages of Creative
Destruction: What Happens before Diffusion Occurs?
What is
Happening
Here?
5. What Happens Before Diffusion Starts?
How do new technologies become economically
feasible?
What is the long term evolutionary process by which
they become economically feasible?
And thus diffuse and cause creative destruction?
We can distinguish between economic feasibility
and the organizational and regulatory challenges
of implementing new technologies
Technologies (or Systems Composed of Them)
that Experience Rapid Improvements are more
Likely to Become Economically Feasible than are
Slowly Improving Technologies
6. Rate of Improvement
ExtentofImprovementNeeded
Small
Large
Slow (e.g., <5% Fast >10%
Technologies that Experience Rapid Improvements
are more Likely to Become Economically Feasible
Now or Probably
Very Soon
Probably
Never
Within
5 to 15 Years?
Within 5-15
Years?
7. Session Technology
1 Objectives and overview of course
2 How do improvements in cost and performance occur? 1) Creating materials
that better exploit physical phenomena; 2) Geometrical scaling
3 What is process by which new technologies become economically
feasible?
4 Future of ICs and Electronic Systems
5 Sensors, MEMs, and the Internet of Things
6 Bio-Electronics, DNA Sequencers, and Health Care
7 Lighting, Lasers, Displays
8 Human-Computer Interfaces, Wearable Computing
9 IT and Transportation
10 Nanotechnology and Superconductivity
11 Feedback on Group Slides
12-13 Group Presentations
This is Third Session of MT5009
8. Outline
Supply and demand curves and economic feasibility
Two models of technology change
Model of Invention
Silicon Valley Model (Improvements in components lead to
emergence of new systems)
Billion Dollar Startup Club
Problems with over-emphasizing model of invention: Predictions
made by MIT’s Technology Review
Myths of technology change
S-Curves for performance
Slowdowns in old technologies lead to improvements in new
technologies
Learning and Experience Curves
A-U Model
11. What are some problems with last
slide?
Previous slide assumes performance is unimportant
In reality, performance is important
Market evaluates products and services in terms of
price and a variety of performance dimensions
Difficult to represent multiple dimensions on a two-
dimension graph, so most graphs only show price and
quantity
Many use the term value proposition to capture price
and all dimensions of performance
12. Simple Definition of Value Proposition
Value to
the
target
market
Benefits to
the
target
market
Price to
the
target
market
=
Relative
to
A simple and clear statement of what the new technology provides and that the
existing technology does not: better performance, features, or price
Such a statement involves performance (including features) and cost
13. Superior Value Propositions
Might involve lower price
Higher performance
Speed, ease of Use
Durability, portability
Maintainability
Reliability, Aesthetics
Specific Features
Your projects should consider many
aspects of value
But for this session, let’s simplify the
discussion and just focus on performance
and price
15. Price, Performance, and Demand
Price and performance determine the amount of
demand and supply
Rising performance often leads to growing demand
Falling price often leads to growing demand
What changes over time and how do these curves look
before there is a market (i.e., no commercial
production)?
When performance is too low?
Or when price is too high?
How can we represent these dynamics with supply and
demand curves?
16. Quantity (Q)
Price (P)
q
p
Diffusion often starts in segments/users that are willing to pay
more for products and services than are other segments/users
Demand
Curve
Supply Curve
Typical movement of
supply curve over time Typical
movement
of demand
curve over
time
17. Quantity (Q)
Price (P)
q
p
Maximum Threshold of Price:
the maximum price that the market will
pay for a new technology
Demand
Curve
Supply Curve
Typical movement of
supply curve over time
18. Quantity (Q)
Performance
(P)
q
p
Sometimes, diffusion starts in segments/users that have
lower performance expectations than other segments/users
Supply
Curve
Demand
Curve
Typical movement
of supply curve
over time?
19. Quantity (Q)
Performance
(P)
Minimum Threshold of Performance: the minimum
performance the market will accept for a new technology
Supply
Curve
Demand
Curve
Typical
movement
of supply
curve over
time
20. Whether we Focus on Performance or Price
Demand and supply curves help us think about important issues
Impact of falling price or increasing performance on demand
Levels of performance and price that are needed before a
technology becomes economically feasible
Other factors impact on diffusion such as standards, regulations,
and organizational issues
Demand and supply curves can also help us to think about the first
Products to diffuse
First value propositions
First designs
Markets to accept this diffusion
First customer segments
First customers within segments
First sales channels
21. But What Drives Changes in Supply Curves?
The second session focused on
Creating materials that better exploit physical
phenomena
Geometrical scaling
Some technologies directly experience improvements
through these two mechanisms while others indirectly
experience them through improvements in specific
“components”
Let’s place these improvements in a larger context
What is the long term evolutionary process by
which new technologies become economically
feasible?
