Content uploaded by Jan Konietzko
Author content
All content in this area was uploaded by Jan Konietzko on Jul 16, 2020
Content may be subject to copyright.
ACCEPTED VERSION
(Wordcount: 9065)
CIRCULAR BUSINESS MODEL EXPERIMENTATION: DEMYSTIFYING ASSUMPTIONS
Jan Konietzkoa, Brian Baldassarrea, Phil Browna, Nancy Bockenab, Erik Jan Hultinka
a Delft University of Technology, Landbergstraat 15, 2628 CE Delft, Netherlands; j.c.konietzko@tudelft.nl, B.R.Baldassarre@tudelft.nl,
P.D.Brown@tudelft.nl, N.M.P.Bocken@tudelft.nl, H.J.Hultink@tudelft.nl
b Internati onal Ins titute f or Industrial Environmental Economics (IIIEE),
Lund University, Tegnérsplatsen 4, Lund, Sweden; nancy.bocken@iiiee.lu.s e
Corresponding author: Jan Konietzko, j.c.konietzko@tudelft.nl, Tel.: +4915756013100
PLEASE CITE AS:
Konietzko, J.; Baldassarre, B.; Brown, P.; Bocken, N.; Hultink, E.J. Circular business model
experimentation: Demystifying assumptions. J. Clean. Prod. 2020, 122596.
ABSTRACT
Circular business model experiments may help firms transition towards a circular economy.
Little is known about how the participants of experimentation – entrepreneurs,
intrapreneurs, innovation managers – develop and test their assumptions during the
experimentation process to achieve more circular outcomes. Using a design-science
approach, we investigate this process and develop principles to improve it. This is done during
three workshops in different contexts: an innovation festival with 14 early-stage circular
startups, a workshop with a health technology incumbent, and a workshop with six growth-
oriented startups. We find that analyzing their available means – what they find important
and prefer to happen (part of their identity), what they know (their skills and knowledge), and
whom they know (their social network) – helps to understand how the participants develop
and test their assumptions. We show how the mindset and awareness of the participants
impact how much attention they pay to the circularity potential of their envisioned circular
business models. Based on these insights, we propose a set of principles to prepare the
innovation participants for experimentation, and to increase their ability to reflect on their
circularity assumptions. Future research is needed to further grow our understanding of the
types of principles that can guide meaningful experimentations towards a circular economy.
Keywords: Business model, Business model innovation, Circular economy, Lean Startup,
Effectuation, Experimentation, Sustainability
ACCEPTED VERSION
1. INTRODUCTION
Firms are in need of methods and approaches to innovate their business models towards a
circular economy (Blomsma and Brennan, 2017). In a circular economy, firms maximize the
value of the material resources and minimize the overall resource use, waste, pollution and
emissions that are associated with their business activities (Geissdoerfer et al., 2017).
Designing and conducting business model experiments – small-scale and cost-effective ways
to test the underlying theories and hypotheses about new business models – has become a
promising approach to innovate towards a circular economy (Antikainen et al., 2017a; Bocken
et al., 2019; Weissbrod and Bocken, 2017).
Most existing research on circular business model experimentation has used approaches that
operationalize the ‘The Lean Startup’ (Ries, 2011), a popular approach in entrepreneurship
practice (Antikainen et al., 2017a; Bocken et al., 2019; Bocken et al., 2017; Bocken et al., 2018;
Weissbrod and Bocken, 2017). This research has shown that experimentation can help speed
up action and decision-making towards sustainability in organizations. It has also revealed
that the decision-making process during experimentation may be more opportunistic and
messy than originally intended (Bocken et al., 2017). Participants often make intuitive
judgements and decisions (Foss et al., 2019), rather than rely on the decision criteria of the
experiment designs (Bocken et al., 2019). It also appears that the term experimentation may
lead participants to adopt a more ‘scientific’ language, but not necessarily a more rigorous
approach to innovation (Weissbrod and Bocken, 2017). In addition, collecting and analyzing
data during experimentation may result in unexpected events and surprises that require fast
changes of the experiment designs (Antikainen et al., 2017b). Some have suggested that
approaches like The Lean Startup fail to guide how the participants can develop and test their
hypotheses; that is, how they develop the underlying theory of value about their proposed
business models (Felin et al., 2019). Moreover, it appears that there is a gap between the
intended formality of experimentation approaches like The Lean Startup (Ries, 2011), and the
opportunistic and intuitive nature of how decisions are made during experimentation (Felin
et al., 2019; Foss et al., 2019; Sarasvathy, 2001).
The goal of this study is two-fold: first, we aim to better understand how the participants
develop and test their assumptions during circular business model experimentation; second,
we use this understanding to propose a set of principles that can help improve the process.
This is guided by two research questions: How do the participants develop and test their
assumptions during circular business model experimentation? How can a better
understanding of this help improve the process? Through a design-science approach for
entrepreneurship research (Romme and Reymen, 2018), we design and validate contexts and
principles for circular business model experimentation. This is done in the course of three
different workshops: a circular oriented innovation event with 14 novice student
entrepreneurs; an incumbent from the health technology sector and nine participants; and
six growth-oriented startups as part of a startup program, with twelve participants.
We find that analyzing their available means – what they find important and prefer to happen
(part of their identity), what they know (their skills and knowledge), and whom they know
(their social network) – helps to understand how the participants develop and test their
ACCEPTED VERSION
assumptions during experimentation. These available means (Sarasvathy, 2001) influence
what they focus on – whether they focus on, for example, the desirability of a value
proposition, or the contribution of an envisioned business model to a circular economy. Based
on these insights, we propose a set of principles to improve the process. This includes, for
instance, the importance of recognizing the available means of the participants, and to
prepare them if these means are not conducive to more circular outcomes. Future research
can use and further develop these principles to better understand how to experiment with
new business models towards a circular economy.
2. CONCEPTUAL BACKGROUND
In this section, we introduce the key concepts of this study: the business model, business
model experiments, and circular business model experiments. This leads us to identify the
research gap and the intended contribution.
2.1 BUSINESS MODEL
A business model helps to describe, investigate, and design how firms do business (Baden-
Fuller and Morgan, 2010; Magretta, 2002). It contains three essential elements: the value
proposition (what a firm offers and to whom), value creation and delivery (how it creates and
delivers the offering), and value capture (how it earns money and other forms of value with
it) (Bocken and Short, 2016; Richardson, 2008). From a design perspective, these three
elements can be desirable, feasible and viable (Brown, 2008; Calabretta et al., 2016).
Desirability is a property of the value proposition: how desirable a value proposition is to, for
example, intended users, customers or investors. Feasibility is a property of value creation
and delivery: how feasible it is to organize the needed activities and resources to create and
deliver the value proposition. Viability is a property of value capture: how the business model
can generate enough revenue to sustain the cost of creating and delivering the value
proposition (Figure 1) (Richardson 2008; Bocken and Short 2016; Calabretta et al. 2016). We
refer to the properties desirability, feasibility and viability because they are useful in the
context of experimentation, i.e. they can be tested. For example, you can test the desirability
of a value proposition, or the viability of a business model, to inform the right of course of
action during the design process (Simon, 1996).
