Silverman Research: Collective Intelligence In Organisations Report

Collective
Intelligence

in
Organisations

Uses
and
Challenges

Executive
Report

2

London,
June
2015

This
report
provides
an
overview
of
Collective
Intelligence
in
organisations.
It
brings
together

perspectives
from
different
areas
of
study
in
what
is
a
diverse
and
multi-‐disciplinary
topic.
First

we
look
at
Collective
Intelligence
in
general
terms
–
what
it
is,
how
it
is
defined,
the
impact
of

social
and
digital
technologies
and
how
organisations
are
using
it
to
address
a
variety
of
business

issues.
Then
we
discuss
the
various
challenges
to
overcome
in
designing
and
using
Collective

Intelligence
tools
in
an
organisational
context.

The
report
centres
on
Collective
Intelligence
systems
and
tools
that
involve
some
form
of

discussion,
ideation
or
interaction
as
well
as
interactive
voting
or
rating
mechanisms.
This
is

because
tools
that
possess
these
functions
are
reasonably
common
and
also
that
they
are
often

the
most
complex
in
terms
of
design,
administration
and
analysis.

One
issue
that
emerges
quite
strongly
is
that
not
all
Collective
Intelligence
is
created
equally
–

there
are
levels
of
Collective
Intelligence.
Simple
survey
methodologies
do
provide
Collective

Intelligence
of
a
sort,
but
what
they
lack
is
a
social
element
that
enables
participants
to
evaluate

each
other’s
responses.

This
is
important
because
when
this
type
of
aggregation
is
provided
it
is
possible
to
obtain
Social

Collective
Intelligence
-‐
and
that’s
quite
a
different
proposition
in
terms
of
the
richness
of
insight

produced.
If
a
group
is
interacting
in
a
system
that
has
been
carefully
designed
to
optimise
their

input
and
extract
insight
then
the
level
of
Collective
Intelligence
obtained
can
be
far
greater
than

any
survey.

I
hope
the
report
is
useful
to
people
working
in
Market
Research,
Digital
Social
Innovation,

Computer
Science,
Internal
Communications
or
Human
Resources.
Thanks
to
Milly
Picton
for
help

with
the
report.

Michael
Silverman

Managing
Director

Silverman
Research

3

Contents

Executive
Summary

4

1.
Introduction

6

1.1
What
is
Collective
Intelligence?

1.2
The
Emergence
of
Social
Collective
Intelligence

1.3
The
Wisdom
of
Crowds
and
Crowdsourcing

2.
Using
Collective
Intelligence

9

2.1
Ideation
and
Innovation

2.2
Knowledge
Sharing
and
Workload
Distribution

2.3
Predicting
the
Future

2.4
Coordination
and
Collaboration

3.
Challenges
for
Collective
Intelligence

11

3.1
Design
Challenges

3.2
Socio-‐Cognitive
Challenges

3.3
Administrative
Challenges

3.4
Managerial
and
Leadership
Challenges

4.
Conclusion

19

4

Executive
Summary

What
is
Collective
Intelligence?

The
Internet
is
giving
rise
to
some
remarkable
technologies
that
enable
large
groups
of
people
to

collaborate
online.
This
is
having
a
huge
impact
in
the
field
of
Collective
Intelligence
as
it
provides

an
opportunity
to
progress
research
and
practice
in
this
area.
There
is
a
need
to
understand

what’s
important
in
mass
online
interactions
and
discussions
because
intelligence
doesn’t
just

reside
within
us
as
individuals
–
it
also
emerges
from
groups
of
people.
Collective
Intelligence
is
a

broad
area
and
the
term
is
often
applied
quite
liberally
to
a
variety
of
activities
or
systems.

In
the
last
ten
years
there
has
been
an
explosion
in
both
research
and
interest.
The
main
reason

for
this
burst
of
enthusiasm
is
the
rapid
advance
of
social
and
digital
technologies
in
both

business
and
in
the
public
sphere.
These
changes
reflect
the
rise
of
a
new
era
for
Collective

Intelligence,
one
that
is
able
to
fully
take
advantage
of
both
the
increasing
interconnectivity

between
people
and
also
the
increasing
computing
and
storage
capabilities
of
advanced

information
and
communication
technologies.

The
rise
of
social
technologies
has
facilitated
the
rise
of
Collective
Intelligence
because
before

this
it
was
almost
impossible
for
large
and
dispersed
groups
of
people
to
organise
themselves
and

interact.
Today,
people
can
participate
within
technological
systems
that
are
increasingly
able
to

orchestrate
the
collection
and
analysis
of
human
social
activity.
This
is
the
frontier
of
Social

Collective
Intelligence:
networks
of
people
and
computers
acting
together
in
intelligent
ways.

The
term
Crowdsourcing
is
synonymous
with
Collective
Intelligence
and
the
two
are
often
used

interchangeably.
Crowdsourcing
can
be
defined
in
different
ways,
although
it’s
frequently

defined
as
taking
a
task
that
is
typically
conducted
by
one
person
and
outsourcing
it
to
a
large

group
of
people.
Both
Collective
Intelligence
and
crowdsourcing
involve
groups
acting
together

online,
however,
there
are
differences.
Crowdsourcing
can
enable
Collective
Intelligence,

although
that
is
not
always
the
case.
Crowdsourcing
is
usually
more
of
a
process,
the
goal
of

which
is
to
distribute
workload
from
one
to
many.
Collective
Intelligence,
on
the
other
hand,
can

be
better
thought
of
as
a
goal.
The
focus
is
on
the
output.

Using
Collective
Intelligence

There
are
many
areas
of
work
and
types
of
task
to
which
Collective
Intelligence
methods
can
be

applied.
An
area
in
which
Collective
Intelligence
methods
are
commonly
used
is
for
generating

ideas,
discussions
and
solutions
with
networks
of
employees,
customers
and
other
external

parties.
Sometimes,
there
can
be
a
competitive
element
or
other
incentives
for
participation.

In
addition
to
allowing
organisations
to
gather
ideas
and
feedback
from
a
greater
number
and

diversity
of
people,
Collective
Intelligence
approaches
also
help
organisations
to
act
on
those

insights
by
distributing
work
to
people
who
are
best
placed
to
do
it.
This
not
only
has
the

potential
to
enhance
quality,
but
also
to
increase
efficiency
as
tasks
can
be
conducted
in
parallel

by
many
people.
Collective
Intelligence
can
also
be
used
to
predict
the
outcomes
of
future

events.
Aggregating
the
diverse
perspectives,
knowledge
and
experience
of
employees
and

customers
can
improve
the
accuracy
of
predictions.

5

Challenges
for
Collective
Intelligence

Real-‐world
implementations
have
taught
us
a
great
deal
about
some
of
the
obstacles
and
pitfalls

that
need
to
be
addressed
in
the
development
and
use
of
Collective
Intelligence
tools.
The

challenges
for
Collective
Intelligence
come
in
a
variety
of
forms,
which
can
be
roughly
divided

into
four
groups.
These
are
challenges
in
Design,
Social-‐Cognition,
Administration
and
Leadership.