22. Outline
Supply and demand curves and economic feasibility
Two evolutionary models of technology change
Model of Invention
Silicon Valley Model (Improvements in components lead to
emergence of new systems)
Billion Dollar Startup Club
Problems with over-emphasizing model of invention: Predictions
made by MIT’s Technology Review
Myths of technology change
S-Curves for performance
Slowdowns in old technologies lead to improvements in new
technologies
Learning and Experience Curves
A-U Model
23. What was the process by which these
opportunities emerged?
Did they emerge through process of
Invention
Commercialization
Diffusion?
This is most widely described process in economics,
emphasized by Joseph Schumpeter, Nathan Rosenberg,
Giovanni Dosi, Brian Arthur, and others
Science (new explanations of natural or artificial
phenomena) plays critical role in this process
Facilitates creation, demonstration, commercialization, and
improvement of a concept and product design
Did advances in science play critical role in emergence of
opportunities exploited by billion dollar startup club?
24. What was the process by which these
opportunities emerged? (2)
Or did they emerge through a different process?
Improvements in components, particularly those that are
defined as General Purpose Technologies (Paul David,
Timothy Bresnahan, Manuel Trajtenberg, Elhanan Helpman),
enable new products, systems, and services
For example, improvements in integrated circuits enabled new
forms of computers, mobile phones, and other electronic
products
More broadly speaking, improvements in ICs, lasers, glass
fiber, and computers enabled improvements in Internet, which
enabled new forms of content, services, and access devices
(e.g., mobile phones) to emerge*
See for example, http://www.slideshare.net/Funk98/when-do-new-technologies-
become-economically-feasible-the-case-of-electronic-products
25. All of these Models Involve Technological
Discontinuities (significant changes in design)
This term is widely used in courses on technology
management
Disruptive and radical innovations are discontinuities
One reason discontinuities are discussed is because
Incumbents often fail to effectively commercialize
them and thus lose substantial market share in the
new technology
Thus discontinuities represent large opportunities for
new entrants
It is also important to understand the concept that
forms the basis for the technology
One reason is because this helps us understand the
potential for improvements
26. Consider Two Types of Technological Discontinuities
Within Four Types of Innovations
Reinforced Overturned
Core Concepts
Unchanged
Changed
LinkagesBetweenCore
ConceptsandComponents
Incremental
Innovation
Modular
Innovation
Architectural
Innovation
Radical
Innovation
Source: Henderson and Clark (1990)
27. Henderson and Clark’s Innovation Framework
Applied to Ceiling Fans
Reinforced Overturned
Core Concepts
Unchanged
Changed
LinkagesBetweenCore
ConceptsandComponents
Improvements
in Blade or
Motor Design
Completely new
form of motor
Portable Fans Air Conditioners
28. Steam-powered fire engine
Technological Discontinuities: What was change in concepts?
Old
Technology
New
Technology,
i.e.,
Discontinuity
Early Benz (1894)
Wright Brothers (1904)
Gliders (19th Century)
29. Which model can best help us to find new
opportunities?
Each model has its advantages and disadvantages
Let’s look at each model
Then we will use the billion dollar startup club to
understand the relative importance of these models
Recent startups (still private) with valuations of greater
than $1Billion
http://graphics.wsj.com/billion-dollar-club/
Which models help us understand how the opportunities
emerged?
Multiple models may be applicable to a specific
opportunity
Returning to Four Models of Technology Change
30. Outline
Supply and demand curves and economic feasibility
Four models of technology change
Model of Invention
Silicon Valley Model (Improvements in components lead
to emergence of new systems)
Disruptive model of technology change
Billion Dollar Startup Club
Problems with over-emphasizing model of invention :
Predictions made by MIT’s Technology Review
Myths of technology change
32. Different People, Different Terms
Research (Basic, Applied) and Development
Science, Technology, Commercialization
Invention (proof of concept), Innovation
(commercialization of discontinuity/concept), Diffusion of
discontinuity/concept
Technological Discontinuity: based on new concept that
comes from advances in science (sometimes called radical
or disruptive innovation)
Scientific, technical, and economic feasibility
In any case, technologies proceed through stages of
scientific, technical and economic feasibility
advances in science often continue throughout these stages and
contributes towards improvements
33. Advances in science play a key role
First step in process by which new technology becomes
Scientifically feasible
Technically feasible
Economically feasible
Science
provides basis for technology (including invention) and
how it works (concept or paradigm) (Dosi, 1982)
facilitates refinement and improvement of concepts and
prototypes (Arthur, 2007, 2009)
resulting in cost and performance trajectories (Dosi, 1982;
Funk and Magee, 2015)
Model of Invention
34. Engineers and Scientists focus on this model
Mostly consider technologies that involve advances in science
Ignore other technologies that might be considered trivial in
terms of advances in science (e.g., smart phone, tablet
computer)
Example presented later on MIT’s Technology Review
Many scientists focus on scientific feasibility when
they discuss future
Michio Kaku: The Future of the Mind, Physics of the Future,
Physics of the Impossible
Peter Diamandis: Abundance
Model of Invention
35. Examples of technologies for which model of invention
provides important insights
Bio-technology
Organic light emitting diodes (OLEDs)
Organic transistors and solar cells
Quantum dot solar cells and displays
New forms of non-volatile memory, carbon nano-tubes
Superconductors, quantum computers, graphene,
Holograms, nano-fiber and particles (see following slides)
Patent literature heavily emphasizes linear model
Science measured with papers
Innovation measured with patents
Model of Invention
36. 0.1
1
10
100
1000
1985 1990 1995 2000 2005
Green
Yellow
Blue
White
Lumens/Watt
Why might costs fall as luminosity per Watt rises?