ACCEPTED VERSION
Figure 1 – The business model (based on Richardson 2008; Bocken and Short 2016; Calabretta et al. 2016)
2.2. BUSINESS MODEL EXPERIMENTS
Business model experiments can be defined as small-scale and cost-effective ways to test the
underlying theories and hypotheses about the desirability, feasibility and viability of a new
business model (based on Calabretta et al., 2016; Camuffo et al., 2019; Osterwalder et al.,
2014; Ries, 2011). Most business model experiments with start-ups and established business
can be characterized as ‘quasi-experiments’ (Cook and Campbell, 1979), as they cannot be
easily controlled in a business environment (Bocken et al., 2018; Weissbrod & Bocken, 2017).
Experiments influence the experience and perception of entrepreneurs and organizations,
and help to form more accurate beliefs and expectations about the ‘right’ course of action
(Felin and Zenger, 2009). An experimental approach to business modelling makes it more
likely that entrepreneurs scrutinize the profitability of their ideas, that they pivot faster, and
that they increase their chances of high returns (Camuffo et al., 2019). Business model
experiments are important because from the outset, the probabilities of success are not
known (Knight, 1921), and the potential outcome is unclear (Kerr et al., 2014). These
conditions characterize business modelling as a highly uncertain process. Investors therefore
tend to value experimentation, because they enable them to fund startups and new business
models in stages. For each stage, experiments have to reveal new data that inform the quality
and likely profitability of the new business model. The benefit of experimentation in a
situation of high uncertainty is two-fold: one can assess projects without having to invest large
amounts of money upfront, and pursue projects without having to go for an all-or-nothing
bet (Kerr et al., 2014).
One of the most popular approaches for business model experimentation is The Lean Startup
(Blank, 2013; Felin et al., 2019; Osterwalder et al., 2014; Ries, 2011). This approach proposes
a formalized build-measure-learn cycle to conduct business model experiments: build a
‘minimum viable product’, measure how interested potential customers are in this product,
and use the results to learn whether an idea may work or not (Ries, 2011). This is often done
by using workshop material like ‘experiment cards’ that define the hypothesis, the test to
verify the hypothesis, the metric to measure success, and the decision criteria to further
pursue an idea (Osterwalder et al., 2014). Examples of such experiments include
Value
creation and
delivery
Value
capture
Value
proposition
Is it desirable? Is it feasible? Is it viable?
ACCEPTED VERSION
conversational interviews through a quasi-ethnographic approach with a potential partner,
or online A/B tests, where two landing pages with different value propositions are tested to
understand which element of the value proposition may gain more traction among potential
customers (Camuffo et al., 2019; Osterwalder et al., 2014).
A further important design approach to business model experimentation is effectuation
(Sarasvathy, 2008). Effectuation is a theory of entrepreneurship that explains how expert
entrepreneurs develop successful ventures. According to this theory, entrepreneurs start
with a given set of means (what they find important and prefer to happen, what they know,
and whom they know) to prototype new business models. These prototypes are shaped
through continuous negotiations to get the commitment and buy-in from external parties
(Sarasvathy, 2008). Effectuation poses that an expert entrepreneur follows four principles in
this process of new venture creation: 1) An entrepreneur only invests what she can afford to
lose. This principle reflects an iterative and step-by-step approach, which is similar to The
Lean Startup; 2) she seeks strategic alliances that provide commitment and buy-in for her
ideas. This stresses the importance of securing commitment and is also similar to the Lean
Startup approach, where direct payments or sign-ups are possible signs of commitment of an
experiment; 3) she captures value from unexpected situations. This principle emphasizes the
spontaneous and messy nature of the entrepreneurial process; 4) she controls an
unpredictable future by building a safe network of supporting stakeholders. This highlights
the need for a strong social network to sustain and grow the business (Sarasvathy, 2001).
Based on these principles, we pose that effectuation can be seen as an intuitive and less
formalized approach to experimentation (Bocken and Antikainen, 2019).
2.3. CIRCULAR BUSINESS MODEL EXPERIMENTS
Business model experiments have been increasingly conducted in the context of a circular
economy. Most of the existing research on circular business model experimentation has used
The Lean Startup as an underlying approach (see, for example, Antikainen et al., 2017b;
Bocken et al., 2018; Weissbrod and Bocken, 2017). A circular economy seeks to maximize the
value of products, components and material over time, and minimize the overall resource
use, associated emissions, waste and pollution (Geissdoerfer et al., 2017). Firms can
experiment with four inter-related circular strategies (Bocken and Antikainen, 2019): they can
narrow (use less material and energy during design, production, use and end-of-life), slow
(use products and components longer), close (use wasted products, components and
materials again) and regenerate (use non-toxic materials, renewable energy and manage
critical ecosystem services) the material and energy flows associated with their business
activities (Figure 2) (Konietzko et al., 2020a). Firms can use these strategies to develop new
circular business models, and then test how these business models can contribute to
circularity – in parallel to how desirable, feasible and viable they are. The goal is to develop
new business models that provide superior customer value, and that help to maximize the
value of products, components and materials over time, and to minimize the overall
associated resource use, waste, emissions and pollution (Bocken and Antikainen, 2019).
ACCEPTED VERSION
Figure 2 – A circular economy: narrow, slow, close and regenerate material and energ y flows (Konietzko et al., 2020a)
The existing research on this topic has shown that circular business model experimentation
can help stimulate innovation and action towards circularity in organizations. It has the
potential to promote an iterative ‘getting things done’ attitude among the participants
(Bocken et al., 2017). On a spectrum of what can be done to learn about new business models,
experiments are situated between fast learning (e.g., paper sketches, interviews) and slow
learning (e.g., business plans, pilots, market studies) (Bocken and Antikainen, 2019). The
success of circular business model experiments may depend on the following: a careful
selection of the participants (Bocken et al., 2017), internal buy-in from staff and top level
management, experimentation capabilities within the organization, as well as commitment
from relevant partners who can develop complementary products and services (Antikainen
et al., 2017b; Weissbrod and Bocken, 2017). It is also necessary to incorporate ‘circularity
checks’, to make sure that experimentation is geared towards higher circularity (Bocken et
al., 2018).
These ‘circularity checks’ are especially important. This is because circularity – a situation in
which the value of products, components and materials is maximized, and in which the overall
resource use, waste, emissions and pollution are minimized – is a property of a higher-order
system, rather than a property of an individual product or business model (Konietzko et al.,
2020a). For example, a car may be made lighter and more durable. But if the overall number
ACCEPTED VERSION
of cars on the road increases and the cars stand idle 95 % of the time, then the overall
resource use, waste and emissions are not minimized. Providing a car sharing service, that is,
changing the business model, may decrease the overall number of cars on the road. But if the
cars are powered with fossil fuels and still have an idle time of 60 %, then the overall resource
use, emissions and waste are not minimized. Instead of focusing on products and business
models only, circularity thus needs to be approached from an ecosystem perspective.