Researchers
who
are
involved
with
the
study
of
Human-‐Computer
Interaction
(HCI)
and

Collective
Intelligence
focus
on
how
people
and
computers
can
be
connected
to
enhance

usability
and
output.
The
design
choices
made
regarding
the
user
interface
of
tools
are
extremely

important.
Potential
participants
will
only
get
involved
if
the
user
interface
guides
them
in
a

straightforward
and
meaningful
way.
Similarly,
those
interested
in
leveraging
Collective

Intelligence
require
a
means
by
which
to
administrate,
analyse
and
extract
insights.
Ultimately,

Collective
Intelligence
requires
the
design
of
both
technical
infrastructure
and
human-‐human

interaction
–
a
socio-‐technical
system.

It
seems
almost
intuitive
to
us
that
a
group
will
probably
make
a
better
decision
than
an

individual
and
it’s
commonly
thought
that
the
performance
of
a
group
will
be
equivalent
to
that

of
its
best
members.
While
these
assumptions
seem
obvious,
however,
there
also
needs
to
be

acknowledgement
that,
under
some
circumstances,
groups
can
make
terrible
decisions.
Various

social
pressures
and
cognitive
biases
can
affect
the
way
people
behave
in
a
group.
In
certain

circumstances
this
can
be
disastrous.

There
are
also
administrative
and
operational
hurdles
to
overcome
in
obtaining
Collective

Intelligence.
Research
has
made
a
concerted
effort
to
understand
how
best
to
grow
and
maintain

active
participation
in
online
communities.
Overcoming
these
challenges
is
crucial
because

human
input
is
fundamental
to
Collective
Intelligence
–
without
enough
fuel
(and
the
right
sort
of

fuel),
the
engine
cannot
run.
Collective
Intelligence
in
organisations
poses
some
specific

challenges
for
management
and
leadership
because
tapping
into
Collective
Intelligence
means

taking
a
more
open
and
transparent
approach
–
and
this
requires
significant
cultural
change.

Collective
Intelligence
is
nothing
new.
Successful
organisations
have
been
encouraging

collaboration
and
interaction
since
the
first
modern
organisations
were
formalised.
However,
it
is

Social
Collective
Intelligence,
resulting
from
the
rapid
development
of
social
and
digital

technologies,
that
is
both
new
and
exciting
in
this
area.
Organisations
are
only
just
beginning
to

appreciate
the
potential
that
Collective
Intelligence
can
offer.
There
are
pockets
of
good
practice

appearing
as
organisations
start
to
implement
various
Collective
Intelligence
activities.
However,

these
early
examples
of
Social
Collective
Intelligence
are
likely
to
be
just
the
start
of
the
story.

6

1.
Introduction

The
Internet
has
given
rise
to
some
remarkable
technologies
that
enable
people
to
collaborate
en

masse.
In
barely
a
decade,
the
biggest
encyclopaedia
in
human
history
has
been
written
by

millions
of
authors
with
no
centralised
control.
In
under
a
year,
ordinary
people
classified
more

than
50
million
photos
of
distant
worlds
to
help
astronomers
understand
how
galaxies
are

formed.
All
over
the
world,
organisations
are
starting
to
open
up
social
networks
and

collaborative
environments
from
which
emanate
an
endless
stream
of
data
and
insight.
The

potential
that
lies
within
interconnected
networks
of
computers
and
humans
is
only
just
starting

to
be
realised.

1.1
What
is
Collective
Intelligence?

We
all
possess
intelligence,
however,
intelligence
can
be
thought
of
not
only
as
something
that

arises
within
human
brains
–
it
also
arises
in
groups
of
people.
This
is
Collective
Intelligence
–

individuals
acting
together
to
combine
their
knowledge
and
insight.
Collective
Intelligence
is
an

emergent
property.
It
emerges
from
the
group,
it
does
not
reside
within
any
individual
members.

Defining
Collective
Intelligence
is
troublesome
because
of
the
historical
debate
surrounding
the

definition
of
intelligence
as
a
concept
on
its
own
right.
This
situation
is
exacerbated
because
the

study
of
Collective
Intelligence
is
not
a
well-‐defined
area
–
the
research
literature
provides
a

variety
of
definitions.
This
is,
in
a
large
part,
due
to
the
range
of
people
who
are
interested
in

Collective
Intelligence
–
it
is
a
multi-‐disciplinary
topic
of
study
which
has
received
attention
from

researchers
in
Psychology,
Computer
Science,
Animal
Behaviour,
Human
Computer
Interaction,

Crowdsourcing
and
Artificial
Intelligence
to
name
but
a
few.
Despite
their
differences,
what

draws
these
disciplines
together
is
a
focus
on
creating
or
exploring
tools,
systems
and
platforms

that
give
rise
to
Collective
Intelligence.

1.2
The
Emergence
of
Social
Collective
Intelligence

Although
the
term
Collective
Intelligence
is
thought
to
have
materialised
in
the
1800s,
the
first

academic
references
didn’t
appear
until
the
1970s.
However,
in
the
last
10
years
there
has
been

an
explosion
in
both
research
and
interest.
The
main
reason
for
this
burst
of
enthusiasm
is
the

rapid
advance
of
social
and
digital
technologies
in
both
business
and
in
the
public
sphere.

Within
organisations,
openness
and
transparency
are
becoming
vital
business
characteristics.

Similarly,
considering
society
at
large,
the
subjects
of
Digital
Social
Innovation
and
Open

Democracy
are
growing
in
prominence
as
they
aim
to
capture
how
contemporary
techno-‐social

trends
can
be
harnessed
towards
solving
the
most
important
challenges
facing
modern
societies.

These
changes
reflect
the
rise
of
a
new
era
for
Collective
Intelligence,
one
that
is
able
to
fully
take

advantage
of
both
the
increasing
interconnectivity
between
people
and
also
the
increasing

computing
and
storage
capabilities
of
advanced
information
and
communication
technologies.

This
is
the
frontier
of
Social
Collective
Intelligence:
networks
of
people
and
computers
acting

together
in
intelligent
ways.

7

The
rise
of
social
technologies
has
facilitated
the
rise
of
Collective
Intelligence
because
before

this
it
was
almost
impossible
for
large
and
dispersed
groups
of
people
to
organise
themselves
and

interact.
Today,
people
can
participate
within
technological
systems
that
are
increasingly
able
to

orchestrate
the
collection
and
analysis
of
human
activity.
Real
world
interactions
between

humans
are
also
subject
to
many
social
constraints,
of
which
technology
is
helping
to
overcome.

As
a
result,
we
are
witnessing
the
creation
of
new
and
exciting
ways
of
enhancing
collaboration

and
generating
Collective
Intelligence.

The
term
social
machine
has
also
been
used
to
describe
Collective
Intelligence
systems
and
tools.

It
is
helpful
to
think
about
Collective
Intelligence
emanating
from
social
machines
because
it

emphasises
the
joint
involvement
of
both
computers
and
the
people
using
them
as
part
of
the

machinery.
The
technology
is
the
engine
and
the
people
are
the
fuel.
When
a
particular
social

media
platform
(eg
Twitter,
Facebook
or
LinkedIn)
is
described
as
a
social
machine,
it’s
the

platform
and
the
people
who
use
it
that
constitute
the
social
machine.

