Jakobsen, Henning Seje ; B ix, Jacob; Jakobsen, Rune Seje
A icle
Un a eling da a om an idea managemen sys em o 11
adical inno a ion po olios: Key lessons and a enues o
a i icial in elligence in eg a ion
Jou nal o Inno a ion and En ep eneu ship
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Jakobsen, Henning Seje ; B ix, Jacob; Jakobsen, Rune Seje (2024) : Un a eling
da a om an idea managemen sys em o 11 adical inno a ion po olios: Key lessons and a enues
o a i icial in elligence in eg a ion, Jou nal o Inno a ion and En ep eneu ship, ISSN 2192-5372,
Sp inge , Heidelbe g, Vol. 13, Iss. 1, pp. 1-24,
h ps://doi.o g/10.1186/s13731-024-00368-6
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/290331
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Open Access
© The Au ho (s) 2024. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License, which pe mi s
use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal
au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i changes we e made. The images o o he hi d
pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line o he ma e-
ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended use is no pe mi ed by s a u o y egula ion o
exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://
c ea i ecommons.o g/licenses/by/4.0/.
RESEARCH
Jakobsene al.
Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
h ps://doi.o g/10.1186/s13731-024-00368-6
Jou nal o Inno a ion and
En ep eneu ship
Un a eling da a omanidea managemen
sys em o 11 adical inno a ion po olios: key
lessons anda enues o a i icial in elligence
in eg a ion
Henning Seje Jakobsen1* , Jacob B ix2 and Rune Seje Jakobsen3
Abs ac
In s a egic and adical inno a ion, he deg ee o unce ain y and he amoun o com-
plexi y is much highe compa ed o ‘business as usual’. The e o e, idea managemen
sys ems a e o en used o suppo such inno a ion p ocesses. An in e es ing ques ion
is wha we can lea n om s udying da a in such idea managemen sys ems and wha
po en ial implica ions we can de i e om he inno a ion managemen li e a u e. In
his s udy, we we e allowed o access and analyze da a om he same idea manage-
men sys em used in 11 adical inno a ion p ojec s om he yea s 2012–2018. Ou
analysis un a els 8 indings ha in di e en ways nuance o challenge cu en esea ch
on inno a ion managemen . Finally, we discuss how he in eg a ion o a i icial in el-
ligence (AI) in idea managemen sys ems can suppo inno a ion eam membe s
in inc easing he inno a ion po en ial o he ideas ha a e elabo a ed.
Keywo ds: Idea managemen , Inno a ion managemen , Radical inno a ion, Ac ion
con e gence, Pa e n ecogni ion, A i icial in elligence, Abso p i e capaci y, Inno a ion
capaci y building
In oduc ion
The abili y o explain and p esc ibe how es ablished o ganiza ions c ea e adical inno-
a ion is inc easingly ge ing schola ly a en ion (Lei e e al., 2000; Lassen e al., 2006;
Sale no e al., 2015; O’Conno e al., 2018; Goduschei & Faullan , 2018). The scien i ic
a en ion o adical inno a ion has, o example, been aimed a explaining he need o
speci ic compe encies (U e back, 1994), he di e en s ages an idea mus go h ough
om i s ecogni ion o i s de elopmen and comme cializa ion (Lei e e al., 2000; B ix
& Jakobsen, 2013), and how measu emen s can be used o assess ideas in adical inno a-
ion po olios (K is iansen & Ri ala, 2018). In his con ex , adical inno a ion is con-
cep ualized as a sequen ial p ocess, whe e he s a egic di ec ion o new p oduc s is
speci ied. Howe e , he ou pu is unclea , and he p ocess o c ea ing new ou pu s con-
ains di e en ypes and deg ees o unce ain y (O’Conno & Rice, 2013). Ce ain ac i i-
ies a e conside ed i al o no lea ing adical inno a ion o chance, such as applying
*Co espondence:
[email p o ec ed]
1 Depa men o Biological &
Chemical Enginee ing, Facul y
o Technical Science, Aa hus
Uni e si y, Aa hus, Denma k
2 Aalbo g Uni e si y Business
School, Aalbo g Uni e si y,
Aalbo g, Denma k
3 Cen e o Clinical Resea ch,
No h Denma k Regional