From a circular ecosystem perspective, a firm can experiment with a set of complementary
products, services and business models (Konietzko et al., 2020b). For instance, to maximize
the capacity use of cars, a car sharing provider may try and connect business-to-business fleet
operators that have previously had their own fleets. The same cars can also be made
accessible for end users through a joint car sharing platform, as well as for the staff of the
involved companies through a corporate car sharing program. The car sharing provider can
then work together with a local energy provider and make sure the cars are fueled with
renewable energy. The batteries in these cars, once they are below a certain quality
threshold, can then be installed in office spaces to provide heating and thereby prolong their
useful lives. As this example illustrates, several different actors need to be activated and
aligned to jointly contribute to circularity as a collective outcome.
Due to the complexity of this collaborative and uncertain process, understanding how
circularity can be achieved is a major challenge (Brown et al., 2019). It is therefore important
that the innovation participants develop accurate assumptions about the circularity potential
of their envisioned business models. In other words, they need to develop a critical and
reflective mindset, not only with regards to how desirable something is for the user, but also
with regards to the circularity of their proposed circular business models. To develop such a
mindset, it is first necessary to understand how the participants develop and test their
assumptions during experimentation – to then see how this process can be organized to
achieve more circular outcomes.
2.4. RESEARCH GAP AND CONTRIBUTION
Previous research on circular business model experimentation has found that structured
experimentation may often be more messy and opportunistic than originally intended
(Bocken et al., 2017). There seems to be a gap between the intended formality of quasi-
experimental approaches like The Lean Startup and the intuitive and opportunistic nature of
judgements during experimentation (Felin et al., 2019; Foss et al., 2019). In particular, it is not
clear how the participants build their hypotheses and underlying theories of value about the
possible desirability, viability, feasibility of their envisioned circular business models, as well
as their contribution to circularity (Felin et al., 2019). Our study addresses this gap about the
process of developing and testing assumptions during circular business model
experimentation. It is important to better understand this, because it influences the
circularity outcomes of the envisioned business models. In this study, we therefore want to
better understand how the participants develop and test their assumptions during circular
business model experimentation; and to use this understanding to develop principles that can
help improve it.
ACCEPTED VERSION
3. METHOD
This study applies a design science framework for entrepreneurship research (Romme and
Reymen, 2018) to research how the participants develop and test their assumptions during
circular business model experimentation (Figure 3). The purpose of the framework is to
develop knowledge that is both theoretically sound and practically useful (Denyer et al., 2008;
Van de Ven, 2007). The research output from this study is a better understanding of how the
participants develop and test their assumptions during this process, and a set of principles to
improve it. The framework serves to specify how to design and validate this research output
within a continuous research cycle: how to create and evaluate (together: design), and how
to generalize and justify it (together: validation). It is important to note that these four steps
are complementary and researchers may jump from one step to another.
Figure 3 – A framework to design and research workshop forma ts for circular business model experiments (Romme and
Reymen, 2018)
3.1. DESIGN AND VALIDATE THE CONTEXTS OF EXPERIMENTATION
The first step of the study is to design and validate the contexts of circular business model
experimentation. This is done through three workshops. Each workshop represents a
different context of experimentation: one with sustainability-minded novice entrepreneurs,
one with an incumbent firm that communicates ambitions to innovate towards a circular
economy, and one with more experienced and sustainability-minded entrepreneurs. Each
one in turn:
Design
Validation
How participants develop
and test their
assumptions during
circular business model
experimentation
ACCEPTED VERSION
1) The first workshop was created for a ten-day innovation event for the circular
economy in the North of The Netherlands (event name: DORP). The event hosted 14
early-stage start-up ideas for a circular economy, posed by novice entrepreneurs, and
around 70 participants (most of them master students with design or engineering
backgrounds) who formed groups around the 14 ideas. Examples of the startups
include two architects who developed a modular furniture set that can be playfully
turned into twelve different furniture types (e.g., armchair, coffee table, bench, office
table or display); a start-up that has developed packaging material based on wood
from certified forests in Sweden; a startup that develops a service model to replace
disposable plates and cutlery with reusable ones; a firm that turns old, otherwise
wasted bread into beverages.
2) The second workshop was created for nine participants from a Dutch incumbent in
the health technology sector. The goal of the company is to become a circular
economy pioneer and it has a defined circular economy strategy that needs to be
implemented by the different sections of its business portfolio. The participants of the
workshop focused on a business section that sought to turn a consumer product from
a sales into a product-as-a-service business model.
3) The third workshop was created for 12 participants from six circular oriented startups
during an accelerator program of the Impact Hub in Zurich, Switzerland. Examples of
these startups include a firm that rescues left-over yields from farm lands and turns
them into a vegetable box subscription, a firm that provides baby clothing as a service,
and an online platform where users can share everyday goods.
3.2. DESIGN AND VALIDATE PRINCIPLES FOR EXPERIMENTATION
The first set of principles, applied within a workshop format, was designed for and validated
during a ten-day innovation event for the circular economy in the North of The Netherlands.
The initial set of principles was derived from the business literature and based on what has
been used in earlier research on circular business model experimentation. The principles
included: 1) formulate the assumptions you have about how and why an envisioned business
model may work in reality (Ries, 2011), 2) test your assumptions early outside of your
organization’s boundaries, rather than plan thoroughly ‘at the desk’ (Blank, 2013), 3) iterate
fast and several times through the build-measure-learn cycle (Ries, 2011). These principles
were instantiated in the form of a list of possible test methods and instructions (Table 1)
(retrieved from Schuit et al. 2017; Bocken et al. 2018; Ries 2011; Osterwalder et al. 2014), as
well as test cards to formulate assumptions and a validation graph to prioritize the tests
(Figure 4) (based on Osterwalder et al., 2014).
Method
Instruction
Brainstorming
Get a multi-disciplinary team and perspectives from outsid e the company a nd sit together to
brainstorm about the assumption
Conversational interview
Interview the person of i nterest to learn from them
Online A/B test; split-test
experim ents
Get budget for ad-campaign and a content-writer for ads, write ads and launch them on e.g.
Facebook, Google, etc. Mak e different versions to test different assumptions
Booklet interview
Make a product/service booklet and hand it to a potential c ustomer to get feedback
Ethnographic observation
Get into the field where your customer/user/partner is and observe what they do and how they
do things
ACCEPTED VERSION
Creative session with users
Invite users/cust omers/partners who ar e able and willing to discuss openly to have a creative
session about the problem/potential solution
Moderated online discussion with
community members
Find an online forum about your probl em and learn from posts, star t a discussion about the
learning you are trying to gain
Co-create session with
stakeholders
Find a location and schedule a meet-up with relevant stakeholders to co-create a solution
Rapid service
prototyping/mini mum viable
product
Make a first physical and/or digital prototype (e.g. paper mock-up, web landing page, cardb oard
mock-up), get in front of customers and learn from their reactions
Landing page with Video + option
to sign up
Make a short video wher e you pitc h your idea and create a landing page with a call to action
(e.g. sign up for the newsletter, early ordering option for product, etc.)
Concierge MVP: "fake it until you
make it"
Try to fake the product/service through human actions, help the customer out right away
without having any product, improvise
Field experiment
Find a test ground (e.g. a festival), user group, and create an experiment set-up
Wizard of Oz testing
Take humans who can provide the service that you want to provide instead of machines to gain
learning
Table 1 - List of possible tests that was available for the first workshop (retrieved from Schuit et al. 2017; Bocken et al.