Social
machines
involve
the
participation
of
humans
and
technological
components,
with
the

expectation
that
each
component
will
provide
their
distinctive
capabilities
to
complement
each

other.
For
some
social
machines,
it
is
humans
that
perform
the
clever
or
creative
work
and
the

machine
that
does
the
administration.
However,
the
likelihood
is
that,
in
the
future,
the

technology
will
increasingly
take
on
more
human
tasks
and
will
not
be
there
just
to
organise

human
input.

1.3
The
Wisdom
of
Crowds
and
Crowdsourcing

The
Wisdom
of
Crowds
is
simply
another
term
for
Collective
Intelligence.
It
stems
from
work
the

British
scientist
Sir
Francis
Galton.
At
a
livestock
exhibition
in
1906,
Galton
observed
a

competition
where
people
had
to
guess
the
weight
of
an
ox.
At
the
end
of
the
contest,
Galton

gathered
all
the
guesses
and
calculated
the
average
number
–
this
is
essentially
a
simple
survey

methodology.
The
crowd’s
estimate
turned
out
to
be
near
perfect
(the
crowd’s
estimation
was

1,197
pounds
and
the
ox’s
actual
weight
was
1,198).
Galton
referred
to
this
as
the
collective

wisdom
of
the
crowd:
the
fact
that
groups
of
people
can
be
more
intelligent
than
an
intelligent

individual
and
that
groups
do
not
always
require
intelligent
people
to
reach
a
smart
decision
or

outcome.
Although,
as
we
will
see,
the
simplicity
of
asking
people
to
guess
a
number
bears
little

resemblance
to
the
real-‐world
complexity
involved
in,
for
example,
analysing
thousands
of
peer-‐
evaluated
comment
threads
in
an
online
discussion.

The
term
crowdsourcing
is
synonymous
with
Collective
Intelligence
and
the
two
are
often
used

interchangeably.
Crowdsourcing
can
be
defined
in
different
ways,
although
it’s
frequently

defined
as
taking
a
task
that
is
typically
conducted
by
one
person
and
outsourcing
it
to
a
large

group
of
people.
Both
Collective
Intelligence
and
crowdsourcing
involve
groups
acting
together

online,
however,
there
are
differences.
Crowdsourcing
can
enable
Collective
Intelligence,

although
it
doesn’t
always.
Crowdsourcing
is
usually
more
of
a
process,
the
goal
of
which
is
to

distribute
workload
from
one
to
many.
Collective
Intelligence,
on
the
other
hand,
can
be
better

thought
of
as
a
goal.
The
focus
is
on
the
output.

Crowdsourcing
can
be
categorised
into
three
broad
types:
Directed,
Self-‐Directed
or
Passive.
In

directed
crowdsourcing
activities,
a
requestor
(the
person
or
group
running
the
exercise)
directs

8

the
input
of
participants
(crowd
workers)
by
providing
a
specific
brief
or
asking
a
particular

question.
There
is
usually
is
an
external
incentive
for
participation,
which
can
either
be
social
(eg

recognition
for
coming
top
of
the
leader
board)
or
financial
(eg
vouchers
or
money).

Self-‐directed
crowdsourcing
activities
are
characterised
by
the
absence
of
a
requestor,
coupled

with
an
intrinsic
motivation
for
participation.
This
is
where
communities
come
together,
of
their

own
volition,
to
discuss
various
topics
or
decide
on
a
specific
course
of
action.
On
occasion,

crowdsourcing
can
also
be
passive
–
this
is
where
the
output
is
a
side
effect
of
the
work
being

done
–
the
output
is
not
an
explicit
objective
of
the
participant.
For
example,
in
spam
protection,

reCAPTCHA
is
a
simple
tool
found
on
many
websites
where
people
have
to
copy
a
simple
code
to

prove
they
are
human.
Yet,
this
offers
more
than
just
spam
protection,
each
one
completed
can

help
digitise
text
or
annotate
images
even
though
the
participant
is
not
aware
of
this
passive

objective.

9

2.
Using
Collective
Intelligence

There
are
many
areas
of
work
and
types
of
task
to
which
Collective
Intelligence
methods
can
be

applied.
These
are
outlined
below,
grouped
according
to
the
organisational
challenges
they

address.
There
are
a
number
of
different
ways
in
which
various
uses
can
be
categorised
or

described
–
and
there
is
also
a
great
deal
of
overlap
between
uses.
These
methods
are
not

mutually
exclusive,
the
most
appropriate
approach
may
involve
a
combination
of
methods.

2.1
Ideation
and
Innovation

One
of
the
areas
in
which
Collective
Intelligence
methods
have
been
most
commonly
used
is
for

generating
ideas
and
solutions
for
an
infinite
range
of
issues.
Organisations
are
applying

Collective
Intelligence
tools
with
networks
of
employees,
customers
and
other
external
parties

for
a
range
uses
within
the
area
of
ideation
and
innovation.
For
example:

Discussion
and
commentary

Many
Collective
Intelligence
tools
consist
of
virtual
environments
where
participants
can
interact

to
discuss
ideas
and
opinions
or
provide
feedback
on
particular
topics.
Discussions
can
be
on-‐
going
or
fixed
for
a
specific
length
of
time.
Typically,
participants
can
make
suggestions,
receive

feedback
on
their
suggestions
from
other
participants,
as
well
as
rate
and
comment
upon
the

suggestions
of
others.
As
discussed
later,
community
ratings
and
commentary
on
participants’

suggestions
can
be
analysed
identify
themes
and
comments
that
resonate
most
with
the

community.

Competitions
and
challenges

Here,
people
are
invited
to
submit
ideas,
designs
or
solutions
in
response
to
a
given
brief.

Participants
(individuals
or
groups)
then
compete
to
provide
a
winning
entry
that
is
judged
by
the

person
or
group
that
has
initiated
the
competition
(the
requestor).
In
some
instances,

participants’
offerings
can
also
be
evaluated
and
rated
by
other
participants.
However,
there
is

typically
no
interaction
between
participants
in
the
actual
creation
of
the
submission.

Collaborative
design
and
insight
communities

Unlike
competitions
and
challenges,
the
focus
of
collaborative
design
is
on
participants
working

together
to
produce
designs
and
solutions.
Participants
can
evaluate
each
other’s
input
and
offer

suggestions
for
improvement.
Shared
interests
can
also
bring
people
together
in
a
community
to

enhance
knowledge,
share
best
practice
and
propose
new
ideas.
Various
collaborative
tools
are

often
used
to
support
insight
communities
by
enabling
social
networking
and
discussion.