Hospi al, Hjø ing, Denma k
Page 2 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
meaning ul sys ema ic p ocesses and me hods o he inno a ion p ocesses (Sale no
e al., 2015). Applying meaning ul sys ema ic p ocesses and me hods is conside ed a co -
ne s one o oday’s adical inno a ion (Sale no e al., 2015). Such sys ema ic p ocesses
equi e, e.g., (1) acknowledging he exis ence o unce ain y, (2) allowing ha unce -
ain y o be pa o he inno a ion p ocess, and (3) managing he p ocess o educing
unce ain y (O’Conno & Rice, 2013). C ea ing adical inno a ion is a complex, con ex -
dependen phenomenon (Funnell & Roge s, 2011) o which a linea cause–e ec model
canno gua an ee adical inno a ion ou comes will occu (B asil e al., 2021). P ocesses
o c ea ing adical inno a ion equi e di e en ypes o managemen a di e en poin s
in ime (O’Conno e al., 2018) as well as he e is an inc easing demand o a manage-
men sys em o iden i ying and educing unce ain y as pa o a knowledge-c ea ing
lea ning p ocess (B ix and Ho sage , 2022). A g owing and in e es ing aspec o he
managemen o adical inno a ion is how o suppo he inno a ion p ocess using ech-
nology, such as idea managemen sys ems o o e come limi a ions o human in o ma-
ion p ocessing capabili ies (e.g., Hae ne , 2021). Ano he heme ha is eme ging is how
a i icial in elligence can be used as a alue-adding componen o idea managemen sys-
ems (Amabile, 2020), e.g., o he idea ion and de elopmen o ideas in such a way ha
he sea ch ou ines o he pa icipan s in he inno a ion p ojec ge o ame and de ine
new oppo uni ies (Keding, 2021) as a pa o inno a ion capaci y (Mikelsone e al.,
2022a, 2022b).
While empi ical esea ch on adical inno a ion is no uncommon (e.g., O’Conno ,
1998; Lassen e al., 2006; B asil e al., 2021), his s udy epo s on longi udinal da a ha
a e a ely accessible o esea che s and he public: we we e allowed access o 11 adi-
cal inno a ion po olios ha we e sys ema ically elabo a ed using he same idea man-
agemen sys em and sys ema ic inno a ion me hod om 2012 o 2018. By applying an
explo a o y esea ch s a egy, we a e in e es ed in unde s anding he p ocess o elabo-
a ing adical inno a ion ideas om he ea ly s ages o he s ep, whe e business models
a e buil and eady o be comme cialized. The poin o in e es wi h his da a a ailable is
o conduc explo a o y esea ch led by he ollowing esea ch ques ion:
Wha can we lea n om analyzing da a om 11 adical inno a ion po olios, and
wha a e he implica ions o inno a ion managemen esea ch om hese indings?
The heo e ical lens we apply is abso p i e capaci y (Cohen & Le in hal, 1990; Lane
e al., 2006; Zah a & Geo ge, 2002). Abso p i e capaci y is o en desc ibed as an ongoing
p ocess o an o ganiza ion and i s indi iduals ha wan o gene a e no el ideas o new
o highly imp o ed p oduc s o echnological p ocesses (Fab izio, 2009). Abso p i e
capaci y is a ecognized app oach by which an o ganiza ion iden i ies new in o ma ion
and ansla es i in o a comme cial ad an age (Cohen & Le in hal, 1990). The p emise
o abso p i e capaci y is ha indi idual and o ganiza ional lea ning p ocesses a e con-
nec ed. A c ucial mechanism is ha he igh indi idual can iden i y he po en ial alue
o new in o ma ion (Volbe da e al., 2010). Abso p i e capaci y, he e o e, e e s o he
capaci y o an iden i ie (o ecipien ) o assimila e new in o ma ion, bo h in e nal and
ex e nal, o c ea e alue, e.g., by imp o ing compe i i eness o c ea ing new o imp o ed
p oduc s. Inspi ed by Hae ne (2021), we also ind i in e es ing wi h he da a a ailable
Page 3 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
o discuss how a i icial in elligence (AI1) can suppo adical inno a ion p ocesses in
combina ion wi h he use o idea managemen sys ems as a subs an ial ye only spa sely
unde s ood oppo uni y (Fülle e al., 2022; Hae ne , 2021). To ou knowledge, obs acles
ela ed o such a many- aced challenge a e ye o be desc ibed and, i possible, exempli-
ied. Wi h his s udy, we open he lid o his black box.