2018; Ries 2011; Osterwalder et al. 2014)
Figure 4. Initial test cards and the validation graph. Based on Osterwalder et al. (2014)
During the event, the principles were presented in 30 minutes to the participants. The
presentation triggered the group to discuss and reflect on their envisioned business models
in terms of assumptions. Questions we discussed included: “what would need to be true for
your ideas to work in reality?”; “What are your assumptions?” We then went through the
test cards and explained how the participants could use them to develop and test their
assumptions and define tests, metrics and decision criteria. We also introduced the list of
tests they could do and discussed some examples. The validation graph was presented as a
way to plot and prioritize the test cards according to what they would perceive as easiest and
most critical to test. After the presentation, they spread out into groups to use the provided
material in a two-hour workshop session.
3.2.1. Data collection
In the course of the three workshops, we collected different types of data. During the first
workshop, we conducted semi-structured interviews (see Appendix A for an overview of the
Most critical
to test
Most difficult
to test
Easiest
to test
ACCEPTED VERSION
interview questions we asked) (Patton, 2002), made notes to capture observations about the
use of the workshop material, made photos of the filled material and followed up with some
of the participants later on to see what experiments the participants eventually ran, and what
they learned from them. In the second and third workshops, we made notes during the
workshops, took photos of the filled-in material and content from post-its, documented
discussions among the researchers about the workshop afterwards, and for the third
workshop collected a filled-in survey half a year later about the progress and activities since
the workshop. Furthermore, each workshop was evaluated by collecting and analyzing data
on its user acceptance. This was measured in terms of its ease-of-use and perceived
usefulness (Davis, 1989). The participants filled in feedback forms after each session
(Appendix B). The form stated the intended purpose of the workshop (first version:
“understand the assumptions underlying a business idea, and to decide how to test them, and
what to test first”) and then posed two statements: “The material is useful to address the
stated purpose above.” and “The material is easy to use.” Each statement could be rated with
a Likert scale from 1 to 7 (1= fully disagree, 7=fully agree, after the first round we adapted
this to 1-5). We also encouraged the participants to explain their rating through written
feedback. The results were used to validate the ease-of-use and usefulness of the principles
that we proposed for the workshops. Table 2 provides an overview of the collected data
during each of the three workshops.
Collected data
Total length/amount of data
First workshop
Feedback forms
35 filled in forms
Audio/video recorded session
115 minutes
Observations from researchers
145 minutes/4 pages
One interview after session about how easy to use the
session materials were (10 minutes each)
60 minutes
Discussio ns among researchers about the session
60 minutes/two pages
Filled-in test cards
8 test cards
One interview during the testing
55 minutes
Observations from researcher during testing
120 minutes/4 pages
One interview after the testing per group
40 minutes
Second workshop
Feedback forms
9 filled in forms
Photos from post-its and generated ideas and
strategies
22 photos
Observations from researchers in the form of notes
180 minutes, one page summary
Discussio ns among researchers about the session
30 minutes, one page summary
Filled-in workshop material
9 filled in templates
Third workshop
Feedback forms
6 filled in forms
Observations from researcher in the form of notes
180 minutes, one page summary
Filled-in workshop material
6 filled in templates
Company survey after half a year
6 filled in surveys
Table 2 – List of collected data from the three workshops
3.2.2. Data analysis
The data was coded using a mix of descriptive (describe what is being said), In Vivo (which
uses the actual language used by participants and reflects the emotionality of the situation)
and process coding (observing actions performed by the participants) (Saldaña, 2013). We
coded the data according to the three available means of an effectual decision-making logic
(Sarasvathy, 2001): what they find important (part of their identity), what they know (their
ACCEPTED VERSION
skills and knowledge), and whom they know (their social network). The codes were developed
through an iterative coding process that revealed how these available means influenced how
the participants developed and tested their assumptions. For example, one important code
for the category ‘what they find important’ is ‘the business model property’, sub-divided into
the codes desirability, viability and feasibility. This coding enabled us to analyze what business
model property the participants found important to investigate. Figure 5 shows the coding
structure that resulted from the data analysis. The identified codes within the three
categories are not meant to be exhaustive. Rather, they show important elements that had
an influence on how the participants developed and tested their assumptions throughout the
three workshops. The resulting coding structure informs the theoretical research output of
this study, which is detailed in the results section 4.1.
Figure 5 – The resulting coding structure from the data analysis
In addition to the coding structure, each workshop was evaluated through the feedback
forms. We did this to ensure the practical relevance of the principles that we applied
throughout this research. The feedback form gave us insights on the usefulness and ease of
use of the proposed principles, and served to develop an evaluated set of principles as a
practical research output of this study. This output is detailed in the results sections 4.2.
4. RESULTS
We present the results in terms of theoretical and practical relevance. The first section (4.1.)
presents the theoretical results that address the first research question: how the participants
develop and test their assumptions during circular business model experimentation. The
second section (4.2.) presents the practical outcomes, in the form of principles, that address
the second question: how a better understanding of this can help improve the process.
What they find important
What they know
Whom they know
Business model properties
Relevant background knowl edge
Information provided during the
workshops
Typ es of learning
Existing networks
Desirability
Feasibility
Viability
Fast learning
Slow learning
Impact assessment methods
Scientific method
CATEGORY CODES SUB-CODES
List of possible tests
List of metrics
Circularity
ACCEPTED VERSION
4.1. HOW THE PARTICIPANTS DEVELOP AND TEST THEIR ASSUMPTIONS
We find that the participants develop and test their assumptions in terms of what they need
to find out about their envisioned business models, and how they can find it out. The decision-
making logic that underlies this process is influenced by their available means: what they find
important (part of their identity), what they know (their skills and knowledge), and whom
they know (their social network).
4.1.1. What they find important
The participants decided what they needed to find out and how they could find it out based
on what they found important. This related to, for example, if they found circularity important
to investigate, the business model property (desirability, feasibility, viability), or if they prefer
fast and/or slow learning.
Circularity: Across all three workshops, experimenting with and investigating circularity was
not considered most important. This was in spite of the fact that the workshops were about
developing business models for a circular economy. In the first workshop, when prompted,
the participants found it difficult to pinpoint which sustainability problem they were trying to
address. They stated: "We are assuming it is more sustainable than current offers"; “it's not
crucial right now”. In the second workshop, circularity was defined by the whole organization
under the term ‘circular revenues’. A product-as-a-service model, for example, would count
as ‘circular revenues’. When asked to reflect on why product-as-a-service models were
circular, the participants noted down: “subscription enables refurbishment, personalized
offering (buy only what you need), access over ownership”; “obvious”; “first step to service
concept, investigate need for update and refurbishment”; “we remain the owner, closed loop
logistics, reusing basic materials, owning materials”; “service model, we own the product”.