2.2
Knowledge
Sharing
and
Workload
Distribution

In
addition
to
allowing
organisations
to
gather
ideas
and
feedback
from
a
greater
number
and

diversity
of
people,
Collective
Intelligence
approaches
also
help
organisations
to
act
on
those

insights
by
distributing
work
to
people
who
are
best
placed
to
do
it.
This
not
only
has
the

potential
to
enhance
quality,
but
also
to
increase
efficiency
as
tasks
can
be
conducted
in
parallel

by
many
people
at
the
same
time.
The
main
techniques
are
outlined
below:

10

Parallel
task
processing

This
is
a
feature
of
many
Collective
Intelligence
tools
whose
approach
is
to
reduce
complex

problems
into
simple
tasks
that
can
be
worked
on
simultaneously
by
a
large
number
of

individuals.
This
approach
works
especially
well
for
tasks
that
computers
find
extremely
difficult

to
do.
For
example,
labelling
large
numbers
of
uncatalogued
photographs
or
categorising

ambiguous
photos
of
outer
space.
Task
can
be
completed
very
quickly
in
this
way.

Distributed
questions
and
answer
(Q&A)

This
involves
posting
questions
in
a
virtual
forum
that
can
be
answered
by
other
participants.

Participants
are
then
able
to
further
elaborate
and
rate
answers.
Many
organisations
use

distributed
Q&A
to
reduce
direct
support
costs
and
provide
answers
to
common
questions.
For

example,
before
providing
contact
details
for
support,
an
organisation
will
often
direct
customers

to
a
community
forum
where
they
can
see
if
other
people
have
had
a
similar
problem.

Games

Although
good
examples
of
this
are
uncommon,
online
simulations
and
games
can
allow

participants
to
apply
their
knowledge
and
expertise
to
complex
problems.
One
good
example
is

Foldit,
where
an
online
game
was
used
to
engage
scientists
in
uncovering
the
structure
of

particular
proteins.
Using
a
protein-‐folding
game
called
Foldit,
thousands
of
scientists
from
a

variety
of
educational
backgrounds
were
able
to
solve
a
problem
that
eluded
scientist
for
years.

2.3
Predicting
the
Future

Collective
Intelligence
can
also
be
used
to
predict
the
outcomes
of
future
events.
Aggregating

diverse
perspectives,
knowledge
and
experience
of
employees
and
customers
can
improve
the

accuracy
of
predictions.
This
allows
organisations
to
make
more
informed,
evidence
based

decisions
that
can
complement
traditional
forecasting
approaches.
For
example,
in
prediction

markets,
participants
are
typically
given
tokens
or
toy
currency
to
invest
in
the
likelihood
of

future
events
or
outcomes
(eg
for
a
new
product’s
future
success
or
the
result
of
an
election).

Analysis
can
then
determine
forecast
probabilities
for
the
various
outcomes.
Participants
can
also

be
rewarded
on
the
basis
of
the
accuracy
of
their
predictions.

2.4
Coordination
&
Collaboration

Coordination
is
when
a
community
is
created
quickly
with
the
aim
spreading
information
quickly

and
widely.
Increasing
connectivity
means
that
groups
are
able
to
cooperate
and
achieve
what

once
was
the
exclusive
domain
of
large
centralised
organisations.
In
organised
industrial

disputes,
social
media
has
the
power
to
rapidly
organise
and
mobilise
employees
–
even
when

they
are
geographically
or
temporally
dispersed.
Cloud-‐based
collaborations
and
productivity

suites
such
as
Google
Apps
for
Work
and
Dropbox,
Internal
social
networks
like
Yammer
or

Chatter,
or
other
social
collaborative
task-‐based
tools
like
Asana
or
Huddle,
are
becoming
wide

spread.
Clearly
there
is
a
lot
of
difference
across
such
tools,
although
they
all
have
a
focus
on

connecting
people,
allowing
people
to
interact
and
to
help
with
collaboration,
communication,

project
management
and
many
others.
These
tools
have
the
potential
to
contribute
to
Collective

Intelligence
in
lots
of
different
ways
(eg
discussion,
idea
generation,
distributing
work,
knowledge

sharing
and
Q&A).

11

3.
Challenges
for
Collective
Intelligence

Although
Collective
Intelligence
has
evolved
alongside
humans,
the
rise
of
social
technologies

over
the
last
decade
has
caused
the
rapid
development
of
a
wide
range
of
tools,
systems
and

platforms.
Real-‐world
implementations
over
this
time
have
taught
us
a
great
deal
about
some
of

the
obstacles
and
pitfalls
that
need
to
be
addressed
in
the
development
and
use
of
Collective

Intelligence
tools.
The
challenges
for
Collective
Intelligence
come
in
a
variety
of
forms.
These
can

be
roughly
divided
into
four
groups:
Challenges
in
Design,
Social-‐Cognition,
Administration
and

Leadership.
These
are
discussed
in
turn
below.

3.1
Design
Challenges

Researchers
who
are
involved
with
the
study
of
Human-‐Computer
Interaction
(HCI)
and

Collective
Intelligence
focus
on
how
people
and
computers
can
be
connected
to
enhance

usability
and
output.
The
design
choices
made
regarding
the
user
interface
of
tools
are
extremely

important
because
potential
participants
will
only
get
involved
if
the
user
interface
guides
them

in
a
straightforward
and
meaningful
way.
Similarly,
those
interested
in
leveraging
Collective

Intelligence
require
a
means
by
which
to
administrate,
analyse
and
extract
insights
from
the

system.
Ultimately,
Collective
Intelligence
requires
the
design
of
both
technical
infrastructure
and

human-‐human
interaction:
a
socio-‐technical
system.

This
is
a
salient
point
because
Social
Collective
Intelligence
tools
must,
by
their
very
nature,

involve
some
means
of
interaction
between
participants.
This
typically
takes
the
form
of
an

online
discussion
or
comments
section
where
participants
provide
their
own
input
as
well
as

exploring,
evaluating,
or
replying
to,
the
input
of
others.
Online
discussions,
also
known
as

discussion
forums,
message
boards
or
comment
sections,
provide
participants
with
the
ability
to

have
conversations
in
the
form
of
posted
messages.
Online
discussions
are
almost
always

hierarchical,
or
tree-‐like,
in
structure
and
are
typically
displayed
in
a
list
format
or
thread.
Online

discussions
are
ubiquitous
across
the
Internet
–
most
websites,
whether
they
are
news
sites,

blogs,
review
sites
or
social
networks
will
have
some
form
of
discussion
forum
or
comment

section.

It
is
worth
considering,
for
a
moment,
the
evolution
of
online
commentary
and
discussion.
In
the

late
1970s,
even
before
the
Internet
was
established,
simple
bulletin-‐board
systems
were
being

developed.
Users
were
able
to
look
at
a
list
of
posts
and
if
they
found
something
interesting
they

could
contribute
to
the
post.
This
was
the
first
time
that
user-‐generated
content
was
the

fundamental
element
of
a
software
application.

Over
the
years,
as
various
technological
advances
were
made,
the
functionality
of
online

discussions
evolved
into
the
more
mainstream
social
technologies
that
we
see
today.
Most

modern
online
commentary
or
discussions
still
contain
basic
elements
that
were
produced
early

in
the
creation
of
the
Internet
and
this
poses
some
difficult
challenges
for
Collective
Intelligence.