start of commercial production
Luminosity Per Watt for Organic Light Emitting Diodes
37. 0.000001
0.0001
0.01
1
100
1980 1985 1990 1995 2000 2005 2010
Mobility of Single Crystal and Polycrystalline Organic transistors
Single
crystal
Poly
crystalline
Mobility(cm2xsec)
Start of
Commercial
Production
Why might costs fall as
mobility rises?
38. 1998 2002 2006 2010 2014
Organic
Quantum
Dots
Efficiency of Organic and Quantum Dot Solar Cells
25%
5%
0%
Efficiency
Why might costs fall as efficiency rises?
Perovskite solar cells now have about 20% efficiency
but without any commercial production
39. 1990 1995 2000 2005 2010 2015
Red
Blue
Orange
Yellow
Green
Efficiency of Quantum Dot Displays for Different Colors
10%
1%
.1%
.01%
Efficiency
Start of
Commercial
Production
Why might costs fall as
efficiency rises?
40. 0.001
0.1
10
2001 2003 2005 2007 2009 2011 2013
Phase Change
RAM
Ferro Electric
RAM
Magnetic
RAM
Resistive
RAM
Number of Memory Bits (Gb) per RAM
(Random Access Memory) Chip
StorageCapacityperChip(Gb)
Why do costs
fall as storage
capacity rises?
41. 0.01
0.1
1
10
100
0.01
0.1
1
10
100
1995 2000 2005 2010 2015
Density with
Inconsistent
Feature Size
Density with
Consistent
Feature Size
Density(CarbonNanotubes(permicrometer)
Purity(%Contaminant)
Purity (left axis) and Density (right axis) of Carbon Nano Tubes for
Transistors. Density is for Consistent and Inconsistent Feature Size
Purity
Start of
Production
43. 1
10
100
1000
1985 1990 1995 2000 2005 2010 2015
YBaCuO
BiSr CuO
Current (Amps) x Length (km) for Two Types of
Superconducting Cables
AmpsxLength
Start of
commercial
production
Why might costs fall
as current
x length increases
44. 0.001
0.01
0.1
1
1
10
100
1000
1990 1995 2000 2005 2010 2015
Bit Energy (left axis) and Clock Period (right axis) for Super-
conducting Josephson Junctions
BitEnergy(FemtoJoules)
ClockPeriod(PicoSeconds)
Clock Period
Bit Energy
Start of
commercial
production
Why do costs fall
as speeds increase and energy
consumption falls?
45. 0.001
0.1
10
1000
100000
1998 2002 2006 2010 2014
Relaxation Time
Coherence Time
Cavity
Lifetime
QuBit Lifetime for Several Definitions of "Lifetime"
Lifetime(nanoseconds)
Start of
Commercial
Production
47. 1
10
100
1000
2000 2004 2008 2012
NumberofBits
Number of Qubits in Quantum Computers
(mostly prototypes)