To c ea e new knowledge o he p oblems iden i ied abo e, we epo on da a om an
idea managemen sys em ha was used in 11 adical inno a ion p ojec po olios om
he yea s 2012–2018. These 11 adical inno a ion p ojec s ollowed he same sys ema ic
inno a ion me hod o which an idea managemen sys em named Rose a was designed.
This sys ema ic me hod and use o he same idea managemen sys em enables us o build
an embedded case s udy as a esea ch s a egy (Yin, 2013). Mo eo e , since ou analyzes
deal wi h knowledge and he in ol emen o inspi a ion, we explo e and discuss how AI
may con ibu e o cen al unc ions o idea explo a ion in he Rose a idea managemen
sys em. This e lec i e exe cise o p o ide heo e ical unde pinnings and unde s and i s
place in he schola ly wo k o he abso p i e capaci y li e a u e p o ides help o inno a-
ion p o essionals in hei managemen and elabo a ion o ideas o adical inno a ion.
Theo e ical backg ound
Fo mo e han 30yea s, abso p i e capaci y (Cohen & Le in hal, 1990) has been sub-
jec o sc u iny and acknowledged o i s g ea impo ance wi hin se e al esea ch ields,
such as lea ning, inno a ion, and manage ial cogni ion (Volbe da e al., 2010), whe eas
abso p i e capaci y has been he bed ock o heo ies o inno a ion and he capaci y o
inno a e (Zoue al., 2018). While abso p i e capaci y can help o ganiza ions p ocess
in o ma ion mo e e icien ly owa ds alue c ea ion, he app oach also has i s limi a-
ions. I s use ulness depends on he abili y and ingenui y o he employees who use he
me hod and on he o ganiza ional con ex , and he in e nal wo k lows and p ocesses
(B ix, 2019). The e o e, abso p i e capaci y is o en seen as employee-d i en inno a-
ion, whe e he collec ion, s uc u ing, and elabo a ion o ideas a e pu in o a sys em
o ensu ing assessmen , pu sui , and o en e en emune a ion (Knoppen e . al., 2022;
Sale no e al., 2015). In he sense o ans o ming in o ma ion—o new ideas—in o alue
c ea ion and new p oduc s o se ices(Duan e al., 2020). We iew abso p i e capaci y
as he unk ha holds he b anch o inno a ion capaci y (Zou e al., 2018).
In ela ion o he use o idea managemen sys ems wi h an abso p i e and inno a ion
capaci y pe spec i e, endencies ha e gone om p oposing sugges ion boxes o much
mo e ad anced p og ams and amewo ks (Go ski & Heinekamp, 2004). This way, idea
managemen sys ems can be de ined as “ o malized me hods o cap u ing, examining,
nu u ing and de eloping ideas c ea ed wi hin an o ganiza ion” (Nilsson e al., 2002,
p. 500). While mul iple s udies examine di e en ypes o idea managemen sys ems—
especially open-sou ce communi ies (e.g., Dahlande & Gann, 2010), we do no know
much abou he ole o idea managemen sys ems in he con ex o adical inno a ion.
So a , limi ed empi ical esea ch is a ailable conce ning idea managemen sys ems a
he ea ly s ages o inno a ion (Be e a, 2019). A ecen publica ion by Mikelsone e al.
(2022a) dis inguishes be ween idea managemen sys ems dependen on whe he ideas
1 We use he e m AI as an abb e ia ions o A icicial In elligence. We use he no ions AI and machine lea ning in e -
changeably.