These
are
outlined
below:

12

List-‐based
discussions

One
of
the
most
basic
challenges
is
that,
due
to
the
accessibility
of
social
technologies,
Collective

Intelligence
tools
need
to
be
able
to
cope
with
large
groups
of
people.
This
may
sometimes
be
as

small
as
50
or
100
people,
but
it
could
also
be
thousands,
or
even
millions,
or
people.
As
a

consequence,
Collective
Intelligence
tools
that
present
participants’
input
as
comments
or
posts

in
a
list
have
serious
limitations.
The
worst
is
that
they
quickly
become
impossible
to
navigate
and

can
quickly
grow
to
overwhelming
proportions.
Interacting
with
other
people
in
a
list-‐based

discussion
is
ok
from
a
user
perspective
if
there
are
a
small
number
of
people
(say
20

participants).
However,
when
the
number
of
comments
grows
beyond
this
it
becomes
an

enormous
problem.

Without
a
means
of
effective
navigation,
many
comments
or
ideas
will
never
be
properly

evaluated
–
participants
do
not
have
an
equal
chance
of
being
heard.
In
online
discussions,
lists

of
comments
tend
to
hide
the
true
diversity
of
opinion
that
exists,
favouring
early
birds
or
those

with
extreme
views.
Technological
advances
in
graphical
user
interfaces
and
gamification
are

enabling
more
engaging
ways
for
employees
to
express
their
opinions
about
various
work-‐related

issues.
Social
media
that
use
these
techniques
have
huge
advantages
over
traditional
lists
of

comments
used
in
discussion
forums
and
message
boards.
Algorithms
that
share
out
comments

are
also
becoming
increasingly
sophisticated
and
can
ensure
each
comment
has
been
given
an

equal
amount
of
attention
by
the
community.
Interactive
visualisations
also
provide
a
compelling

and
engaging
experience
for
participants
who
can
themselves
get
instant
and
authentic
feedback.

This
can
help
to
increase
participation
rates
in
more
discussion-‐based
interactions.

Statistical
confidence
of
evaluations

Employing
a
system
that
ensures
an
equal
distribution
of
participants’
input
means
that
the

subsequent
analysis
can
be
more
statistically
confident
that
each
idea’s
evaluation
ratings
are

valid
and
really
do
reflect
the
views
of
the
community.
For
example,
if
an
idea
has
only
been

evaluated
by
a
handful
of
people,
the
system
may
need
to
obtain
more
ratings
to
be
confident

that
the
community
really
do
feel
that
way
about
a
particular
idea.
Collective
Intelligence
systems

can
apply
algorithms
that
include
dynamic
control.
These
work
by
deciding
whether
a
comment

requires
more
ratings
to
reach
a
specific
confidence
interval
and
adjusts
the
frequency
with

which
that
comment
is
displayed.

Inadequate
evaluation
mechanisms

The
methods
through
which
participants
evaluate
each
other’s
ideas
and
suggestions
can
make
a

big
difference
to
the
output
obtained
from
Collective
Intelligence
tools.
Typically,
evaluations
are

obtained
by
rating,
voting,
liking
or
assigning
a
proportion
of
virtual
tokens.
However,
as
outlined

above,
when
there
are
thousands
of
ideas
displayed
in
a
list,
only
the
ones
towards
the
top
of
the

list
that
will
be
read
and
evaluated.
Some
systems
allow
participants
to
sort
the
list
by
the
ratings

they
have
received,
however,
this
is
still
problematic
because
it
is
only
the
ideas
near
the
top
of

the
list
get
evaluated.

Some
evaluation
mechanisms
are
very
basic.
The
most
basic
evaluation
criteria
are
binary

mechanisms
such
as
a
simple
like
or
upvote
button.
The
trouble
with
these
is
that
they
are
over-‐
simplistic
and
do
not
provide
any
detail
about
what
it
actually
is
that
participants
like
about
an

idea
–
is
it
because
they
agree
with
it?
Or
that
they
think
it
is
insightful?
Or
that
it
will
be
easy
to

implement?
We
just
don’t
know.
Asking
for
ratings
using
more
detailed
response
scales
can
help,

13

although
they
still
are
only
able
to
assess
one
particular
evaluation
criteria.
Using
multiple

response
scales
that
refer
to
different
evaluation
criteria
can
help
enormously
in
identifying
the

best
ideas
or
comments.

Aggregation

In
order
to
obtain
Collective
Intelligence,
there
must
be
mechanisms
that
transform
private

judgements
into
collective
decisions
–
this
is
aggregation.
The
interconnectivity
offered
by

modern
social
technologies
has
huge
implications
for
the
way
in
which
individuals’
input
is

translated
into
collective
decisions.
There
are
two
main
ways
of
aggregating
information
or

knowledge
in
organisations.
The
most
common
way
of
reaching
a
collective
decision
is
by
adding

all
of
the
individual
responses
together.
For
example,
a
survey
can
provide
a
job
satisfaction
score

for
the
organisation
by
simply
providing
an
average
of
individual
responses
(a
similar
approach

we
saw
with
Galton
and
his
livestock).

It
seems
reasonable
to
assume
that
distributing
a
survey
questionnaire
and
analysing
the
results

will
produce
Collective
Intelligence
–
and,
to
some
extent,
it
does.
With
a
traditional
survey,
a

group
of
people
are
asked
to
provide
feedback
in
isolation
and
these
individual
pieces
of

feedback
are
then
aggregated
by
a
researcher
who
will
analyse
the
data
and
report
the
findings.

The
intelligence
gained
from
traditional
surveys
can
be
considered
‘collective’
in
the
sense
that

the
data
is
being
provided
by
multiple
persons
within
the
group.
The
findings
of
a
survey
can
also

be
considered
‘emergent’
because
the
intelligence
produce
does
not
reside
within
any
members

of
the
group.
However,
it
can
be
argued
that
the
level
of
Collective
Intelligence
provided
by

traditional
surveys
is
limited
–
the
crux
of
the
matter
is
the
means
by
which
the
data
is

aggregated.

Advancements
in
technology
have
allowed
new
means
of
aggregation
to
emerge.
In
particular,

through
social
technologies,
large
groups
of
people
can
rate
each
other’s
written
responses
on

specific
evaluation
criteria.
When
these
ratings
are
analysed
along
the
text
(using
modern
text

analysis
software)
it
is
possible
to
identify
which
comments
and
themes
resonate
most
with
the

community.
As
a
result,
the
output
is
not
shaped
by
an
individual
analyst
or
team
of
researchers,

it
is
shaped
authentically
by
members
of
the
group.
Aggregation
in
social
technologies
marks

such
a
shift
because
aggregators
can
be
used
to
help
shape
a
community’s
voice,
as
well
as

represent
and
analyse
it.
It
is
only
through
these
new
technologies
and
algorithms
that
we
are

starting
to
see,
for
the
first
time,
genuinely
bottom–up,
large-‐scale
and
collective
opinion

forming,
rather
than
building
responses
to
questions
that
reflect
the
constraints
of
a
particular

researcher’s
thinking.