Only one participant, who had previously worked with environmental life cycle assessments,
questioned: “Is refurbishing more circular and better for the environment?” In the third
workshop, one participant investigated the circularity potential of their idea by asking how
many use cycles they could achieve with their baby clothing-as-a-service model, compared to
the current average number of cycles. Apart from that, most of the participants assumed that
their solutions are ‘better’ for the environment compared to existing offerings, and did not
find it important to better investigate this assumption. They only seriously investigated and
documented their assumed circularity or environmental improvements when they had to fill
in a dossier for a startup award. These findings show that ‘circularity checks’ are influenced
by whether participants find circularity relevant and important to their process.
The business model properties: The most important business model properties are
desirability, feasibility and viability. Most participants in the first workshop focused on the
desirability of their envisioned business models. They paid a lot of attention on how they
could sell their products and services, and what value they would provide for their customers.
The other two workshops were more mixed, with attention on several business model
properties and no clear preference for either of them. For example, in the second workshop,
one participant with a user-centered approach was interested in the desirability of a
refurbished product. Another participant wanted to investigate how feasible their idea was
ACCEPTED VERSION
in terms of the hygiene of returned products. Yet another was curious to investigate if the
lower price point would get a customer to buy a refurbished product. In the third context,
most of the participants considered it important to focus their experiments on the desirability
and viability. Table 3 contains a selection of quotes that show which business model
properties the participants found important to investigate. They illustrate which business
model property the participants found important.
Workshop 1
Workshop 2
Workshop 3
Desirability
“Are they potentially
interested?”
“Will they understand our
story?”
“Will they like our product?”
“How to make the product
more appealing for potential
customers?”
“How can we turn our service
into an experience for the
customer?”
“What is a good name for
this product?”
“What drives the consumer?
What do they want in
refurbished products?”
“How many of these products
does an average customer
buy in a lifetime?”
“What is the customer
perception of refurbished
products?”
“What is our target group?”
“How many of our customers
are willing to pay for this
offering?”
“How many will sign up if we
advertise this service?”
Viability
“What are people willing to
pay?”
“Does the price drive the
decision to buy a refurbished
product?”
“What are our costs?”
“How can we price our
service?”
“Is this financially viable?”
Feasibility
“Does the service model
work?”
“What are the challenges of
delivering this service?”
“Will reused products be bio-
contaminated?”
“Can the product be fully
modular?”
“Does refurbishment affect
product safety?”
“Will the customers clean the
product themselves?”
“How can we get our users to
act autonomously?”
Table 3 – Questions that the participants in the three workshops found important to investigate
Fast and/or slow learning: Fast learning during business model innovation can be gained, for
example, via paper sketches, quick interviews or try-outs. Slow learning happens through, for
instance, business plans, market studies or pilots. In the first context, most participants found
fast learning more important than slow learning. This is likely related to the context of the
workshop: an innovation festival in the summer with prototyping facilities and a near-by
music festival to test the prototypes. The founder of a service model for reusable plates, for
example, noted that they were “just trying stuff quickly” to see what worked and what did
not. Another participant noted during the testing: “you just change things quickly and see
what happens”. In the second and third workshops, participants had mixed preferences for
both fast and slow learning. In the second one, some participants were eager to act and
ACCEPTED VERSION
organize a fast experiment and interviews with some of their employees in their office
building. Another participant preferred to conduct a life cycle assessment on the possible
environmental impacts of selling refurbished products. In the third workshop, some found it
important to make an elaborate cost calculation to design and plan an experiment. Another
participant decided to focus on quick changes to the website design, search engine
optimization and customer journey optimization. These examples illustrate that whether the
participants prefer fast and/or slow learning influences what they want to test and how.
4.1.2. What they know
The participants determined what they had to find out and how based on what they knew.
This related to, for example, their relevant background knowledge or the provided
information during the workshops.
Relevant background knowledge: Relevant background knowledge refers to the skills and
knowledge that the participants bring into the experimentation process. In the first workshop,
most of the participants did not follow the suggested quasi-experiment approach and
rigorously collected data, but instead wanted to learn by doing. For example, the leader of a
startup that offered multifunctional furniture noted that there is “no need to be too rigid
about things”. The team was simply looking to get customers to sign up. This can be partly
explained by a lack of background knowledge of and experience in experiment design. A team
member from the service model for reusable plates concluded from the testing that the
service model did not really work, because the plates did not meet the aesthetic requirements
of their client. Again, this was not based on a carefully designed experiment, but came from
the direct, intuitive experience. In the second workshop, the participant who had previously
worked with Life Cycle Assessments suggested to conduct such an assessment for the
envisioned business model around refurbished products. Another participant with a design
background focused on the user-centered methods for value proposition design. In the third
workshop, some participants with a mechanics background focused on the feasibility of
repairing a certain number of products as part of their envisioned business model. Others
with a marketing background focused on how they could optimize their online channels to
attract more customers. As these examples show, the background knowledge has an
influence on what the participants want to test and how they want to test it.
Provided information during the workshops: This refers to the information that the
participants receive during the experimentation, for example in the form of concepts and
methods that they can use. In the first workshop, the list with available testing methods
contributed to the participant knowledge about how to test their assumptions. During the
testing, they used methods such as conversational interviews to understand how much
potential customers were willing to pay, how well they understood the story, ethnographic
observations to see how users interacted with their prototypes, A/B tests to understand
preferences, competitor comparisons, and ‘Wizard-of-Oz’ testing (“fake it until you make it”).
In the second workshop, the participants used the provided information to formulate why
customers would be interested in their proposed solutions, or what they could do to test their
assumptions. In the third workshop, participants were triggered to select a concrete metric
that they wanted to improve through their experiments. They used this information to
concretize their ideas for experiments. For example, one startup that wanted to monetize
ACCEPTED VERSION
left-over yields from farm lands decided to measure buy-in from a potential retail partner
through an experiment to launch a weekly veggie box subscription (“Will they accept the price
offering?”). Another startup that provided baby clothing as a service defined the circularity
metric ‘number of use cycles’ to measure its comparative impact in the baby clothing market
(where there is generally one use cycle). This shows that the available information during
experimentation influences what the participants want to test and how.
4.1.3. Whom they know
The participants decided what they had to find out and how based on whom they knew. This
related to, for example, their existing network.
Existing network: This refers to how the social network of the participants can help support
the experimentation process. In the first workshop, the existing network had an influence on
how the participants prioritized what assumptions to test. For example, the founder of one
startup noted that she could “easily take this one to our partner and discuss". Towards the
end of the workshop session, another participant noted that “it is interesting that a lot if this
really boils down to the network”. Whom they knew had an influence on how they prioritized
what assumptions to test first. One participant noted that "there is actually someone here we
can ask about this". The founder of the startup that offered multifunctional furniture noted
that it was easy to find out how their furniture adds value to the brand experience of their
potential clients: she already had a client who used their furniture for this purpose, and could
go and ask them for more details about how the furniture added value. The existing network
also helped get further contacts and buy-in from external parties. For example, the startup
with the service model for reusable plates got buy-in to conduct a full experiment at the
festival from the event organizers, because they believed in the idea. They also helped to
connect the startup to the food providers on the festival to co-organize the experiment. In
the second workshop, existing retail partners were mentioned as potential places to conduct
an experiment to try and offer a product-as-a-service model. Also internal staff was
mentioned as a potential test group to conduct some early experiments around user
acceptance for a refurbished product. Similarly, in the third workshop, participants designed
experiments together with existing retail, distribution or promotion partners. It appears that
the network determines which assumptions the participants prioritize, because tapping into
the existing network is immediately actionable. It requires comparatively low efforts to set
up experiments and to get the needed information. This shows that the existing network can
influence how participants want to test their assumptions during circular business model
experimentation.