Page 4 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
a e submi ed wi h a clea inno a ion pu pose o no — e e ed o as passi e o ac i e—
and whe he he ideas o igina e inside he o ganiza ion’s bounda y o no — e e ed o
as in e nal o ex e nal. In ou wo k, we expand he no ion by Mikelsone e al. (2022a)
and discuss he in eg a ion o AI in a s epwise idea managemen p ocess (Mikelsone
e al. (2022b). Hence, a sequen ial idea managemen sys em wi h AI enhances inno a-
ion capaci y, which may conclusi ely inc ease he likelihood o adical inno a ion. In
his scena io, we mus i s acknowledge ha A i icial In elligence and machine lea n-
ing models use algo i hms (ma hema ical models) o ind pa e ns in digi ally a ailable
in o ma ion (da a) and con e hem in o ele an knowledge (Nei o i e al., 2021). We
use he no ions AI and machine lea ning in e changeably. Howe e , implemen ing an AI
model calls o p io conside a ion o he da a p esen ed o a model. Fo ins ance, be o e
de eloping he AI model o ou sequen ial idea managemen sys em, we inspec and
clean he da a, such as conside ing he o ma o he in o ma ion ep esen ing an idea
(s uc u ed s. uns uc u ed), handling ou lie s, and missing alues, e c. In his con ex ,
s uc u ed da a a e s o ed in p ede ined o ma s, such as a da e, sende , and ead/un ead
in an email accoun , whe eas uns uc u ed da a a e he con ex o he email, such as ee
ex o a pic u e. The e o e, o a sequen ial idea managemen sys em, such as ou s, we
need o add ess how ideas a e a ailable o he model, e.g., i p esen ed as ee ex . Ou li-
e s a e obse a ions ha ha e an abno mal dis ance o he emaining samples, such as
a da e ou side he adi ional calenda in a mail. Missing alues a e when we miss an
obse a ion om all ou ea u es, such as i we do no ha e a sende o an email. Han-
dling missing alues in a sequen ial idea managemen sys em may be handled in one o
he h ee adi ional ways: (1) dele ing he column o ensu e comple eness bu wi h he
p ice o losing in o ma ion, (2) impu a ion o bes -guess alue in uni- o mul i a ia e
Fig. 1 Funnel o in o ma ion p ocessing in his a icle. Sou ce: Own line-up
Page 5 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
me hods bu wi h he p ice o simula ing obse a ions, o (3) handling he missingness
in he AI model class bu wi h he p ice o being es ic ed o ewe model choices, e.g., a
sequen ial model o p obabilis ic g aphical model. The e o e, using AI in sequen ial idea
managemen sys ems, we i s conside he ideas we p o ide o he model wi hin ou
sequen ial idea managemen sys em, Rose a, in he hopes o eaching adical inno a-
ion as he indings. Figu e1 illus a es he agile unnel o in o ma ion om abso p i e
capaci y o indings in ou wo k.
Conside a ion and examples o he oppo uni ies, bene i s, and po en ial pi alls o
de eloping and using sequen ial idea managemen sys ems and machine lea ning a e
discussed la e in his pape .
Me hodology
Resea ch s a egy
An embedded mul i-case s udy (Yin, 2013) is applied as a esea ch s a egy. We epo
on da a om a so wa e idea managemen sys em used on 11 cases o adical inno a ion.
All cases we e based on he same inno a ion model, ollowed he same implemen a ion
s uc u e, and had he same ex e nal p ojec manage . The de ails ela ed o hese cases
a e un olded below. Pu posi e sampling was used as a sequen ial design (deVaus& de
Vaus, 2013), whe e he 11 cases ollow each o he sequen ially, he i s s a ing in 2012
and he las in 2018. The ad an age o sequen ial design and mul i-cases is ha he expe-
iences om he i s case can p o ide inspi a ion, new ideas, and me hods ha can
in luence he subsequen empi ical collec ion in he subsequen case s udy (de Vaus &
de Vaus, 2013).
P esen ing hecases andquali ying hesample
All cases we epo on in he s udy we e comple ed ollowing he same sys ema ic inno-
a ion p ocess, he ‘C ea i e Idea Solu ion model’ (CIS). CIS was c ea ed in he la e
1990’ies, i s desc ibed by Jakobsen and Rebsdo (2003) and la e published, e.g., by
B ix and Jakobsen (2013) and by Jakobsen (2021) in a doc o al disse a ion. CIS is a
sequen ial phase model con aining ou phases, whe e each phase has ou ac i i ies o
s eps, de ines wi h bo h a p ocess ( o ollow) and a amewo k ( o do). The da a we epo
on s em om a so wa e sys em named Rose a, which was de eloped as a so wa e Idea
Managemen Sys em ollowing he same phases and ac i i ies as he CIS model ela ed
o he middle phase called P e-jec . We s a by p esen ing he phases, ac i i ies, and
p emises o he CIS model, and he ea e we explain he Rose a idea managemen sys-
em and he da a we ha e a ailable o ou s udy. The pu pose o p o iding an ex ended
explana ion o bo h he CIS model and he Rose a idea managemen sys em is o quali y
ou sample o how and why he esul s om he di e en adical inno a ion p ojec s a e
compa able and ele an o epo on.