Start of
Commercial
Production
48. Who Improved These Technologies?
Many of the improvements were implemented by
university researchers
Primarily motivated by publications and perhaps also
patents and forming firms
Others were implemented by startups and corporate
labs of large firms (e.g., IBM, Samsung)
Motivated partly by publications and mostly by firm’s
desire to commercialize new technologies
49. How were these Technologies Improved? (1)
Creating new materials
Ones with higher luminosity per Watt (both OLEDs and
Quantum dots)
Ones that convert more sunlight to electricity (Organic,
Quantum Dot, and Perovskite Solar Cells)
Ones with higher mobility (organic transistors)
Ones with higher critical temperatures, magnetic fields,
and current densities (superconductors)
New materials also require new processes, so each of
these new materials required new processes
50. How were these Technologies Improved? (2)
New processes (really a subset of creating new
materials)
Higher purity and density of carbon nano-tubes
Longer Qubit lifetimes and number of Qubits per lifetime
When the material is fixed, the improvements primarily
come from changes in processes
Reducing the feature size of memory cells or Josephson
Junctions
Non-volatile memory
Superconducting Josephson Junctions
This also requires changes in both product and process
design. Smaller feature sizes involve new product designs
and they require new processes in order to achieve the
smaller feature sizes
51. Outline
Supply and demand curves and economic feasibility
Two models of technology change
Model of Invention
Silicon Valley Model (Improvements in components lead to
emergence of new systems)
Billion Dollar Startup Club
Problems with over-emphasizing model of invention: Predictions
made by MIT’s Technology Review
Myths of technology change
S-Curves for performance
Slowdowns in old technologies lead to improvements in new
technologies
Learning and Experience Curves
A-U Model
52. Some components such as integrated circuits experience very
rapid improvements (Funk, 2013; Funk and Magee, 2015)
These improvements enable new forms of systems, such as
computers, mobile phones, other electronic products to emerge
(Bresnahan and Trajtenberg, 1995; Funk, 2013)
Rapidly improving components with many applications are
often called general purpose technologies by economists
(David, 1989; Bresnahan and Trajtenberg, 1995; Helpman, 2003; Lipsey et al,
2005)
Internet is considered general purpose technology by most economists
For new systems
costs and/or performance depend primarily on components
This was discussed in session 2
Improvements in Components Enable New
Systems to Emerge
53. Laptops MP3 Players
Calculators Video Set-top boxes E-Book Readers
Digital Games Web Browsers Digital TV
Watches Mobile Digital Cameras Smart Phones
Example: Better Integrated Circuits Make New Forms of
Electronic Products Economically Feasible
MT5009
focuses
on the
future
54. Internet has Experienced Rapid
Improvements
In speed, bandwidth and cost
These improvements have enabled new forms of
Content: from text to pictures, videos, flash content
Software: from assembly code to higher level languages
Cloud computing and Big Data
Advertising, pricing, recommendation techniques
Similar things have occurred with mobile networks
and phones
55. Outline
Supply and demand curves and economic feasibility
Two models of technology change
Model of Invention
Silicon Valley Model (Improvements in components lead to
emergence of new systems)
Billion Dollar Startup Club
Problems with over-emphasizing model of invention: Predictions
made by MIT’s Technology Review
Myths of technology change
S-Curves for performance
Slowdowns in old technologies lead to improvements in new
technologies
Learning and Experience Curves
A-U Model
56. Global Startups (sometimes called Unicorns)
valuations over $1 Billion
still private (no IPO yet)
have raised money in past four years
at least one venture capital firm as investor
122 firms as of 25 September 2015
With 21 other startups that recently exited (IPOs, acquisitions or
decreasing value), total of 143 firms
High valuations mean investors believe these firms offer
something valuable, unique, hard to copy
Some of them will
lead to “creative destruction”
have $100 Billion plus market capitalizations in the future, like
the strongest hi-tech startups: Apple, Google, Amazon, and
Microsoft
Wall Street Journal’s Billion Dollar Startup Club
57. Company Latest Valuation Total Equity Funding Last Valuation
Uber $51.0 billion $7.4 billion August 2015
Xiaomi $46.0 billion $1.4 billion December 2014
Airbnb $25.5 billion $2.3 billion June 2015
Palantir $20.0 billion $1.6 billion October 2015
Snapchat $16.0 billion $1.2 billion May 2015
Didi Kuaidi $16.0 billion $4.0 billion September 2015
Flipkart $15.0 billion $3.0 billion April 2015
SpaceX $12.0 billion $1.1 billion January 2015
Pinterest $11.0 billion $1.3 billion February 2015
Dropbox $10.0 billion $607 million January 2014
WeWork $10.0 billion $969 million June 2015
Lufax $9.6 billion $488 million March 2015
Theranos $9.0 billion $400 million June 2014
Spotify $8.5 billion $1.0 billion April 2015
DJI $8.0 billion $105 million May 2015
Zhong An Online $8.0 billion $934 million June 2015
Meituan $7.0 billion $1.1 billion January 2015
Square $6.0 billion $495 million August 2014
Stripe $5.0 billion $290 million July 2015
ANI Technologies (Ola Cabs) $5.0 billion $903 million September 2015
Snapdeal $5.0 billion $911 million August 2015
Stemcentrx $5.0 billion $250 million September 2015
Zenefits $4.5 billion $596 million May 2015
Cloudera $4.1 billion $670 million March 2014
Dianping $4.0 billion $1.4 billion March 2015
The Top 25 Firms as of 25 September, 2015
58. Category U.S. Europe China India Other Total
Software 38 1 2 41
E-Commerce 12 3 9 2 2 28
Consumer
Internet
18 6 7 2 4 37
Financial 7 4 3 1 15
Hardware 7 2 1 10
BioTech, Bio-
Electronics
7 1 8
Energy 2 2
Space 1 1
Retail 1 1
Total 92 14 22 6 9 143
Number of Startups, by Category and Country
Most are Internet Related (122)
Note: some of the startups were redefined and the smaller categories were combined,
based on the descriptions by the Wall Street Journal and other sources
59. 0
5
10
15
20
25
30
before
2001
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Number of Startups Founded by Year and Category
software
consumer internet
e-commerce
financial
hardware
healthcare
60. What was the process by which these
opportunities emerged?
