Presentation given to a Finnish eMBA group during their visit at mediaX at Stanford.
Presenters: Kaisa Still, Jukka Huhtamäki
Session chair: Martha G. Russell
1. The
Transforma,on
of
Innova&on
Ecosystems
in
Global
Metropolitan
Areas
A
Data-‐Driven
Perspec,ve
Martha
G
Russell,
Jukka
Huhtamäki,
Kaisa
S,ll
Innova,on
Ecosystems
Network
TUT
eMBA
Visit
to
Stanford
University
Martha
Russell,
Rahul
C.
Basole,
Neil
Rubens,
Jukka
Huhtamäki,
Kaisa
S,ll
2. Transforming
Innova,on
Ecosystems
Through
Network
Orchestra,on:
Case
EIT
ICT
Labs
Dr.
Kaisa
S,ll,
VTT
Technical
Research
Centre
of
Finland
In
collabora,on
with
Marko
Turpeinen
and
others
at
EIT
ICT
Labs
Helsinki
3. Need
for
innova,on
indicators:
tradi,onal
measures
and
metrics
are
limited
Innova,on
ac,vi,es
rarely
carried
out
within
a
single
organiza,on:
Network
approach
to
understand
the
complex
systems
of
innova,on
Unprecedented
amount
of
data
about
the
complex
innova,on
system
and
its
actors:
Social
media,
socially
constructed
data
Possibili,es
of
SNA
and
visualiza,ons
Computer
power
4. • EIT
ICT
Labs
aims
“to
build
European
trust
based
on
mobility
of
people
across
countries,
disciplines
and
organiza,on”
• People,
their
knowledge
and
the
financial
flows
are
networked,
all
contribu,ng
toward
poten,al
of
innova,on
-‐>
Analysis
should
not
be
limited
to
labor
mobility
• How
to
measure,
analyze
and
visualize
mobility
of
people,
money
and
technology
in
the
European
ICT
innova,on
ecosystem?
EIT
ICT
Labs’
mission
is
to
turn
Europe
into
a
global
leader
in
ICT
Innova,on
5. Mobility
is
a
central
theme
5
nodes
working
together
Student
and
teacher
mobility,
Doctoral
School,
Mobility
programs
Ini,al
analysis
of
mobility
(S#ll
et
al
2010)
for
baseline:
with
geospa,al
representa,ons
of
networks
and
a
metric
of
betweenness
Highligh,ng
few
individuals,
more
investors,
less
so
of
universi,es,
and
the
role
of
Silicon
Valley
as
connectorà
(1)
new
”requirements”
for
data/
process
of
next
network
visualiza,on,
and
(2)
ini,al
insights
for
network
orchestra,on
6. Two
Studies
§
Using
IEN
Dataset
§
Betweenness
Centrality
§ number
of
,mes
that
a
given
node
is
included
in
the
shortest
path
between
any
two
nodes
in
the
network
(Wasserman
and
Faust,
1994)
§ point
out
investors,
individuals
and
educa,onal
ins,tu,ons
that
operate
in
between
the
six
EIT
ICT
Labs
Nodes
§ Coupled
with
the
modeling
applied,
can
be
used
as
a
metric
for
actor
mobility
in
an
innova,on
ecosystem
§
Note:
analysis
does
not
show
the
mobility
of
people
within
individual
companies
§
Two
consecu,ve
analysis:
first
in
2011
and
the
second
in
2012,
with
refined
segng
and
updated
data
(Gray,
2012)
7.
S,ll,
Russell,
Huhtamäki,
Turpeinen,
Rubens
(2011).
Explaining
innova#on
with
indicators
of
mobility
and
networks:
Insights
into
central
innova#on
nodes
in
Europe
Mobility
and
Educa,onal
Ins,tu,ons
2011
8. S,ll,
Russell,
Huhtamäki,
Turpeinen,
Rubens
(2011).
Explaining
innova#on
with
indicators
of
mobility
and
networks:
Insights
into
central
innova#on
nodes
in
Europe
Mobility
and
Financial
Flows
2011
9. Analysis
round
#2:
Trento
included
as
the
sixth
node,
more
ci,es
connected
to
coloca,on
centers,
updated
data
and
transforma,on
in
place
|
S,ll,
Huhtamäki,
Russell,
Rubens
(2012).
Transforming
Innova#on
Ecosystems
Through
Network
Orchestra#on:
Case
EIT
ICT
Labs
10. Finally,
adding
San
Francisco
Bay
Area
as
“the
seventh
EIT
ICT
Labs
node”
for
contrast,
interconnec,ons,
comparison
and
benchmark
S,ll,
Huhtamäki,
Russell,
Rubens
(2012).