Time
wasting

In
online
discussions
that
have
large
numbers
of
participants,
the
same
ideas
are
often
repeated

and
this
can
sometimes
prevent
the
discussion
from
progressing.
Similarly,
it
can
be
difficult
to

organise
and
categorise
the
enormous
amount
of
ideas.
Too
much
time
can
also
be
wasted

discussing
suggestions
that
are
unworkable
or
impractical,
meaning
that
good
ideas
can

sometimes
be
overlooked
or
underdeveloped.

14

3.2
Socio-‐Cognitive
Challenges

The
idea
of
synergy
–
that
the
whole
goes
beyond
the
sum
of
its
parts
–
provides
a
powerful

incentive
to
consult
a
group
rather
than
an
individual.
It
seems
almost
intuitive
to
us
that
a
group

will
make
a
better
decision
than
an
individual
and
it’s
commonly
thought
that
the
performance
of

a
group
will
be
equivalent
to
that
of
its
best
members.
While
these
assumptions
seem
obvious,

however,
there
also
needs
to
be
acknowledgement
that,
under
some
circumstances,
groups
can

make
terrible
decisions.

Various
social
pressures
and
cognitive
biases
can
affect
the
way
people
behave
in
a
group,
and
in

certain
circumstances
this
can
be
disastrous.
Groups
can,
and
often
will,
correct
individual

mistakes
in
order
to
arrive
at
the
best
final
decision,
but
sometimes
group
psychology
can
act
to

amplify
these
mistakes.
In
many
cases,
these
effects
are
more
pronounced
in
face-‐to-‐face
group

situations.
Nevertheless,
the
impact
of
socio-‐cognitive
factors
can
also
be
felt
in
online

interactions
(depending
on
the
use,
design
and
context
of
the
exercise).
Some
of
the
most

important
factors
to
be
aware
of
are
outlined
below:

Group
decision-‐making

Social
pressures
felt
within
a
group,
such
as
social
conformity
and
obedience,
can
lead
members

to
keep
quiet
in
order
to
avoid
negative
penalties
such
as
disapproval
from
peers
and
isolation

within
the
group.
This
is
further
exacerbated
by
participants’
interpretation
of
information
signals

within
the
group
(ie
who’s
said
what,
and
when).
These
signals
can
lead
people
to
hold
back

information
they
know
out
of
respect
to
those
who
have
already
shared
their
insights.
Take
for

example,
an
authoritative
figure
passionately
outlining
their
argument
at
the
beginning
of
a

meeting
–
it’s
likely
that
a
number
of
group
members
will
subsequently
hold
their
tongues.
These

influences
serve
to
exacerbate
a
number
of
problems
that
arise
in
groups.
Errors
are
soon

amplified
rather
than
corrected
by
those
avoiding
social
pressures
or
respecting
and
accepting

information
provided
by
other
group
members.
Groups
can
also
become
victim
to
cascade
effects

–
this
is
when
a
group
accepts
the
decisions
or
follows
the
actions
of
those
who
spoke
first,
even

if
these
decisions
or
actions
are
wrong.

Cognitive
biases

Cognitive
biases
are
mistakes
in
cognitive
processing
that
can
lead
to
illogical
or
irrational

decision-‐making.
These
biases
can
result
as
a
consequence
of
the
limited
processing
capacity
of

our
brains,
but
they
also
enable
us
to
make
quick
decisions.
Biases
exist
in
online
social

interactions
as
they
do
in
real
life,
however,
there’s
much
more
scope
in
online
interactions
to

design
systems
to
overcome
many
of
these
biases.
Even
when
biases
remain,
there
is
certainly

more
scope
to
measure
them
and
design
ways
by
which
the
output
can
be
recalibrated
to
take

account
of
their
impact.
There
are
at
least
a
hundred
cognitive
biases
that
have
been

demonstrated
empirically
by
psychologists.
Some
of
these
are
particularly
relevant
in
considering

online
discussion
and
evaluation:

• The
backfire
effect:
Participants
can
often
strengthen
their
particular
beliefs
when
presented

with
evidence
that
does
not
support
their
view.

• The
bandwagon
effect:
Participants
are
more
likely
to
adopt
particular
behaviours
or

perceptions
simply
because
many
other
participants
do
(this
is
similar
to
social
herding).

15

• The
curse
of
knowledge:
Participants
who
have
more
knowledge
and
expertise
in
a
particular

subject
area
can
find
it
almost
impossible
to
think
about
the
subject
from
the
perspective
of

a
participant
who
is
completely
ignorant
about
the
subject
matter.

• The
overconfidence
effect:
Participants
are
likely
to
be
overconfident
in
their
own
solutions

and
answers
to
various
problems.

• Reactive
devaluation:
Participants’
evaluations
of
other’s
suggestions
can
be
negatively

influenced
by
the
simple
fact
that
the
suggestion
was
made
by
an
adversary
or
someone

with
an
opposing
perspective.

• In-‐group
bias:
Participants
tend
to
give
more
positive
evaluations
to
suggestions
from

members
of
their
own
group
–
even
when
the
groupings
are
arbitrary
or
irrelevant.

• Shared
information
bias:
Participants
often
spend
more
time
debating
issues
that
all

participants
are
familiar
with,
as
opposed
to
issues
that
only
a
minority
of
participants
are

aware
of.

Fragmentation
and
polarisation

When
debating
divisive
or
controversial
issues,
groups
can
often
split
into
one
or
more
opposing

groups.
As
the
discussion
progresses,
fewer
participants
remain
neutral
as
newcomers
are
more

likely
to
side
with
a
particular
group.
This
forces
the
opposing
groups
further
and
further
apart.

When
polarisation
occurs,
interactions
across
groups
can
become
even
more
disagreeable,

forcing
the
groups
further
apart.
As
a
result,
evaluations
made
after
a
group
discussion
can

sometimes
be
more
extreme
than
those
made
by
individual
participants
prior
to
any
interaction.

In
line
with
social
pressures,
participants
can
also
become
polarised
as
they
attempt
to
conform

to
the
norm
of
their
particular
group.

3.3
Administrative
Challenges

There
are
a
number
of
administrative
and
operational
hurdles
to
overcome
in
Collective

Intelligence
and
there
has
been
a
great
deal
of
research
to
understand
how
active
participation
in

online
communities
can
be
enhanced
and
maintained.
Overcoming
these
challenges
is
crucial

because
human
input
is
fundamental
to
Collective
Intelligence
–
without
enough
fuel
(and
the

right
sort
of
fuel),
the
engine
cannot
run
properly.
The
main
administrative
challenges
are

outlined
below:

Lack
of
participation

Different
participants
will
have
varying
levels
of
commitment,
expertise
and
availability.
Ideally,

the
value
of
individual
contributions
should
not
depend
on
either
the
frequency
of
contribution

by
an
individual,
or
the
time
that
the
person
can
spend
reading
and
contributing
to
the
exercise.