Did they emerge through process of
Invention; Commercialization; Diffusion?
Science (new explanations of natural or artificial
phenomena) facilities improvements in technology and
often the creation of new concept
Or did they emerge through a different process?
Improvements in components enable new products,
systems, and services
For example, improvements in integrated circuits enabled
new forms of computers, mobile phones, and other
electronic products
61. Methodology
Which of these processes enabled the opportunities
to emerge that were exploited by the billion dollar
startup club?
Invention and advances in science
Examine the U.S. patents held by members of startup club
and scientific papers cited by the patents
Scientific papers are defined as papers published in
journals that are in science citation index
Contrast papers in physical and life science journals with
those in engineering (e.g., IEEE) and computer science
(e.g., Association of Computing Machinery) journals
Compare importance of scientific papers across categories
62. Methodology (2)
Which of these processes enabled the opportunities
to emerge that were exploited by the billion dollar
startup club?
Improvements in components
By reading descriptions provided by the Wall Street Journal
and other sources, what number of startups can be defined
as Internet related startups or in general those that benefit
from improvements in ICs?
Define sub-categories for each category of Internet-related
startups
Explain how improvements in components, particularly
those that can be defined as general purpose technologies,
enabled these opportunities to emerge?
63. Category Total
Number
Percentage of Startups with Patents
≥ 1 patent ≥ 10 patents ≥ 50 patents
Software 41 63% 29% 7.3%
E-Commerce 28 3.4% 3.4% 0%
Consumer
Internet
37 23% 17% 0%
Financial 15 6.6% 6.6% 6.6%
Hardware 10 90% 70% 20%
BioTech, Bio-
Electronics
8 88% 50% 0%
Energy 2 100% 100% 50%
Space 1 100% 0% 0%
Retail 1 0% 0% 0%
Total 143 39% 23% 4.9%
Percentage of Startups All, by Numbers of Patents
64. Category Total
Number
Percentage of Startups with Patents
≥ 1 patent ≥ 10 patents ≥ 50 patents
Software 38 66% 32% 7.9%
E-Commerce 12 8.3% 8.3% 0%
Consumer
Internet
18 39% 28% 0%
Financial 7 14% 14% 14%
Hardware 7 86% 72% 28%
BioTech, Bio-
Electronics
7 86% 56% 0%
Energy 2 100% 100% 50%
Space 1 100% 0% 0%
Total 92 54% 34% 7.6%
Percentage of U.S. Startups, by Numbers of Patents
65. Category Total
Number of
startups
Percentage of Startups with Patents Citing
Scientific Papers (SPs)
Citing ≥ 1 SP Citing ≥ 10
different SPs
Citing ≥ 50
different SPs
Software 41 27% 2.5% 0%
E-Commerce 28 0% 0% 0%
Consumer
Internet
37 2.7% 0% 0%
Financial 15 0% 0% 0%
Hardware 10 50% 0% 0%
BioTech, Bio-
Electronics
8 88% 75% 75%
Energy 2 50% 50% 0%
Space 1 0% 0% 0%
Retail 1 0% 0% 0%
Total 143 18% 5.6% 4.2%
Percentages of All Startups,
by Numbers of Scientific Papers Cited in Patents
66. Category Total
Number
Percentage of Startups with Patents Citing
Scientific Papers (SPs)
Citing ≥ 1 SP Citing ≥ 10
different SPs
Citing ≥ 50
different SPs
Software 38 28% 7.5% 0%
E-Commerce 12 0% 0% 0%
Consumer
Internet
18 0% 0% 0%
Financial 7 0% 0% 0%
Hardware 7 63% 0% 0%
BioTech, Bio-
Electronics
7 88% 75% 75%
Energy 2 50% 50% 0%
Space 1 0% 0% 0%
Total 92 23% 7.6% 5.4%
Percentages of U.S. Startups, by Numbers of Scientific
Papers Cited in Patents
67. Few Startups have Patents that Cite
Scientific Paper
Only 4.2% of all startups have patents that cite 50 or
more different papers
Only 5.6% of all startups have patents that cite 10 or
more different papers
Even for U.S. startups, the percentages are low
Only 5.4% of U.S. startups have patents that cite 50 or
more different papers
Only 7.6% of U.S. startups have patents that cite 10 or
more different papers
68. Startups for a Few Categories Have Patents
that Cite Scientific Papers
75% of BioTech/Bio-Electronic startups had patents
that cited more than 50 scientific papers
One of two (50%) of energy startups had patents
that cited more than 10 scientific papers
All of these startups are U.S. startups
In comparison
Only 7.3% of the software startups
None of the e-commerce, consumer Internet, or hardware
startups
had patents citing 10 or more scientific papers
69. Pure Science vs. Engineering and Computer
Science
The BioTech/Bio-Electronics and energy startups
had patents that cited papers in pure scientific
journals
Physics
Chemistry
Biology
Only one startup outside of these categories cited
papers in pure scientific journals
Palantir, a software startup
It cited more than 10 scientific papers including Nucleic
Acids Research and Bioinformatics
70. Summary
Advances in science did not directly play an
important role in the emergence of opportunities
exploited by startups
Few scientific papers cited in patents
Of the few papers cited in patents, they were mostly for
startups in Bio-Tech/Bio-Electronic category
Few science-based products
No carbon nanotubes, graphene, nano-particle, quantum
dot, superconductor, display, lighting, new forms of
integrated circuits, membranes, or quantum computers
https://www.youtube.com/watch?v=yesyhQkYrQM
These types of science-based technologies are emphasized
by engineering schools
71. Summary (2)
Thus, traditional process (invention,
commercialization and diffusion) does not explain
how opportunities emerged
Few science-based prototypes
Few improvements in performance and cost for these
new prototypes driven by advances in science
What about improvements in components?