4.2. PRINCIPLES TO HELP IMPROVE CIRCULAR BUSINESS MODEL EXPERIMENTATION
We have shown how analyzing their available means – what they find important, what they
know and whom they know – can help to better understand how the participants develop
and test their assumptions during circular business model experimentation. Based on this
better understanding, we propose a set of principles for before experimentation, and a set of
principles for during experimentation.
ACCEPTED VERSION
4.2.1. Before experimentation
Recognize what the participants find important: to ensure that circular business model
experiments aim at higher circularity (or lower environmental impact), it is important to
involve participants who care about circularity and the minimizing of environmental impact.
The more the participants think it is important to ensure that their envisioned business
models reduce environmental impact and resource use, the more likely they are to be critical
and scrutinize their assumptions about the circularity of the proposed ideas. If some of the
participants do not think that circularity is relevant and important, then they need to be
supported in developing a stronger awareness about it.
Recognize what the participants know: to ensure that circular business model experiments
aim at higher circularity, it is important to involve participants who know about the
environmental impacts of their business activities, and how this impact – and the potential
impact of the proposed business model changes – can be measured using concrete metrics.
In addition, the more the participants know how to apply the principles of the experimental
method (how to formulate a hypothesis or theory, and how to test it rigorously), the more
likely they are to avoid false negatives: where they disconfirm the potential of an opportunity
where there is one; and false positives: where they confirm an opportunity where there is
none.
Recognize who the participants know: to ensure that circular business model experiments aim
at higher circularity (or lower environmental impact), it is important that the participants
explore and develop a supportive network that can help inform and conduct the experiments.
A supportive network can, for example, make the experiments more actionable (partners can
provide space to experiment), more collaborative (partners can co-develop complementary
products and services), more cost-effective to organize (known partners mean lower
transaction cost because of existing ties), and more meaningful (knowledge partners can, for
example, help assess the circularity of the experiments).
4.2.2. During experimentation
Formulate assumptions in terms of what you need to find out: in the first workshop, we
proposed test cards and a validation graph to the participants as a way to develop and test
their assumptions. The average rating of perceived usefulness was 4.8 (out of 7), and of ease-
of-use 5.1 (out of 7). Many who provided a rating indicated that they did not use the methods
(25%). The test card’s ability to stimulate immediate action was limited. Thinking in terms of
assumptions was often not perceived as helpful. As one participant pointed out: "I feel like
we don't end up anywhere if we point out all these assumptions”. Instead, the participants
developed an intuitive alternative to the test cards to formulate their assumptions. They
simply asked: “what do we need to find out to see if this can work?” In the second workshop,
the participants used this technique to post their assumptions on a wall. This was perceived
as a useful way to document the things they did not know and that they wanted to find out.
Prioritize assumptions in terms of what you can do right now, with what is available: the
participants in the first workshop tried to answer their questions by looking at currently
available means. One noted: "the question is really what we can test here and now". Another
ACCEPTED VERSION
participant commented that “it is true that there is a lot that you can do, but it is also about
what is it that you can do right now”. In the second workshop, the prompts to document
possible immediate actions (“what can we do right now to find out?”) were captured on post-
its and collected on a wall. They provided an intuitive and easy-to-use way to generate a
concrete action plan for the next experiment.
Define key metrics: in the second workshop, the metric of ‘circular revenues’ (e.g., revenue
from a product-as-a-service model) was defined as a key metric to guide experimentation. In
the third workshop, we asked for feedback on the usefulness and ease-of-use of using
concrete metrics to guide circular business model experiments. These were perceived as
useful (average rating of 4.25 out of 5) and moderately easy to use (3.5 of 5). The moderate
rating on ease of use was because one team needed more time to define meaningful metrics,
and another participant who had to leave earlier and could therefore not use the workshop
material as intended. The use of circularity metrics prompted the participants to focus on one
key metric that can help them specify how each action further grows the business and
increases circularity. For example, one startup that developed a baby-clothing-as-a-service
model focused on ‘number of use cycles’ as a circularity metric. They found that the
subscription model may lead to six use cycles, compared to one cycle in the sales model.
Another startup that developed a sharing platform for everyday goods measured the number
of items on its platform and the number of times they have been rented out to make
inferences about avoided sales of these items. The participants noted that defining metrics
to guide their experiments helped to “decide what to focus on” and that “it was very helpful
to decide on goals for the coming time”.
5. DISCUSSION
This study makes two contributions to the existing research and practice of circular business
model experimentation. First, to research, it adds an improved understanding of how the
innovation participants – entrepreneurs, innovation managers, business managers, designers
– develop and test their assumptions during the experimentation process. Second, for
practice, it adds a set of principles – based on this improved understanding and the workshop
evaluations – that can help to improve the experimentations. We discuss both contributions
and the limitations of this study in the following sections.
5.1. CONTRIBUTION TO CIRCULAR BUSINESS MODEL EXPERIMENTATION RESEARCH
The findings from circular business model experimentation research show that the
experimentation reality is less formal than what may be desirable according to The Lean
Startup (Ries, 2011), confirming earlier findings on the application of Lean startup in the
circular economy context (Bocken et al., 2017). In general, approaches like The Lean Startup
lack an understanding of, and guidance on how the participants – entrepreneurs, innovation
managers, business managers, designers – develop an underlying theory of value about their
envisioned business models (Felin et al., 2019). In this study, we seek to contribute to a better
understanding of this process. In particular, we show that their available means influence how
the participants move through the experimentation process. Decisions on what to test, how
to test it, and what to conclude from the tests are influenced by an effectual logic and
behavior: what they find important (part of their identity), what they know (their skills and
ACCEPTED VERSION
knowledge) and whom they know (their social network) (Sarasvathy, 2008). This supports the
findings from the circular business model experimentation literature (Bocken et al., 2017). It
also fits with the understanding that the innovation process is often driven by subjective and
intuitive judgements (Foss et al., 2019). It is therefore important to recognize this underlying
process of developing and testing assumptions, to make the participants aware of it, and in
turn to develop a more reflective and rigorous process.
It is important to highlight that these findings do not intend to discredit the merits of a more
formalized approach to entrepreneurship. We are aware of earlier research that has
demonstrated the potential positive influence of a more formal approach to business
venturing (Camuffo et al., 2019). Rather, we argue that a better understanding of the
subjective nature of decision-making during experimentation can help to make the process
more rigorous. With regards to circularity, this relates to making sure that the participants
have strong sustainability and circularity aspirations; that they have the skills and knowledge
that are necessary to experiment towards circularity; and that they have a supportive network
to achieve their aspirations. This adds to previous findings about the importance of carefully
selecting the participants who join the efforts (Bocken et al., 2017). It is important to
understand that they never enter into the process with a blank slate. Rather, they have a set
of predetermined means – their identity, their skills and knowledge, and their social network
– that influence it. We argue that recognizing and leveraging these means can help improve
the process.