Transforming
Innova#on
Ecosystems
Through
Network
Orchestra#on:
Case
EIT
ICT
Labs
11. Conclusions
§
Geospa,al
social
network
visualiza,on
make
it
possible
to
share
and
show
special
characteris,cs,
significant
actors
and
connenc,ons
in
the
innova,on
ecosystem
§
Betweenness
centrality
(how
central
a
node
is
within
a
network)
can
be
used
to
measure
innova,on
poten,al
of
an
ecosystem
§
Our
framework
can
be
used
for
understanding
the
transforma,on
and
for
bringing
transparency
§ At
the
same
,me,
when
interpreted
in
the
context,
our
approach
can
be
used
to
suggest
possibili,es
for
network
orchestra,on
12. Networks
of
innova&on
rela&onships:
mul&scopic
views
on
Finland
Presented
at
ISPIM
Helsinki
2012
Kaisa
S,ll,
VTT
Jukka
Huhtamäki,
TUT
Martha
G.
Russell,
Stanford
mediaX
Rahul
C.
Basole,
Georgia
Tech
Jaakko,
Salonen,
TUT
Neil
Rubens,
University
of
Electro-‐
Communica,ons
Jukka
Huhtamäki,
Tampere
University
of
Technology
13. Networks
of
innova,on
Approach
By
whom
The
shik
of
innova,on
from
a
single
firm
toward
an
increasingly
network-‐centric
ac,vity
Chesbrough
2003
Importance
of
collabora,on
and
value
co-‐crea,on
Ramaswamy
and
Goullart
2010
Resul,ng
networks
of
rela,onships
between
individual
and
organiza,onal
en,,es
Kogut
and
Zander
1996,
Vargo
2009
Studies
of
innova,on
ecosystems
Iansi,
and
Levien
2004,
Russell
et
al.
2011,
Basole
et
al.
2012,
Hwang
and
Horowio
2012,
Marts
et
al.
2012
14. From
data
with
visualiza,on
to
insights
Sense-‐making
and
storytelling
Boundary
specifica,on
Computa,on,
analysis
and
visualiza,on
Metrics
iden,fica,on
Analysing
a
business
ecosystem
21. Sense-‐making
and
storytelling:
So
what?
• Visualiza,ons
of
metrics
and
networks
can
be
seen
to
model
the
skeleton
of
an
ecosystem
• Tacit
knowledge
about
networks
(and
the
roles
of
certain
actors)
becomes
explicit
and
shared
22. Visualizing an Open Innovation
Platform: The structure and
dynamics of Demola
Huhtamäki, Luotonen, Kairamo, Still, Russell
TUT // New Factory // VTT // Stanford
Academic MindTrek 2013:
"Making Sense of Converging Media”
http://bit.ly/mt2013visualizingdemola // @jnkka
23. In this presentation
• Case context description: what is Demola?
• Challenges in measuring Demola & open
innovation
• Use case examples
• Method: data-driven network animation
• Results
• Discussion
• Critique
• Wrap up and future work
24. What is Demola?
• Open innovation platform & ecosystem
engager established in 2008 in Tampere
• By 2013, 86 companies and 1200 students
from 3 universities have participated in
250+ projects
• The Demola network is expanding
internationally
• This study focuses in Demola Tampere
25. Open innovation platform & ecosystem engager
established in 2008 in Tampere
By 2013, 86 companies and 1200 students from
3 universities have participated in 250+ projects
The Demola network is expanding internationally;
this study focuses in Demola Tampere
26. Challenges in measuring Demola
…and open innovation in general:
• Tradition: A linear view on innovation;
• Measuring inputs (money) and outputs
(patents, products, new companies);
• Survey-based methods, aggregate
measures
How does one measure the performance
of an ecosystem engager?
Still, K., Huhtamäki, J., Russell, M. & Rubens, N.
2012. Paradigm shift in innovation indicators—from
analog to digital. Proceedings of the 5th ISPIM
Innovation Forum, 9-12 December, Seoul, Korea.
27. Use case examples
Who wants to
measure?
Why do they want to measure? What will the do with the
measurement insights?