Can they explain the emergence of opportunities that
were exploited by startups?
Let’s look at the startups for each category, beginning
with e-commerce
72. Sub-
Category
Number Names of Firms
Clothing and
Accessories
9 Fanatics, Vancl, Gilt Groupe, Mogujie,
JustFab, LaShou, Zalando, Global Fashion
Group, Lazada
Broad Variety
of Products
7 Flipkart, Contextlogic, Snapdeal, Coupang,
Koudai Shopping, Quikr, JD.com
Furniture,
Interior
Design
5 Home24, Honest Co, FarFetch, Wayfair, Fab
Other
Specialty
Sites
5 Warby Parker, Blue Apron, Beibei,
Pluralsight, We Work
Discount
coupons
2 Coupons.doc, Meituan
Total 28
Number of E-Commerce Startups by Sub-Category
73. What Enabled these Opportunities?
Fifteen of 28 startups focus on fashion related products
clothing, accessories, furniture, and interior design.
Although music, videos, books, electronic products
dominated early online sales in U.S. and elsewhere, fashion
related products have experienced rapid growth in online
sales over the last 10 years
their U.S. sales were 50% higher than those of music, videos,
books, electronic products in 2013
Improvements in Internet speed and bandwidth
Enabled more complex and aesthetically pleasing web pages
and thus fashion related opportunities (see below)
Enabled new users and thus new opportunities
The Ongoing Evolution of US Retail, Journal of Economic Perspectives, Ali Hortacsu and Chad Syverson, Vol 29, No 4, pp. 89-112
74. The Ongoing Evolution of US Retail, Journal of
Economic Perspectives, Ali Hortacsu and Chad
Syverson, Vol 29, No 4, 2015 pp. 89-112
Furniture,
Sporting Goods,
Clothing
Fashion, clothing,
furniture
E-Commerce Changed from Digital Products to Fashion, Clothing, Furniture
75. Increases in Speed Enable Increases in Web Page Size
And Number of Objects (pictures, videos, flash files)
76. Large number of 28 startups in two largest emerging
economies and are mobile related
9 from China and 2 from India
16 of them primarily depend on access from mobile phones
Both these trends consistent with improvements in
Internet and Internet-related devices
As cost and performance of Internet improved, Internet
diffusion spread to countries such as China and India and
to new access devices such as mobile phones
Smart phone-targeted services attract young people and
other fashion-conscious shoppers to the Internet and this
impacts on popularity of fashion-related products
Two Other Key Trends
77. Sub-Category Number Names of Firms
Ride Sharing 7 Uber, Didi Dache, Kuaidi Dache, Ola Cabs,
Lyft, Grabtaxi, BlaBla Car
Food (delivery,
restaurant search)
