Idea in a bottle – A new method for creativity in Open Innovation
Matthias R. Guertler, Christopher Muenzberg, Udo Lindemann
Institute of Product Development
Technische Universität München
Boltzmannstr. 15, 85748 Garching
guertler@pe.mw.tum.de
Abstract
This paper presents an approach to increase the creativity
of ideas/solutions in an idea contest. Analog to a letter
in a bottle tasks are distributed in a randomized way
to potential problem solvers. The idea contest is a
method from Open Innovation which opens a company’s
innovation process to its environment (e.g. customers,
suppliers). By using idea contests the creative
potential of a large crowd of people can be used for developing
innovative solutions for a specific task. Nevertheless,
based on experience from industry projects we
found that creativity often is limited. This paper presents
an approach for increasing the creative potential
of participants. The new integrated method combines
idea contest with lead user’s methods and aspects from
synectics and communication.
Introduction
Open Innovation integrates a company’s environment into
its innovation process, e.g. in terms of customers or suppliers,
and enables new innovations (Chesbrough et al. 2006).
A popular Open Innovation method is the, usually webbased,
idea contest which allows companies to publish a
specific issue/task to a large crowd of people. These develop
and post potential solutions for the issue. The idea behind
is using the diversity of the crowd to generate creative
and innovative solutions (Keinz et al. 2012). By giving
participants/users the possibility to review other posts they
can evaluate them as well as advance them. However, in
industry projects we found that submitted solutions often
are relatively homogeneous, of small number and of low
degree of creativity.
In order to improve participants’ creativity and the quality
of posts, we developed the approach “Idea in a Bottle”
based on the creativity method Synectics and Shannon’s
model of communication. The idea is to break up entrenched
processes within an idea contest where users/problem
solvers choose tasks to contribute. This is
done by allocating the four phases of synectics (see next
chapter) to different persons or groups and instrumentalize
the primarily negative “noise source” of Shannon’s model
in a positive manner. We propose, by randomly allocating
issues from idea-seekers to other users, their creativity is
stimulated. The confrontation with an unexpected, nonself-chosen
task helps overcoming our assumption that
users usually choose issues they are familiar with. To direct
the randomized process into efficient channels the
Pyramiding method from the lead user concept is utilized.
Thus, the first recipients of the issue do not solve it but act
as agents and forward it to users they consider to be suitable
and experienced on the specific field. These users submit
suggested solutions to the idea seeker who evaluates
the usefulness.
The proposed approach is applicable for issues/tasks of low
and medium complexity. This means the improvement or
new development of everyday products or the solution of
medium complex problems. All issues should be processable
without the need of highly specialized expertise or
know-how.
The paper starts with a rough overview of the state of the
art of Open Innovation, different user integration concepts,
synectics, and Shannon’s communication model. Based on
this we present our I aB approach. We close the paper with
a discussion about the planned evaluation of our approach
by integrating I aB into a web-based idea contest platform.
State of the art
This chapter shortly explains the underlying concepts of
the proposed approach “Idea in a Bottle”. The basic elements
are Open Innovation, synectics, Shannon’s communication
model, analysis-of-stimulating-word and pyramiding.
Open Innovation and Crowdsourcing
Open Innovation opens a company’s innovation process to
its environment (Chesbrough et al. 2006). The interaction
with the environment enables innovations inside and outside
the company. A concept focusing on the innovative
potential of a large group of people is Crowdsourcing
(Sloane 2011). The crowd can help elaborating and solving
Proceedings of the Fourth International Conference on Computational Creativity 2013 194
specific issues and tasks by using the diversity of persons
with their individual backgrounds, mindsets, abilities and
knowledge (Keinz et al. 2012). A popular Crowdsourcing
method is the usually web-based idea contest. Companies
or individuals can publish issues on a web-platform. Users
of the platform look at the issues and post ideas for solutions.
Other users review these posts, advance them or get
inspiration for new ideas. The goal is obtaining a large
number of advanced ideas.
Lead User
According to von Hippel et al. (2006) lead users are characterized
by (1) their capability for innovation as they are
ahead of the market, and (2) their motivation for contribution.
Several methods were developed to identify these
innovative users. One method based on the snowball effect
is the method Pyramiding. It is based on the assumption
that people who are interested in a topic know other people
who are more expert than themselves. Thus, Pyramiding
starts with an initial group of people who name other people
they consider to be more expert. These persons again
name persons considered to be more expert. After some
iterations potential lead users are gained (von Hippel et al.
2006).
Synectics
Synectic is a creativity technique based on brainstorming
and was developed by W.J.J. Gordon in 1960 (Daenzer
and Huber 2002). By postulating analogies from different
fields, e.g. literature, nature, or symbols, users of this
method are supported to find new solutions spaces for a
stated problem. Synectic is a group technique with a proposed
maximum of 10 participants who are instructed by a
skilled moderator (Daenzer and Huber 2002). Synectics is
structured into four phases which are passed through sequentially.
The four phases are:
1. In the Analysis phase the group exposes the problem
and states a problem definition. Also first solutions
will be gathered and documented. Finally the problem
should be restated.