Lack
of
detail
in
participant
input

Depending
on
the
specifics
of
the
community
and
the
topic
being
discussed,
participants
can

often
provide
input
that
is
lacking
in
clarity
or
detail,
with
only
a
minority
making
carefully

considered
responses.
Of
course,
participants
need
to
be
able
to
answer
the
question
or
brief

they’re
given,
but
more
importantly,
they
need
to
be
bothered
to
craft
considered
input
and
to

read
and
respond
to
other
participants’
input.

16

Anonymity

A
common
problem
is
that
many
organisations
do
not
allow
anonymised
comments
in
tools
and

systems
from
which
Collective
Intelligence
can
be
elicited.
In
some
instances,
for
example

ideation
and
innovation,
being
named
may
enhance
the
quality
of
the
feedback,
as
the
social

recognition
obtained
from
coming
up
with
a
winning
idea
can
be
highly
motivating.
However,

when
discussing
many
organisational
issues,
what’s
mostly
required
is
honest
feedback,
which

may
be
critical
of
the
organisation
or
its
leadership.
Many
platforms
in
organisations
do
not
allow

for
anonymised
comments
and,
as
a
result,
this
means
that
employees
are
unlikely
to
be
candid

in
their
feedback.

Dealing
with
conflict
and
abuse

Without
a
doubt,
most
Collective
Intelligence
tools
will
produce
some
conflict.
Wherever
there
is

interaction,
there
can
be
conflict.
A
certain
amount
of
conflict
can
be
beneficial
if
it
helps
people

to
understand
a
more
diverse
range
of
perspectives.
Some
Collective
Intelligence
systems
are

designed
do
this
by
forcing
people
to
read
the
comments
of
people
who
are
different
to
them
in

some
way
and
using
visualisations
to
illustrate
the
diversity
of
opinion
that
is
largely
hidden
in

list-‐based
discussions.

Some
systems
can
also
weight
participants’
ratings
of
each
other’s
input
to
take
into
account

various
social,
attitudinal
or
demographic
biases.
However,
in
any
online
discussion
there
can

often
be
some
participants
who
do
not
share
the
general
view
of
the
wider
community,
or
even

worse,
individuals
who
have
set
out
to
deliberately
abuse
other
participants
or
disrupt
the

system.
These
people
can
often
dominate
discussions
as
the
attention
of
other
participants
is

directed
away
from
progressing
discussions
or
ideas
to
petty
arguments.

In
this
regard,
Godwin's
Law
is
an
Internet
saying
which
states
that
as
an
online
discussion
grows

longer,
the
probability
of
a
comparison
involving
Nazis
or
Hitler
approaches.
While
this
is
more

relevant
to
external
discussions
(clearly,
this
is
less
relevant
in
organisations
where
people
can

get
into
serious
trouble
for
or
being
offensive
on
social
media),
Godwin’s
Law
can
help
pinpoint

when
a
discussion
has
gone
on
for
too
long.
It’s
useful
to
know
that
the
rule
of
thumb
in
online

discussions
is
that
if
a
person
invokes
Godwin’s
Law
and
makes
the
comparison,
they

automatically
lose
the
argument.

Recruiting
participants

There
needs
to
be
careful
consideration
of
which
people
are
recruited
to
participate
in
Collective

Intelligence
activities
and
how
they
are
selected.
For
a
start,
any
collective
intelligence
tool
needs

to
have
a
minimum
number
of
active
users
to
have
meaningful
interactions
and
maintain

engagement.
Populations
can
be
as
small
as
20
to
30
people
and
potentially
as
big
as
everyone

who
is
able
to
read,
write
and
access
the
Internet.
Obtaining
good
participation
rates
is
important

because
having
a
rich
diversity
of
perspectives
is
an
important
pre-‐requisite
for
obtaining
the

best
Collective
Intelligence.
There
also
need
to
be
participants
who
are
progressive
thinkers
or

who
can
challenge
the
status-‐quo.
Participants
also
need
to
have
the
knowledge,
understanding

and
experience
to
contribute
to
the
exercise.
The
question
is
whether
meaningful
contributions

can
be
made
with
little
of
no
specific
subject
matter
knowledge,
or
whether
participants
require

specific
knowledge
and
experience.

17

Motivating
participants
and
incentivisation

After
recruiting
the
right
people,
participants
need
to
be
encouraged
to
share
their
input
and

interact
with
other
participants.
Participants
should
be
clear
what
it
is
they
are
a
part
of,
why

they
should
engage
and
what
the
value
is
to
both
themselves
and
the
organisation.
Intrinsic

motivators
such
as
social
recognition,
being
heard
(visibility
of
contributions),
loyalty,
having
a

common
purpose
or
being
interested
in
the
subject
matter,
can
interplay
with
more
extrinsic

motivators,
such
as
vouchers,
money
or
prizes.
Within
organisations,
acting
on
the
insights

obtained
from
Collective
Intelligence
exercises
is
typically
a
good
predictor
of
future
intentions
to

participate.

3.4
Managerial
and
Leadership
Challenges

Collective
Intelligence
in
organisations
poses
some
specific
challenges
for
management
and

leadership
because
tapping
into
Collective
Intelligence
means
taking
a
more
open
and

transparent
approach
–
and
this
requires
significant
cultural
change.
The
main
challenges
are

outline
below:

Loss
of
control

Taking
into
account
perspectives
and
suggestions
from
across
organisational
boundaries
and

hierarchies
may
leave
some
managers
feeling
vulnerable.
As
more
work
takes
on
elements
of

crowdsourcing
–
moving
tasks
from
one
to
many
–
this
could
be
construed
by
some
managers
as

a
sign
that
they
are
not
doing
their
jobs
properly.
There
is
also
a
risk
that
senior
figures
may

perceive
the
open
methods
used
to
gather
Collective
Intelligence
as
reducing
their
power
to

make
decisions.
A
common
theme
when
talking
about
leadership
in
the
digital
era
is
that
it

necessitates
a
different
way
of
doing
things.
However,
leaders
often
lack
skills
and
awareness

about
Collective
Intelligence
tools
that
stifles
their
view
of
tangible
organisational
benefits,

prevents
them
from
driving
change
and
makes
them
dismissive
and
distrusting.
In
terms
of

leadership,
this
means
finding
an
approach
that
fits
a
more
authentic
style
of
leadership
–

embracing
change,
being
open
to
experimentation,
demonstrating
transparency,
working

collaboratively
and
promoting
dialogue.

Overcoming
resistance

It
is
not
always
the
most
senior
leaders
in
a
company
who
are
most
resistant
to
a
more
authentic

way
of
operating.
Often
it
can
be
middle
managers
that
are
more
hesitant
about
embracing
an

open
approach.
This
is
perhaps
because
managers
at
this
level
are
most
fearful
about
any
critical

feedback
being
attributed
to
them.
In
addition,
it
is
frequently
middle
managers
that
own
the

work
streams
to
which
Collective
Intelligence
can
be
best
applied.