6 Delivery Hero, Zomato, Instacart, Hello Fresh,
Ele.me, Dianping
Audio/Video 6 Spotify, Shazam, Pinterest, Snapchat, KIK
Interactive, Tango
Games, Digital
Entertainment
5 Kabam, Garena Online, Fan Duel, Draft Kings,
Legendary Entertainment
Social Networking 5 Houzz, NextDoor, Eventbrite, Lamabang, Vox
Media
Hotels 2 Airbnb, Tujia
Health Care 2 Oscar Healthcare, ZocDoc
Other 5 Rocket Internet, Yello Mobile, APUS, BuzzFeed
Total 37
Number of Consumer Internet Startups by Sub-Category
78. What Enabled these Opportunities?
All 37 provide services that were not available during early
years of Internet (late 1990s and early 2000s)
because Internet did not have sufficient bandwidth and/or
mobile phones did not have sufficient capability
This is particularly true of startups that were founded in
China or India or that depend on mobile phones
9 from China or India
25 of 37 services mostly depend on smart phones and often
on smart phone apps
this includes all ride sharing, food, and hotel-related
services, and most audio/video, social networking, and other
services
79. What Enabled these Opportunities? (2)
Taxi apps (e.g., Uber) have become most famous of
these apps
Also food (delivery, restaurant search), audio/video
(music photos), some games, social networking
(specialized sites)
Many of these startups can also be defined as
members of the demand or sharing economy
Taxi, food services, Airbnb, Tujia
Consumers are paying for services that provide taxis,
food, and hotels on demand and all of these services
clearly require mobile phones
81. Sub-Category Number Names of Firms
Sales, Human
Resource, Inventory
Enterprise software
14 Shopify, Apttus, Coupa, Qualtrics, Zenefits,
Automattic, CloudFlare, InsideSales.com, Sprinklr,
Deem, AppDynamics, Slack, Medalia, Domo
Security software 5 Tanium, Good Technology, Lookout, Okta, Zscaler
Database, data storage
software
6 Nutanix, Simplivity, MarkLogic, PureStorage,
MongoDB, Actifio
Big Data Software and
Services
4 Palantir, Cloudera, Hortonworks, MuSigma
Online Ad Software 3 InMobi, AppNexus, IronSource
Cloud storage 2 Dropbox, Box
Software Dev. Tools 2 Twilio, Github
Other Tools 3 DocuSign, Evernote, New Relic
Integration Platforms 1 MuleSoft
Internet of Things
Platform
1 Jasper Technologies
Total 41
Number of Software Startups by Sub-Category
82. What Enabled these Opportunities?
All these opportunities involve cloud computing
Cloud computing emerged as improvements in Internet speed
and bandwidth occurred
Various types of enterprise, software, software development and
other tools, and platforms are accessed or downloaded via cloud
by organizations and to lesser extent individuals
Organizations use software for internal use or on website. This
includes enterprise software for human resources, sales,
marketing, operations, human resources
Purchase Big Data services or use cloud storage services
Economic feasibility of these opportunities depended on
improvements in Internet bandwidth and speed
83. What Enabled these Opportunities? (2)
Most of these opportunities involve big data
Big data is broad term for data sets so large or complex that
traditional data processing techniques are inadequate
It tests much more complex models with many more
independent variables than does traditional data analysis
It emerged as improvements in Internet speed and
bandwidth occurred
Organizations purchase
big data software and services
software services for sales, human resource, inventory
enterprise software suppliers that includes big data functions
84. What Enabled these Opportunities? (3)
Organizations purchase big data software/services
Purchase big data results through services or do big data internally
with software or services from
Palantir, Cloudera, MuSigma, Hortonworks
Use data base software from
Nutanix, Actifio, Simplivity, MarkLogic, PureStorage, MongoDB
Organizations also purchase software services for sales,
human resource, inventory enterprise software suppliers
that includes big data functions
this includes Shopify, Apttus, Coupa, Deem, AppDynamics,
Sprinklr, Qualtrics, InsideSales.com
Opportunities emerged as improvements in Internet
speed and bandwidth occurred
85. What Enabled these Opportunities? (4)
Improvements in Internet speed and bandwidth along
with emergence of cloud computing and big data caused
other opportunities to emerge
Security has become more important
Tanium, Good Technology, Lookout, Okta, Zscaler
Online ads have become more sophisticated, both in
presentation and delivery
InMobi, AppNexus, IronSource
Software development and integration have become
more expensive and important
Twilio, Github, MuleSoft
86. What Enabled these Opportunities? (5)
Five of the 41 software startups (InMobi, Good
Technology, Lookout, Evernote, Twilio) also
depended on emergence and diffusion of smart
phones such as iPhone and Android phones
Emergence of these phones and networks depended
on improvements in microprocessors, flash
memory, and displays
What’s Next? To be discussed in Session 4
See for example, http://www.slideshare.net/Funk98/when-do-new-technologies-become-economically-feasible-the-case-of-electronic-
products
87. Sub-
Category
Number Names of Firms
Peer-to peer
lending
5 Lufax, Prosper Marketplace, Social Finance,
Funding Circle, Lending Club
Mobile
payment
4 Stripe, One97 Communications, Adven, Square
E-commerce
Payment
2 Powa and Klarna
Other 4 Zhong an Online (insurance), Hanhua Financial
(credit guarantor), Credit Karma (credit scores),
Sunrun (solar leasing)