5.2. CONTRIBUTION TO CIRCULAR BUSINESS MODEL EXPERIMENTATION PRACTICE
The practical research output of this study is a set of principles that can help improve circular
business model experimentation. The first three principles relate to the effectual logic and
behavior of the participants before the process: what they find important, what they know
and whom they know. Recognizing these elements can be used to compose stronger teams
for experimentation. In particular, it can be used to identify participant profiles with useful
capabilities, for example: a strong personal drive to innovate towards sustainability and a
circular economy, good knowledge of the scientific method, an understanding of
environmental impact assessments, and a network of supportive actors that can be used to
support and widen the perspective of the process.
We also propose a set of principles for during experimentation. During experimentation, the
participants can formulate their assumptions in terms of what they think they need to find
out about their ideas. They can prioritize which assumptions to test by looking at what they
can do right now, and whom they know who can support or who is needed for the inquiry
process. The participants benefit from defining concrete metrics to guide their search
process. This is to ensure an element of rigor and goal orientation within a largely effectual
process. We provide an example set of metrics (Appendix C) that can be used as inspiration
to find an appropriate metric. The search for an appropriate metric can be guided by
questions such as: how do we know if we are on the right track? What do we want to achieve?
How do we measure progress? We learned throughout the three workshops that defining a
key metric for each experiment helps to focus the efforts, and that it helps to be clear about
the intended outcome of an experiment. This is in line with earlier propositions for a metric-
based approach to business model experimentation (Croll and Yoskovitz, 2013; Heikkilä et al.,
ACCEPTED VERSION
2016). It is important to highlight that metrics do not have to be quantitative. Sometimes,
qualitative metrics are more meaningful, especially when a business model is new and has no
prior history (Antikainen et al., 2017b). These principles need further research to understand
when and how they can be used to experiment more successfully.
5.3. LIMITATIONS OF THIS STUDY
We highlight several limitations of this study. First, It is important to note that we conducted
three workshops: two in the Netherlands and one in Switzerland; one with novice
entrepreneurs, one with an incumbent and one with growth-oriented and more experienced
entrepreneurs. This provides a solid data foundation, but is limited in terms of organizational
(no mid-sized company, for example) and cultural richness (no emerging or developing
country context). Second, there are potentially other ways to explain and describe the
decision-making logic during business model experimentation. We found an effectual logic
and behavior to be useful in this context. This does not mean that other theoretical
frameworks may not also shed light on the underlying logic of how the participants form a
theory of value about their envisioned business models. Third, the proposed principles need
further testing and refining, especially with regards to the metrics. Previous research has
collected a set of metrics to guide business model experimentation (Croll and Yoskovitz, 2013;
Heikkilä et al., 2016). It is important to better understand how metrics can be used during
circular business model experimentation, especially how they can help to conduct ‘circularity
checks’ (Bocken et al., 2018).
6. CONCLUSION
This study has shown that analyzing their available means – what they find important, what
they know, and whom they know – can help to better understand how the participants
develop and test their assumptions during circular business model experimentation. We also
showed how a better understanding of this underlying process can help improve it. In
particular, before experimentation, it can help to form a strong circular oriented team with
participants who care about circularity, know about it, and have a network of supporting
stakeholders to explore circularity from an ecosystem perspective. Moreover, during
experimentation, we propose that the participants can formulate their assumptions in terms
of what they need to find out about their ideas, that they can prioritize what to test based on
what they can do right now with what is available, and that they benefit from defining
concrete metrics to guide their search process. Future research is needed to further increase
our understanding of the experimentation process. In particular, it is important to further
investigate, for example, how to compose effective experimentation teams, how to choose
an appropriate metric for an experiment, and how to organize more inter-organizational
business model experimentations for ecosystem level change towards sustainability and a
circular economy.
REFERENCES
Antikainen, M., Aminoff, A., Kettunen, O., Sundqvist-Andberg, H., Paloheimo, H., 2017a.
Circular economy business model innovation process – Case study, in: Smart Innovation,
Systems and Technologies. https://doi.org/10.1007/978-3-319-57078-5_52
ACCEPTED VERSION
Antikainen, M., Aminoff, A., Paloheimo, H., Kettunen, O., 2017b. Designing circular business
model experimentation - Case study. ISPIM Innov. Forum 1–14.
Baden-Fuller, C., Morgan, M.S., 2010. Business models as models. Long Range Plann.
https://doi.org/10.1016/j.lrp.2010.02.005
Blank, S., 2013. The four steps to the epiphany, Fifth edit. ed. K&S.
Blomsma, F., Brennan, G., 2017. The Emergence of Circular Economy: A New Framing Around
Prolonging Resource Productivity. J. Ind. Ecol. 21, 603–614.
https://doi.org/10.1111/jiec.12603
Bocken, N., Miller, K., Weissbrod, I., Holgado, M., Evans, S., 2019. Slowing resource loops in
the circular economy: An experimentation approach in fashion retail, in: Smart
Innovation, Systems and Technologies. https://doi.org/10.1007/978-3-030-04290-5_17
Bocken, N., Schuit, C.S.C., Kraaijenhagen, C., 2018. Experimenting with a circular business
model: Lessons from eight cases. Environ. Innov. Soc. Transitions.
Bocken, N.M.P., Antikainen, M., 2019. Circular business model experimentation: Concept and
approaches, in: Smart Innovation, Systems and Technologies.
https://doi.org/10.1007/978-3-030-04290-5_25
Bocken, N.M.P., Miller, K., Weissbrod, I., Holgado, M., Evans, S., 2017. Business model
experimentation for circularity: Driving sustainability in a large international clothing
retailer. Econ. Policy Energy Environ. https://doi.org/10.3280/EFE2017-001006
Bocken, N.M.P., Schuit, C.S.C., Kraaijenhagen, C., 2018. Experimenting with a circular business
model: Lessons from eight cases. Environ. Innov. Soc. Transitions.
https://doi.org/10.1016/j.eist.2018.02.001
Bocken, N.M.P., Short, S.W., 2016. Towards a sufficiency-driven business model: Experiences
and opportunities. Environ. Innov. Soc. Transitions 18, 41–61.
https://doi.org/10.1016/j.eist.2015.07.010
Brown, P., Bocken, N., Balkenende, R., Brown, P., Bocken, N., Balkenende, R., 2019. Why Do
Companies Pursue Collaborative Circular Oriented Innovation? Sustainability 11, 635.
https://doi.org/10.3390/su11030635
Brown, T., 2008. Design thinking. Harv. Bus. Rev. 86. https://doi.org/10.1145/2535915
Calabretta, G., Gemser, G., Karpen, I., 2016. Strategic Design: 8 Essential Practices Every
Strategic Designer Must Master. BIS Publishers.
Camuffo, A., Cordova, A., Gambardella, A., Spina, C., 2019. A Scientific Approach to
Entrepreneurial Decision Making: Evidence from a Randomized Control Trial. Manage.