2. The second phase, Incubation, is characterized by taking
one step back with the help of building analogies.
For example the group tries to build personal analogies
by thinking how the object of interest feels. The
outcomes of this phase are abstract solutions of the
problem.
3. In the third step the stated analogies get analyzed and it
is tried to transfer the solutions on the original problem.
This can also be done with the help of force fit,
i.e. oppressive reforming of the analogies. The results
of the Illumination are new solutions approaches.
4. In the Verification phase the proposed approaches are
used to elaborate solution concepts.
Presentation of Communication by Shannon
The communication process within an idea contest or synectics,
e.g. the problem description formulated by the ideaseeker
and interpreted by the problem solver, is one of the
success factors for developing appropriate solutions. In
1963 Shannon and Weaver proposed a schematic diagram
of a communication system (Shannon 1998). The proposed
diagram consists of five essentially parts. These are information
source, transmitter, channel, receiver, and the destination
depict in Figure 1.
Information
source Transmitter Receiver Destination
Noise source
Received
signal
Signal
Information Information
Figure 1: Schematic diagram of Communication by
Shannon and Weaver (Shannon 1998)
The information source produces messages or sequences of
messages which should be communicated. These messages
can be of various kinds, e.g. letters or functions (Shannon
1998). The operator produces a suitable signal for transportation.
The channel is the medium which transmits the signal
to the receiver. The receiver reconstructions the signal
and transports it to the destination, i.e. the person for
whom the message is intended.
An important factor in Shannon and Weaver’s diagram is
the noise source introduced in the channel. This source
leads to impacts on the communication. These impacts can
change the original message by new interpretations, extension,
reduction, or adaption (Lindemann 2009).
Analysis of stimulus words
This is a creativity method for developing new ideas by
confronting participants with words not related to the actual
topic. Participants analysis these words spontaneously
by relevant criteria and build links to the original topic
(Lindemann 2009).
A new method for creativity in idea contests:
Idea in a bottle (I²aB)
In order to increase the creativity and quality of ideas developed
during an idea contest, we suggest redesigning the
present communication process on an idea contest platform.
So far, in analogy to Figure 1, an idea-seeker describes
his issue (information source) by a problem description/task
(transmitter) and publishes it on the platform.
Here other users (receivers) can select this task, read it and
derive their understanding of the task (destination). The
following posting of solution ideas proceeds in an analogous
way.
Proceedings of the Fourth International Conference on Computational Creativity 2013 195
Our approach splits up the four steps of synectics and distributes
each to another group in order to increase efficiency
and creativity. The analysis (1) is performed by the
idea-seeker as “owner” of the problem. His analysis and
statement of the issue affect the entire following I aB process.
The incubation (2) is located by users of the platform
who read and interpret the problem statement. Based on
their understatement they link the issue to other users they
consider suitable for the issue. The illumination (3) is conducted
by the recommended users. They develop solution
ideas for the given task based on their own interpretation of
the issue and their personal background. The final verification
(4) of the created solution ideas is performed by the
idea-seeker himself again. Due to the incubation and illumination
stage are not executed by the idea-seeker but by
other users we term them “external”.
The I aB approach instrumentalizes the “noise source” in
terms of a randomized distribution of tasks to users. Instead
of selecting familiar issues users get new tasks. Receiving
unfamiliar topics shall support out-of-the-box
thinking by providing an external perspective on a topic.
To prevent demotivating users by receiving to many unfamiliar
topics the distribution and solving step are separated
by an intermediary Pyramiding step. The primary receivers
of a task forward it to other users they consider to contribute
a value gain to solving the problem. The process of the
idea-seeker putting his issue into the platform without
knowing who is receiving the issue is comparable to a letter
in a bottle thrown into the sea. Hence, the approach was
named “Idea in a Bottle” in analogy.
Figure 2 illustrates the concept of Idea in a Bottle (I aB). It
consists of four stages analog the synectics approach, as
mentioned previously:
In Stage 1 “analysis” idea-seekers phrase their problem/issue
in a written task statement. It can also be enhanced
by a picture or sketch. However, it is the intension
to gain a compact description of the issue which focuses on
relevant aspects. This increases the comprehensibility and
thereby the user’s motivation to deal with the issue. Thus,
the number of words will be limited to abstracts’ length
with ca. 250 words in the beginning. Adding characterizing
keywords supports the later forwarding process by the socalled
agents. All issues are stored on the web-based idea
contest platform.
In Stage 2 “external incubation” the Idea in a Bottle
(I aB) system distributes the issue in a randomized way to
three registered users on the platform. These users act as
agents: they examine and, due to its shortness, interpret the
issue. They are allowed to reply a potential solution idea.
However, primarily their function is forwarding the issue
to another user they consider able to contribute an add value
for solving the issue, e.g. due to their experience/behavior
in other idea contests on the platform. This
forwarding process is based on the pyramiding method of
the lead user concept. The optimal number of agents and
problem solvers needs to be evaluated in practical tests.
The randomized distribution and interpretation of the issue
by the agents equate the noise source of Shannon’s model.