Total 15
Number of Financial Startups by Sub-Category
88. What Enabled these Opportunities?
Peer-to peer lending and “other” financial startups have
benefited from emergence of
Cloud Computing
Big Data
Web sites that specialize in peer-to peer loans offer loans to
customers, considered too risky by banks
set rates using special algorithms and these rates are often lower than
those offered by loan sharks
Improvements in Internet bandwidth and speed helped create more
sophisticated websites
that could assemble credit histories and credit scores of potential borrowers
through Big Data
Similar things happened with “other” and one “e-commerce”
Decisions about credit guarantees, credit scores, insurance, solar
leasing, and one e-commerce payments (Klarna) are based on Big Data
89. What Enabled these Opportunities? (2)
Feasibility of e-commerce payment services have benefited
from increasing market for e-commerce, which has benefited
from improvements in Internet
Feasibility of mobile payment services have benefited from
improvements in mobile phones
including growth in text messaging services in early 2000s
emergence and growth of smart phones in last seven years
Global opportunities
7 in U.S.
4 in Europe
3 in China
1 in India
What’s Next? Discussed some in Session 4
90. Hardware Related Startups
All ten hardware startups provide electronic
products
They are not placed into sub-categories
since each of them provides a different type of electronic
product
Smart phones, rugged cameras, drones
Wearable health, augmented reality glasses
Storage hardware, audio headphones, assisted vision
Gaming mouse, smart thermostats
All of these opportunities depended on
improvements in electronic components
91. Other Startups
Other startups depended less on improvements in
electronic components than did above ones
Energy, space, and retail startup do not benefit
greatly from improvements in electronic
components
Only four of eight 8 bio-tech/bio-electronic startups
might benefit from rapid improvements in
components, including electronic components
Theranos, Intarcia, Proteus Digital Health, 23andMe
Four other startups (Moderna, Stemcentrix, Adaptive
BioTechonlogies, CureVaC) offer drugs (bio-tech)
The drug related products involve advances in science,
and this is consistent with data on patents and papers
presented above
92. Summary
Most startups exploited opportunities that benefited
from improvements in electronic components, which
experience rapid improvements
Internet bandwidth, cost, and speed
Integrated circuits, lasers, photo-sensors, computers
These improvements in electronic components enabled
improvements in Internet speed and bandwidth and
emergence of new forms of
Internet content
Services
Software
121 of 143 are Internet related
10 are related to electronic hardware
93. What’s Next: Much of this will continue
with some changes Many new ones discussed in
Session 4
Software: cloud computing,
enterprise software, big data, online
ads, security, database
e-commerce: more fashion,
cosmetics, drugs, food and beverages
Consumer internet: Taxi (Session
9), social networking, food, games,
music, hotels
Financial services: P2P lending,
mobile payment, micro-financing
Hardware: Phones, drones,
wearable computing
Improvements in
components such as ICs
(Moore’s Law)
Greater use of smart
phones, other mobile
devices
Improvements in online
Internet experience
More social networking
Growth of Internet in new
countries: China, India,
and other countries
94. Internet of Things (Session 5)
Better and cheaper ICs, MEMS, transceivers (WiFi,
Bluetooth), and energy harvesters are enabling all
mechanical products to be connected to Internet
Which products can benefit the most from being attached
to Internet?
Big Data services and software will also emerge and they
will probably become the biggest opportunities
Who will provide these services and software?
Wearable Computing and Health Care (Sessions 6, 7, 8)
Better and cheaper bio-sensors, ICs, MEMS and displays
are enabling more health care related wearable computing
Big Data services and software will also emerge and will
likely provide most of the opportunities
Two Big Opportunities
For more info, see: http://www.slideshare.net/Funk98/presentations
96. The next step in
computing – after
smart phones and
tablets
One big market
for wearables
will be health
care – monitoring
various parts of
your body
For more info, see:
http://www.slideshare.net/Funk98/
presentations
97. For your projects
Can you better explain the types of software, e-
commerce, consumer internet, or financial services
that will likely emerge in the next few years?
Focus on one of the categories and explain the types of
technological changes that are changing the
economics and thus enabling better software, e-
commerce, consumer internet, or financial services to
emerge
Can you use this analysis to explain the types of
changes that will likely occur in the sub-categories?
Will some sub-categories become more or less important?
Will there be new sub-categories?
Or will there be new categories?
98. Outline
Supply and demand curves and economic feasibility
Two models of technology change
Model of Invention
Silicon Valley Model (Improvements in components lead to
emergence of new systems)
Billion Dollar Startup Club
Problems with over-emphasizing model of invention: Predictions
made by MIT’s Technology Review
Myths of technology change
S-Curves for performance
Slowdowns in old technologies lead to improvements in new
technologies
Learning and Experience Curves
A-U Model
99. For the remaining slides
Problems with over-emphasizing the linear model:
Predictions made by MIT’s Technology Review
See predicting breakthrough technologies……
http://www.slideshare.net/Funk98/presentations/2
Myths of Technology Change
http://www.slideshare.net/Funk98/presentations/3