Policy makers Interested in the impact that Demola
has had to the surrounding ecosystem
Evaluate the utility of the
platform for future investments
and the applicability of the
approach
Company
representatives
Utility of Demola engament Decide whether to engage or
not; Select an approach suitable
for their portfolio
Demola operators Activity in general; Companies with
changing (increasing/decreasing)
Demola engagement; Ecosystem
Structure
General Demola introductions,
marketing & sales; Demola key
area development
University students Reviewing opportunities that
participating in a Demola project would
open
Decide whether to participate or
not
University decision-
makers
Impact, new developments in the
ecosystem
Add initiatives for students to
get involved
International actors Impact, engagement, transformation To evaluate the utility of the
process for deciding the
applicability of the approach
28. Method: data-driven network
analysis (& action research)
(Hansen et al., 2009)
Project Detail Example
Project Id Project 115
Name Koukkuniemi 2020
Started 2010-05-04
Ended 2010-10-31
Status Completed
Collaboration Partner City of Tampere
Type of Partner Public
Project Domain Non-profit
Location Tampere
Key Areas well-being, knowledge
management, regional
studies
Project Team Members uta, uta, tut, tut
29. Result 1: Project network
Nodes represent
projects and
companies
Company nodes are
light green; other
colors indicate
cluster membership
Node size shows its
betweenness value
Force-driven layout
30. Result 2: Project domain network
Nodes represent project
domains
Nodes are connected
through domain co-
occurence
Colors show cluster
membership
Node size shows its
betweenness
32. Discussion
• Technical challenges exist when using
internally collected data for network
visualization and animation
• Visualization development challenges data-
collection procedures and can add value to
existing data
• Demola operators find value particularly in
the animation of the project sphere;
international collaborators have also
expressed an interest in them
34. Acknowledgements & thank you
Time for your comments and questions.
Jukka Huhtamäki <jukka.huhtamaki @tut.fi>
Ville Luotonen
Ville Kairamo
Kaisa Still
Martha G. Russell
Acknowledgements Ville Ilkkala, Meanfish Ltd,
supported animation development. Heikki Ilvespakka
took care of exporting the data from the Demola platform
35. Innovation
ecosystem
Context
Data-driven
visualization
Process
Availability of relational data about innovation activities
(free, easily available public data)
Can be studied as networks (SNA)
Application arena Supporting insights on Highlighting
Network visualization Innovation indicators Indicator ”osoitin”
Network dynamics Relational capital
(Ecosystemic relational capital)
metrics
Various levels:
International
National
Local/regional
Organizational
36. Questions
What would your ecosystem look like based on the publicly
available data?
§ What info is there about you, your organization, your stakeholders– and
the connections between all these?
à Is this relevant for you? Could this have implications for some action?
Would the visualization of your ecosystem be valuable for you?
§ How?
§ What could you do better with that?
§ What could you do that you cannot do now?
Would knowing about your relational capital be valuable for you?
§ How?
§ What could you do better with that?
§ What could you do that you cannot do now?
Where could we find more relational data (easily
available public data, almost free)?
40. Framework
of
network
dynamics
(Ahuja
et
al
2009):
Operate
via
the
mechanisms
of:
• Homophily
• Heterophily
• Prominence
aorac,on
• Brokerage
• Closure
Microdynamics
of
networks
Network
Architecture
Dimension
Network
primi,ves
Micro-‐
founda,ons
of
networks:
Basic
factors
that
drive
or
shape
the
forma,on
and
content
of
,es
in
the
network:
• Agency
• Opportunity
• Iner,a
• Random
&
Exogenous
Causing
changes
in
network
membership
(through
dissolu,on
or
forma,on
of
,es,
changes
in
,e
content,
strength
and
mul,plexity)
Structure
-‐
Ego
network
• Centrality
• Contraint
-‐
Whole
network
• Degree
distribu,on
• Connec,vity
• Clustering
• Density
• Degree
assorta,vity
Content
• Types
of
flows
• Number
of
dis,nct
flows
(mul,plexity)
Architecture
of
any
network
can
be
conceptualized
in
terms
of:
• Nodes
(that
comprise
the
network)
• Ties
(that
connect
the
nodes)
• Structure
(the
paoerns
of
structure
that
result
from
these
connec,ons)
41. Dimensions
of
dynamics