Recognized inno a ion models such as he ‘Disco e y, Incuba ion, and Accele a ion
(O’Conno e . al., 2008), ‘open inno a ion’ (Chesb ough, 2006), ‘dis up ion’ (Ch is ensen,
1997), and ‘Theo y U’ (Scha me , 2007), and he CIS model ocuses on desc ibing he
need o adical, dis up i e, and ans o ma i e inno a ion in a s a egic pe spec i e. The
CIS model is cha ac e ized by he ollowing c i e ia: i s , ideas o adical inno a ion
Page 6 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
in ol e knowledge c ea ion om di e en expe s and compe ence con ibu o s ha a e
bo h in e nal and ex e nal o he o ganiza ion (Zobel, 2017). Second, in he CIS model,
decisions and judgmen s a e suspended (no limi ed o pos ponemen ), meaning ha no
ac i e assessmen o e alua ion p ocess akes place du ing idea and concep de elop-
men as his is eplaced wi h ac ions s imula ion, e.g., he blending me hod (Fauconnie
& Tu ne , 2008), as well as di e en po olio models o ensu e he o e iew (Fig.2).
A hi d cha ac e is ic is ha i is no possible o dele e ideas, and i is no possible
o me ge ideas. I is only possible o add new in o ma ion and knowledge, and i hese
addi ions con ibu e added alue, such as new o be e unde s andings, he ideas can
be mo ed o he nex s ep in he p ocess. The e o e, all eco ds a e s o ed in Rose a o
conduc ing he handling as well as mo emen . The CIS model has he ollowing phases
and ac i i ies:
Phase 1: (la e al) ocus The i s phase in CIS is a la e al ocus, whe e he pu pose is o
de ine he (new) inno a ion ask. The i s ac i i y in his phase is o collec knowledge
ega ding wha is known ela ed o he challenge, p oblem o needs (Chil on e al., 2007),
e.g., using he Fo ce Field analysis (Lewin, 1943). The second ac i i y is o challenge hese
assump ions, done by he eam membe s and he knowledge ha has been collec ed.
The hi d ac i i y is o c ea e new insigh s wi h he pu pose o iden i ying possible new
a eas o changing di ec ion and b eaking he exis ing pa e n (B ix & Jakobsen, 2013).
The ou h ac i i y is o de ine he main inno a ion ask as well as de ining he igh c ea-
i e ools o use and he igh expe s and inspi a ions o implemen , which leads o he
inno a ion ques ion. Ha ing de ined he inno a ion ques ion, he idea ion in phase wo
can be conduc ed.
Phase 2: p e‑jec In he p e-jec phase, he i s ac i i y is o c ea e new inpu s based
on he inno a ion ques ion, mainly using c ea i e echniques in ol ing he igh peo-
ple. The inpu s ha a e added enough knowledge and ega ded as po en ial a e mo ed
o he nex s ep. An idea is, as a esul o his, de eloped. De eloping ideas is based
on pa e n ecogni ion, associa ion, and blending heo y (Fauconnie & Tu ne , 2008;
Oakley e al., 2017) by in ol ing inspi a ion om o he inpu s and ideas. Those ideas
Fig. 2 The C ea i e Idea Solu ion model Sou ce: B ix and Jakobsen (2013)
Page 7 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
ha appea wi h clea s eng hs and o en po en ial challenges a e ans e ed in o he
nex s ep, he concep le el. In his s ep, he concep s a e supplied by new knowledge
and inspi a ion om a eas co e ing he same p inciple o explo e he ull po en ial in a
na ow a ea, and hen/i a po en ial di ec ion opens a o able oppo uni ies, he con-
cep is ans e ed o a new s ep, he design le el. In he design phase, a e ical inno-
a ion p ocess is conduc ed (B ix & Jakobsen, 2013) o b ing he desc ibed ho izon al
p ocessed idea o a holis ic concep handling all aspec s in he e ical chain (Keeley
e al.,; 2013) o p e o yping (B ix & Jakobsen, 2014;Jakobsen & Hansen, 2007). The
CIS model has a s ic sepa a ion o ho izon al and e ical inno a ion, as he p ocess
o ho izon al inno a ion ( inding o iginali y) is o en sabo aged by ye - o-be- ouched
a gumen s om he e ical alue chain (Jakobsen e al., 2008). The in o ma ion and
da a low in he p e-jec phase a e handled in he idea managemen sys em Rose a.