Summarized, the randomization stimulates the creativity of
problem solvers in terms of analysis-of-stimulus-words
(Lindemann 2009). By receiving forwarded issues, we asd)

Verification
a)
Analysis
c)
External
illumination
Idea contest
agent problem
solver
idea
seeker
System
pyramiding
task
distribution
task
distribution
suggestion of solutions/ inspiration for idea seeker
act of attribution
randomized
allocation
problem,
task
task
distribution
b)
External
incubation
Figure 2: Model of Idea in a Bottle
Proceedings of the Fourth International Conference on Computational Creativity 2013 196
sume an increased motivation of problem solvers due to
the honor of being recommended by other users.
Stage 3 is called “external illumination” due to the interpretation
of the issue by other users. As described before,
the potential problem solvers receive a random issue with
the request for solving it. Since the problem solver does
not know the real problem, only the problem statement, he
builds new analogies of the given problem by interpreting
the issue. These new analogies combined with the randomized
distribution should lead to creative solutions which
were not considered by the idea-seeker. Similar to Stage 1
also the solution ideas can be consist of text, photos or
sketches. The size is limited, too. The problem solver is
considered to contribute with solution ideas. Otherwise it is
also possible to submit advices/hints which might indirectly
draw the idea-seekers attention toward alternative potential
sources and directions for a solution. Both the solution
ideas and the hints are submitted electronically via the I aB
system.
In Stage 4, “verification”, the idea-seeker receives potential
solution ideas and evaluates them regarding their applicability
to his problem. In comparison to “classical” idea
contest with a high effort in evaluating the gained ideas
(Kain et al. 2012), we assume the verification effort for
ideas created by I aB being lower since the solutions were
elaborated by qualified system user. In the case of no appropriate
idea the idea-seeker can submit his issue for a
second loop.
Conclusion and next steps
The presented approach supports increasing the creativity
and quality of solution ideas posted in an idea contest. This
is realized by a combination of crowdsourcing, synectics,
creativity techniques and pyramiding. Issues/tasks published
by idea-seekers cannot be chosen by other users as
in “classical” idea contests but are distributed in a randomized
way to users who forward it to potential problem solvers.
This randomized distribution combined with both the
interpretation by the agent and the potential problem solver
supports “out-of-the-box” ideas which might lead to innovative
solutions. At this, the confrontation with unfamiliar
topics acts as an analysis-by-stimulating-words and affects
the problem solver’s creativity. Additionally by being considered
as a kind of expert by other users the motivation
should tend to be high to contribute a solution.
I aB is enhancing, transferring and implementing classical
creativity methods for new media and distributed product
development activities. However, synectics was developed
in the 1960s and is a classical creativity method which can
be used in teams. We try to adapt this method for today’s
multi-media society.
To evaluate and proof these advantages we plan to implement
I aB in a web-based way. The basis will be an idea
contest platform at the institute which is being implemented
at the moment and is specifically designed for testing
new methods in the field of Open Innovation. This platform
allows Open Innovation contest with students as well
as industry as evaluation partners.
Here, we have the possibility to assess I aB in direct comparison
to a “classical” idea contest. At this, the user pool
of the platform can be used, as a sufficient community is
seen as crucial success factor.
Besides others, the following questions need to be addressed:
1. Does the satisfaction and motivation of problem
solvers increase?
2. Are differences regarding the number of replies to
an issue; the quality and usefulness of ideas; the
creativeness and the evaluation effort by the ideaseeker?
3. Is the choice of limitations of the issue description
useful?
4. Are there any specific patterns within the forwarding
process with frequently involved users?
Summarized, the expected key contributions of I aB are (1)
a higher creativity, (2) a higher motivation of problem
solvers and (3) a higher resulting quality of solution ideas.
<references_biblio/>
References
Chesbrough, H., Vanhaverbeke, W. and West, J. 2006.
Open Innovation: Researching a New Paradigm, New
York, Oxford University Press Inc.
Daenzer, W. F. and Huber, F. 2002. Systems Engineering:
Methodik und Praxis, Zürich, Verlag Industrielle
Organisation.
Kain, A., Kirschner, R. and Lindemann, U. 2012.
Utilization of Outside-In Innovation Input for Product
Development. International Design Conference DESIGN
2012. Dubrovnik, Croatia.
Keinz, P., Hienerth, C. and Lettl, C. 2012. Designing the
Organization for User Innovation.
Lindemann, U. 2009. Methodische Entwicklung
technischer Produkte: Methoden flexibel und
situationsgerecht anwenden, Berlin, Springer.
Shannon, C. E. 1998. Communication in the presence of
noise. Proceedings of the IEEE, 86, 447-457.
Sloane, P. 2011. A guide to open innovation and
crowdsourcing : expert tips and advice, London ;
Philadelphia, Kogan Page.
Von Hippel, E., Franke, N. and Prugl, R. Efficient
Identification of Leading-Edge Expertise: Screening vs.
Pyramiding. Technology Management for the Global
Future, 2006. PICMET 2006, 8-13 July 2006 2006. 884-
897.
Proceedings of the Fourth International Conference on Computational Creativity 2013 197