Descrip&on
Meaning
Network
Architecture
Dimension
for
structure
Ego
network
Centrality
has
been
associated
with
a
wide
variety
of
poten,al
benefits
such
as
access
to
diverse
informa,on
and
higher
status
or
pres,ge
(Brass
1985)
Constraint
The
presence
of
structural
hole
is
commonly
related
to
brokerage
possibili,es
(Burt
1992,
Zaheer
and
Soda
2009)
Whole
network
Degree
Distribu,on
reflects
the
rela,ve
frequency
of
the
occurrence
of
,es
across
nodes
or
the
variance
in
the
distribu,on
of
,es
(Jackson
2008
has
been
used
to
signify
the
dis,bu,on
of
status,
power
or
pres,ge
across
organiza,ons
(Gula,
and
Caguilo,
1999;
Ahuja,
Polidoro
and
Mitchell
2009);
may
be
reflec,ve
of
changes
in
the
status
hierarchy
of
the
observed
system
(Ahuja
et
al
2009)
Connec,vity
Is
captured
in
the
diameter
of
a
network
which
in
turn
reflects
the
largest
path-‐distance
between
any
two
nodes
of
the
network
(Jackson
2008)
The
average
path
length
connec,ng
any
two
nodes
in
the
ntework
is
an
indicator
of
the
connec,vity
or
”small-‐wordness”
of
the
network;
as
network
becomes
more
”small-‐wordly”
informa,on
can
diffuse
more
quickly
fostering
outcomes
such
as
inova,on
or
crea,vity
(Schilling
2005,
Schilling
and
Phelps
2007);
as
the
path
length
between
any
two
nodes
of
a
network
diminishes,
it
is
possible
that
informa,on
can
become
more
decomra,zed
and
result
in
a
reduc,on
in
the
informa,onal
advantage
of
any
single
player
(Ahuja
et
al
2009)
Clustering
The
degree
to
which
the
network
is
formed
of
,ghtly
interconnected
cliques
(Ahuja
et
al
2009)
The
emergence
of
inter-‐connected
subgroups
or
cliques
suggests
that
the
network
is
being
differen,ated
into
a
variety
of
dis,nct
sub-‐
networks
or
communi,es
(Ahuja
et
al
2009);
at
inter-‐organiza,onal
level
this
may
represent
the
reclustering
of
clusters
or
constella,ons
of
firms
that
may
be
compe,ng
against
each
other
as
’alliance
network’
(Gomes-‐Cassares
1994);
clique
instability
maybe
a
precursor
of
a
significant
technological
discon,nuity
if
the
network
is
an
interorganiza,onal
technology
network,
or
perhaps
portend
an
imminent
change
in
the
power
structure
of
an
organiza,on
in
an
intraorganiza,onal
employee
network
(Ahuja
et
al
2009)
Density
The
propor,on
of
,es
that
are
realized
in
the
network
rela,ve
to
the
hypothe,cal
maximum
possible
(Ahuja
et
al
2009)
In
organiza,onal
segngs,
higher
network
density
may
be
reflec,ve
of
network
closure,
a
condi,on
that
in
turn
may
be
associated
with
the
development
of
norms;
increasing
density
could
be
reflec,ng
in
a
reduc,on
of
diversity
of
perspec,ves
and
choice
within
the
network
as
the
high
propor,on
of
realized
,es
provide
a
hologenizing
influnce
across
actors
,
and
thus
results
in
increasing
reifica,on
of
ideas
(Ahuja
et
al
2009)
Degree
Assorta,vity
The
degree
to
which
nodes
with
similar
degrees
connect
to
each
other
(Waos,
2004)
Posi,ve
assorta,vity
implies
that
high-‐degree
nodes
connect
to
other
high
degree
nodes
etc.
;
in
an
intra-‐organiza,onal
segng,
assorta,vity
could
be
driven
by
homophily
processes
and
disassorta,vy
by
complimentary
needs
(Ahuja
et
al
2009;
assorta,vity
can
be
associated
with
the
emergence
of
a
core-‐periphery
structure
(Borgag
and
Evereo
1999)
where
a
set
of
densely
connected
actors
cons,tute
a
core
of
an
industry
while
many
of
other
low
degree
actors
cons,tute
a
periphery.
Changes
might
signal
a
shik
in
the
resource
requirements
for
success
in
the
industry
(Powell,
Packalen
and
Whigngton
????)
Microfounda&ons–
d
Agency
Agency
behavior,
choosing
or
not
choosing
to
establish
connec,ons;
The
focal
actor’s
mo,va,on
and
ability
to
shape
rela,ons,
and
create
a
beneficial
link
or
dissolve
an
unprofitable
one
or
shape
an
advantageous
structure
(Sewell
1992;
Emirbayer
and
Goodwin
1994;
Emirbayer
and
Mische
1998)
As
actors
deliberately
seek
to
create
social
structures,
which
is
in
line
iwth
Burt’s
idea
of
structural
holes
as
socfial
capital,
highligh,ng
the
entrepreneurial
role
in
the
crea,on
of
this
valuable
form
os
social
structure
(Burt
1992)
à
Network
structures
emerbe
as
a
result
of
self-‐seeking
ac,ons
by
focal
nodes
and
their
connec,ons,
no,ng
that
actors
can
devise
unique
responses
to
imporve
their
own
situa,ons
in
the
network
(Ahuja
et
al
2009)
Opportunity
Reflects
the
structural
context
of
ac,on
(Blau
1994)
and
includes
the
argument
that
actors
tend
to
prefer
linking
within
groups
rather
than
across
them
(Li
and
Rowley
2002)
Iner,a
Includes
the
pressures
for
persistence
and
change
(Giddens
1984,
Portes
and
Sensenbrenner
1993,