Sci. https://doi.org/10.1287/mnsc.2018.3249
Cook, T., Campbell, D., 1979. Quasi-Experimentation: Design and Analysis Issues for Field
Settings. Houghton Mifflin Company, Boston.
Croll, A., Yoskovitz, B., 2013. Lean Analytics. O’Reilly Media Inc., Sebastopol.
Davis, F.D., 1989. Perceived Usefulness, Perceived Ease Of Use, And User Acceptance of
Information Technology. MIS Q. 13, 319–40.
Denyer, D., Tranfield, D., Van Aken, J.E., 2008. Developing design propositions through
research synthesis. Organ. Stud. https://doi.org/10.1177/0170840607088020
Felin, T., Gambardella, A., Stern, S., Zenger, T., 2019. Lean startup and the business model:
Experimentation revisited. Long Range Plann. https://doi.org/10.1016/j.lrp.2019.06.002
Felin, T., Zenger, T.R., 2009. Entrepreneurs as theorists: on the origins of collective beliefs and
novel strategies. Strateg. Entrep. J. https://doi.org/10.1002/sej.67
Foss, N.J., Klein, P.G., Bjørnskov, C., 2019. The Context of Entrepreneurial Judgment:
Organizations, Markets, and Institutions. J. Manag. Stud.
ACCEPTED VERSION
https://doi.org/10.1111/joms.12428
Geissdoerfer, M., Savaget, P., Bocken, N.M.P., Hultink, E.J., 2017. The Circular Economy – A
new sustainability paradigm? J. Clean. Prod. 143, 757–768.
https://doi.org/10.1016/j.jclepro.2016.12.048
Heikkilä, M., Bouwman, H., Heikkilä, J., Solaimani, S., Janssen, W., 2016. Business model
metrics: an open repository. Inf. Syst. E-bus. Manag. https://doi.org/10.1007/s10257-
015-0286-3
Kerr, W.R., Nanda, R., Rhodes-Kropf, M., 2014. Entrepreneurship as experimentation. J. Econ.
Perspect. https://doi.org/10.1257/jep.28.3.25
Knight, F.H., 1921. Risk, Uncertainty, and Profit. Houghton Mifflin, Boston.
Konietzko, J., Bocken, N., Hultink, E.J., 2020a. A Tool to Analyze, Ideate and Develop Circular
Innovation Ecosystems. Sustainability 12, 417. https://doi.org/10.3390/su12010417
Konietzko, J., Bocken, N., Hultink, E.J., 2020b. Circular ecosystem innovation: An initial set of
principles. J. Clean. Prod. 253, 119942. https://doi.org/10.1016/j.jclepro.2019.119942
Magretta, J., 2002. Why Business Models Matter. Harv. Bus. Rev.
https://doi.org/10.1016/j.cub.2005.06.028
Osterwalder, A., Pigneur, Y., Bernarda, G., Smith, A., 2014. Value proposition design,
Strategyzer series. https://doi.org/10.1017/CBO9781107415324.004
Patton, M.Q., 2002. Qualitative Research & evaluation methodos. SAGE Publ.
https://doi.org/10.1037/a0033788
Richardson, J., 2008. The business model: an integrative framework for strategy execution.
Strateg. Chang. 17, 133–144. https://doi.org/10.1002/jsc.821
Ries, E., 2011. The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to
Create Radically Successful Businesses, Crown Business. https://doi.org/23
Romme, A.G.L., Reymen, I.M.M.J., 2018. Entrepreneurship at the interface of design and
science: Toward an inclusive framework. J. Bus. Ventur. Insights 10.
https://doi.org/10.1016/j.jbvi.2018.e00094
Saldaña, J., 2013. The Coding Manual for Qualitative Researchers, Sage Publication.
https://doi.org/10.1109/TEST.2002.1041893
Sarasvathy, S.D., 2008. Effectuation - Elements of Entrepreneurial Expertise. Edward Elgar
Publishing Limited.
Sarasvathy, S.D., 2001. Causation and effectuation: Toward a theoretical shift from economic
inevitability to entrepreneurial contingency. Acad. Manag. Rev. 26, 243–263.
https://doi.org/10.5465/AMR.2001.4378020
Schuit, C.S.C., Baldassarre, B., Bocken, N., 2017. Sustainable business model experimentation
practices: evidence from three start-ups, in: Bakker, C., Mugge, R. (Eds.), Product
Lifetimes and the Environment (PLATE) 2017 - Conference Proceedings. Delft University
of Technology and iOS Press, Delft, pp. 370–376.
Simon, H.A., 1996. The sciences of the artificlal. MIT Press.
Van de Ven, A.H., 2007. Engaged scholarship: A guide for organizational and social research.
Oxford Univ. Press 344. https://doi.org/10.1017/CBO9781107415324.004
Weissbrod, I., Bocken, N.M.P., 2017. Developing sustainable business experimentation
capability – A case study. J. Clean. Prod. 142, 2663–2676.
https://doi.org/10.1016/j.jclepro.2016.11.009
ACKNOWLEDGMENTS
ACCEPTED VERSION
This work was made possible by the Marie-Sklodowska-Curie Innovative Training Network
“Circ€uit” - Circular European Economy Innovative Training Network, within the Horizon 2020
Programme of the European Commission (grant agreement number: 721909). The authors
gratefully acknowledge the support of the European Commission and the contributions of
partners in this project.
APPENDICES
APPENDIX A - INTERVIEW THEMES AND QUESTIONS FOR THE FIRST WORKSHOP
Interview
Interview themes/questions
One interview after session
The workshop material:
1) What are your assumptions?
2) How do you want to test them?
3) How are you going to measure this?
4) When do you know whether you are on the right track?
Reflection
3) How helpful was it to think in terms of assumptions?
4) How did you formulate assumptions?
5) How did you prioritise them?
One interview during the testing
1) How is the testing going?
2) What are you testing?
3) What exactly are you measuring?
One interview after the testing
1) How did the testing go?
2) What have you tested?
3) What have you learned?
4) How does the testing experience help you move forward?
APPENDIX B – FEEDBACK FORM
ACCEPTED VERSION
APPENDIX C –EXAMPLE METRICS FOR DESIRABILITY, VIABILITY AND CIRCULARITY
Tool assessment form
You just used the test cards and the validation graph (see image below). Its purpose is to
understand the assumptions underlying a business idea, and to decide how to test them,
and what to test first.
Please quickly answer the following questions.
________________________________________________________________________
1. The tool is useful to address the purpose stated above.
Please explain your answer:
________________________________________________________________________
2. The tool is easy to use.
Please explain your answer:
________________________________________________________________________
3. Other remarks:
Test card
Assumption: We believe that...
Test: To test that, we will...
Metric: And measure...
Criteria: We are right if...
ACCEPTED VERSION
Figure 2 – Some examples of desirability a nd viab ility metri cs used to trigger pa rticipants (based on Croll and Yoskovitz,
2013; Heikkilä et al., 2016).
Figure 3 – Some examples of potentia l circu larity metrics used to trigger participants (see Konietzko et al., 2020a for
details).
ACCEPTED VERSION