Ethical
treatment
of
crowd
workers

The
increasing
prominence
of
crowdsourcing
and
Collective
Intelligence
tools
that
facilitate
the

distribution
of
work,
such
as
parallel
task
processing,
have
given
rise
to
new
labour
market

segments.
The
global
labour
market
has
more
freelancers
than
ever
before
and
there
are
a

number
of
well-‐known
websites
that
offer
the
services
of
many
crowd
workers
to
perform
simple

tasks.
As
crowd
workers
of
the
this
type
increase
in
number,
organisational
leaders
must
consider

the
ethical
and
welfare
issues
involved
in
using
these
types
of
workers.

18

Intellectual
property
and
security
issues

Organisations
can
perceive
certain
risks
in
using
Collective
Intelligence
platforms
that
relate
to

intellectual
property,
liability
and
data
protection
issues
in
allowing
discussions
in
an
open
forum.

If
this
is
the
case,
agreement
should
be
sought
on
pre-‐established
terms
of
participation
and

ownership
of
intellectual
property.
Related
to
this,
there
may
also
be
concerns
about
the
open

communication
that
accompanies
Collective
Intelligence
approaches.
Inviting
large
numbers
of

people
to
participate
in
discussing
various
organisational
issues
means
that
more
people
will
be

aware
of
potential
problems
or
solutions.

Lack
of
action
in
response

An
element
with
which
organisations
struggle
the
most
in
asking
for
feedback
is
instigating
some

sort
of
change
in
line
with
the
insights
provided,
and
providing
feedback
to
participants
on
the

findings
and
actions
taken.
In
this
regard,
Collective
Intelligence
is
no
different.
Yes,
the
insights

may
be
richer,
more
valid
and
more
actionable,
and
participants
may
be
more
engaged
with
the

process,
but
unless
leaders
and
managers
act
on
the
insights,
the
Collective
Intelligence
obtained

is
largely
useless.

The
importance
of
leadership
support
in
seeking
solutions
through
open
dialogue
should
not
be

underestimated.
This
is
especially
true
where
organisations
need
to
be
seen
to
respond
and
act

on
feedback
that
is
provided
in
an
open
forum.
A
final
aspect
to
consider
regarding
actionability

is
that
collecting
employee
feedback
in
an
open,
transparent
and
collaborative
environment
has

enormous
potential
for
participants
to
actually
learn
and
share
information
during
data

collection.
This
can
increase
employee’s
readiness
for
change
in
that
they
are
more
aware
of
the

issues
at
hand
and
are
more
likely
to
feel
that
their
voices
have
been
heard.

Creating
specific
roles

Adopting
Collective
Intelligence
methods
can
also
require
the
creation
of
new
and
specific
roles,

such
as
senior
sponsors
to
lend
weight
to
new
ways
of
doing
things,
roles
to
administrate
and

moderate
Collective
Intelligence
activities
and
people
to
analyse
and
implement
the
output.

19

4.
Conclusion

There
are
many
ways
in
which
organisations
are
using
Collective
Intelligence.
From
ideation
to

augmenting
skills
and
predicting
the
future,
an
increasing
number
of
tools
are
allowing

organisations
to
more
easily
address
critical
business
issues.
Whatever
the
specifics
of
the

approach,
Collective
Intelligence
is
enabling
organisations
to
tap
into
the
shared
knowledge
and

expertise
of
their
workforce
(and
beyond)
to
address
their
most
important
business
challenges.

Collective
Intelligence
is
nothing
new.
Successful
organisations
have
been
encouraging

collaboration
and
interaction
since
the
first
modern
organisations
were
formalised.
However,
it
is

Social
Collective
Intelligence,
resulting
from
the
rapid
development
of
social
and
digital

technologies,
which
is
both
new
and
exciting
in
this
area.

Organisations
are
only
just
beginning
to

appreciate
the
potential
that
Collective
Intelligence
can
offer.
There
are
pockets
of
good
practice

appearing
as
organisations
start
to
implement
various
Collective
Intelligence
activities.
However,

these
early
examples
of
Social
Collective
Intelligence
are
likely
to
be
just
the
start
of
the
story.

As
the
area
develops
and
evolves,
those
who
are
developing
systems
and
tools
for
Collective

Intelligence
will
also
need
to
take
into
account
the
various
challenges
and
pitfalls
outlined
in
this

report.
Not
only
in
the
design
of
human-‐computer
interactions,
but
also
taking
into
account
the

psychology
of
social
and
cognitive
processes
that
occur
when
groups
of
people
interact
on
a
large

scale.
Moreover,
the
way
in
which
Collective
Intelligence
systems
are
implemented
and

embedded
within
organisations
needs
to
be
considered
along
side
the
administrative,
managerial

and
leadership
challenges
that
such
approaches
present.

One
issue
that
emerges
quite
strongly
in
this
report
is
that
not
all
Collective
Intelligence
is

created
equally
–
there
are
levels
of
Collective
Intelligence.
Simple
survey
methodologies
do

provide
Collective
Intelligence
of
a
sort,
but
what
they
lack
is
a
social
element
that
enables

participants
to
evaluate
each
other’s
responses.
This
is
important
because
when
this
type
of

aggregation
is
provided
it
is
possible
to
obtain
Social
Collective
Intelligence
and
that’s
quite
a

different
proposition
in
terms
of
the
richness
of
insight
produced.
If
a
group
is
interacting
in
a

system
that
has
been
carefully
designed
to
optimise
their
input
and
extract
insight
then
the
level

of
Collective
Intelligence
obtained
can
be
far
greater
than
any
survey.

Thinking
about
the
future
of
Collective
Intelligence,
it
is
clear
that
developments
in
human-‐
computer
interaction
are
likely
to
play
a
big
role
enhancing
human-‐human
and
human-‐machine

interaction.
However,
for
many
scientists
and
entrepreneurs,
the
ultimate
aim
is
to
develop

hybrid
systems
that
employ
both
human
intelligence
and
advanced
machine
intelligence.

The
progress
being
made
within
the
fields
of
artificial
intelligence
and
machine
learning
is
also

likely
to
be
at
the
forefront
of
Collective
Intelligence
as
it
evolves
in
the
coming
decades.

At
the

moment,
it
is
typically
the
humans
that
do
the
‘intelligent’
work
and
computers
that
do
the

administration.
However,
as
the
processing
and
storage
capabilities
of
computers
increases,
is

there
a
risk
that
human
input
and
creativity
will
be
increasingly
redundant?
Maybe
it
is
the

humans
who
will
be
the
machines’
administrators
in
the
future.

20

©Copyright
Silverman
Research
Ltd
2015

Silverman
Research

Arch
462,
Kingsland
Viaduct

83
Rivington
Street

London

EC2A
3AY

www.silvermanresearch.com

This
report
was
written
by
Michael
Silverman
and
Milly
Picton
at
Silverman
Research:
an
award-‐
winning
research
company
that
works
with
some
of
the
world’s
biggest
organisations
to
harness

the
collective
intelligence
of
employees,
customers
and
consumers.

For
more
information
please
contact:
info@silvermanresearch.com

Silverman Research: Collective Intelligence In Organisations Report

More Related Content

What's hot

Viewers also liked

Similar to Silverman Research: Collective Intelligence In Organisations Report

More from Silverman_Research

Recently uploaded

Silverman Research: Collective Intelligence In Organisations Report