Phase 3: p ojec In he p ojec phase, hose es ablished and s onges desc ibed
designs in ela ion o he desi ed ou come based on de ined ocus a e p ocessed indi-
idually o po en ial implemen a ion. The p ojec phase is no a pa o Rose a.
Phase 4: pos ‑jec The pos -jec phase is iden ical o ocus bu is no pos poned un il
he p oblem a ises wi h is ini ia ed in pa allel wi h implemen a ion (o wha e e else is
ini ia ed a e he p ojec phase). This is o s i e o adical inno a ion based on chal-
lenge, no p oblem-sol ing.
The idea managemen sys em: Rose a
The Idea Managemen sys em ‘Rose a’ wo ks as a sys ema ic po olio ool ailo -
made o he second phase in he CIS model, he p e-jec phase. Rose a is based on
he p emise o abso p i e capaci y, which implies ha an in low o new in o ma ion
mus be guided h oughou he inno a ion p ocess by he indi idual eam membe s
aking pa in he p ojec . All in o ma ion, such as no es, pho os, links o homepages,
e c., a e documen ed igh om he s a o he p e-jec phase and kep un il he inal
ac i i y has been comple ed.
Rose a was o iginally de eloped in he la e 1990’ies. The i s e sion o Rose a was
made in he ela ional da abasePa adox,and he second e sion o Rose a was buil
upon Mic oso Access. Based on he lea ning om wo king wi h hese wo ea ly e -
sions in p ac ice, he cu en e sion o Rose a e sion III, was upda ed in Mic oso
Access ollowing he ou phases desc ibed in he p e-jec phase o he CIS ame-
wo k (B ix & Jakobsen, 2013). Despi e a his o ical pe spec i e wi h imp o emen s,
bo h he desc ip ion and he analysis o he da a deal only wi h da a om Rose a
e sion III, and he sys em is subsequen ly simply called Rose a. Table1 below illus-
a es he numbe o adical inno a ion p ojec s, whe e Rose a, in di e en e sions,
has been he idea managemen sys em suppo ing he inno a ion p ocess (Table2).
Rose a is a ‘Mul i o Mul i’ ela ion. An inpu can esul in se e al ideas, and
an idea can esul in se e al concep s, e c. A he same ime, se e al inpu s can be
used as inspi a ion o an idea, e c. To b ing inpu s o ideas o he nex ac i i y, new
Page 8 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
Table 1 Timeline o he de elopmen o he Rose a sys em wi h in ol ed majo p ojec s
Sou ce: Au ho s’ de elopmen
Yea Pla o m Coun y o use N p ojec s Sec o s/indus ies
1998 - Pa adox (Bo land) Denma k 2 Tele ision, be e age
2005 - Mic oso Access Denma k, Ge many 12 Indus ial p ojec s
2011 – Mic oso Access, designed
in e ace, help unc ion,
manual look-up unc ions,
English languages
Denma k, Ge many,
Sweden, No way, Finland,
Ne he lands, USA, Li huania,
Russia, I aly
60 37% a e indus ial p ojec s
om Denma k, 20% a e
municipali y and go e n-
men al p ojec s om Den-
ma k, and 43% a e indus ial
p ojec s loca ed ou side
Denma k
Table 2 The 11 cases
Yea Desc ip ion Clien
2012 Sus ainable packaging Indus y
2012 20% mo e lea ning in p e-school Municipali y
2013 New pu chasing sys em a a Hospi al Go e nmen
2014 Handicap help sys em Municipali y
2014 Sh edde sys em o was e managemen suppo Indus y
2014 Rehabili a ion Municipali y
2015 Visi a ion sys em o ob ain suppo and assis ance Municipali y
2015 Wo kshop sys em o handling ca du ing he epai Indus y
2016 Rehabili a ion in he heal hca e sec o Municipali y
2017 Nex -gene a ion umble d ye Indus y
2017 New s imulan as an al e na i e o obacco Indus y
2018 Food packaging in he dai y indus y Indus y
Fig. 3 The o e iew page in Rose a ollowing he ou s eps in he CIS P e-jec . F om B ix and Pe e s (2015)
Page 15 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
canno uncondi ionally conclude ha mo e inpu s esul in a mo e adical ou ‑
come. A s onge co ela ion is seen be ween he numbe o ideas and he inally
achie ed adicali y, and an e en s onge co ela ion be ween he numbe o con‑
cep s and he inal adicali y achie ed. Finally, he e is a signi ican co ela ion
be ween he numbe o designs and he o al adicali y achie ed in he designs.
The pa e n is ein o ced u he in he p ocess, whe e he o al adicali y o he
en i e po en ial (de ined as ‘Idea Capi al’) inc eases when he e a e mo e ideas,
concep s, and designs.
The a e age μ o c ea ed concep s ( hi d s ep) o he conduc ed p ojec s is 34 wi h a
s anda d de ia ion σ o 16,00. On a e age, 0,73 ex e nal inspi a ions sou ces ha e been
connec ed o each concep as a sou ce o new ac ion.
I can be seen om Fig.12 ha he e is a connec ion be ween g een and blue con-
cep s, jus as he e is a s onge connec ion be ween blue and ed de ined om he
C-box. Mo e gi es mo e. When adding expe s as a pe son o knowledge o a sou ce o
inspi a ion, he measu ed adicali y is inc easingly signi ican .
Finding 6: When expe s in he domain o sou ces o inspi a ion conce ning
expe s om o he domains co e ing he same p inciple a e ela ed o a concep , he
deg ee o adicali y inc eases signi ican ly. This e ec e en inc eases ma kedly and
signi ican ly when he numbe o associa ed expe s inc eases
Ou da a hus shows ha i makes sense (in some con ex s) o wo k wi h adical inno-
a ion (sea ch o ed ideas and concep s) as his implici ly esul s in mo e blue con-
cep s, and blue esul s in g eene (inc emen al) concep s. Adding new aspec s o he
concep s s ep p o ides a signi ican ly inc eased adicali y and seems o be ex emely
e ec i e. Such aspec s could be a pe son wi h huge knowledge in he new a ea o he
concep o an a ea, whe e he undamen al p inciple in ano he domain is explo ed and
used wi h success. This has simila i y o desc ibed in T iz (Al shulle , 1999).
The a e age μ o c ea ed designs ( ou h s ep) o he conduc ed p ojec s is 18 wi h a
s anda d de ia ion σ o 6,55. The a e age o ideas, di ided by colo , alls o all colo s
om he idea phase o he design phase. Inpu is no included in his calcula ion as
colo s do no appea speci ically in his phase. Howe e , he d op is less om concep
o design han om idea o concep . The e a e co espondingly mo e g een esul s han
Fig. 13 The e olu ion o he numbe o g een, blue, and ed esul s o e he h ee o ou phases in Rose a
(le ), in he middle, added wi h he SIP and C-Box alue, and on he igh , he a e age numbe o designs o
all p ojec s di ided by SIP Po olio
Page 16 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
blue and again ed esul s a all le els, bu he numbe is added wi h he calcula ed alue
se up a SIP (seeFig.4), as illus a ed in he igu e in he middle o Fig.13. This ocused
on SIP wi h conscious p io i iza ion hus has he e ec ha he ex en o esul ing espe-
cially blue and g een concep s, consciously o unconsciously, is jus g ea e ( he igu e
on he igh in Fig.13).
Finding 7: The da a un a eling shows a s ong endency owa d ideas being p o‑
cessed agains he de ined expec a ions as “you ge wha you measu e”. A s ong
indica ion o adicali y in he SIP po olio p o ides a ma kedly in luence on he
inno a ion p ocess owa d a mo e adical goal and a endency o ollow hose wi h
he highes po en ial.
All p ojec s in his analysis a e comple ed ollowing he CIS model (B ix & Jakobsen,
2013) and he phases desc ibed in his model’s P e-jec . Fo all p ojec s in he analysis,
he p ojec s ha e been comple ed by es ablishing eams wi h membe s om he com-
pany o o ganiza ion in ques ion, espec i ely, inno a ion consul an s. The dis ibu ion
can be seen in Fig.14, jus as indica ed he e we e he ollowing (non-pa icipa ing) con-
sul an s. I is no a equi emen o p e equisi e o he implemen a ion o CIS o he use
o Rose a ha i is eamwo k, bu his has been he si ua ion in he as majo i y and all
p ojec s in he p esen analysis. The eams a e es ablished be o e he s a o he p ojec
and ollow he en i e p ocess, including he phase be o e he p e-p ojec phase (la e al
ocus) ega ding he en i e ask p ocessing. When compa ing eam size as a unc ion o
he esul ing sco e, a smalle dec ease in e ec is seen as he la ge he g oup becomes.
This co esponds o he esul s, i.e., shown by Laughlin, Sil e and Boh (2006).
Finding 8: G oup size seems o be a o able o he numbe o c ea ed ideas bu does
no seem o ha e a a o able e ec on he c ea ion o ideas o adical inno a ion—
qui e he opposi e o la ge g oups. A good g oup size seems o be 3–5/6 pe sons.
Discussion
I is bo h a s eng h and a sou ce o o e looking po en ial e o s ha all au ho s o his
pape , in hei way, a e amilia wi h Rose a and he comp ehensi e CIS me hod. This
is as a p ocess consul an / acili a o , an associa ed esea che , o a pa icipa ing expe .
Howe e , his also p o ides he abili y o cha ac e ize Rose a as an Idea Managemen
Fig. 14 The eams’ composi ion o employees and consul an s (le ) and eam size as a unc ion o he inal
esul ( igh )
Page 17 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
sys em (Nilsson e al., 2002) and ecognize he alue o Rose a as an abso p i e capac-
i y sys em ela ed o he 11 adical inno a ion p ojec s conduc ed om 2012 o 2018. In
his ega d, he Rose a sys em used in he inno a ion p ojec s, can be cha ac e ized as
an ac i e, in e nal idea managemen sys em (Mikelsone e al., 2022a, 2022b). In Table3
below, we ha e summa ized he s udy’s key indings and implica ions.
F om Finding 1, we see con e gence can be achie ed, e en hough decisions a e no
included, bu new aspec s and possibili ies a e added, he e na owing is achie able, hus
ac ion con e gence. Simila ly, Findings 2 and Findings 3 show ha he e is a co ela ion
be ween he numbe o ideas and achie ed adicali y especially when he e a e c ea ed
mo e han 300 inpu s. Howe e , he gene al co ela ions be ween c ea ed inpu and he
measu ed adicali y in isola ion a e no signi ican and no unambiguous. This is un a e-
led in he analyses in h ee a eas: (1) indings 3 shows ha hose ideas inspi ed by o he
inpu s o he c ea ion p esen a signi ican ly highe pe o mance ega ding adicali y;
Table 3 Summa y o key poin s
Discussion
Inpu IdeasConcep DesignLeading o:
Ac i i y in Ro-
se a ollowed
he CIS model
p e-jec phase
C ea e inpu by he use o c ea i e
echniques: an asy, imagina ion,
inspi a ion and se endipi y
Concep ualized inpu s ha a e
con e ed in o an idea in a na ow
a ea wi h po en ial o explo a ion
Ideas ha a e concep ualized
usually by concep s ha a e
unde s ood wi h he meaning in i s
en i e y
Concep s a e p ocessed e ically
o p o ide a holis ic o e iew
and e eal s a egic pe spec i es
Rose a could wi h ad an- age
be ex ended wi h bo h la e al
ocus ( eason) and Ve ical
inno a ion (ou come)
P ominen
heo y in he
s ep
Associa ion heo y based on he
de ined pa e n ecogni ion ( ocus)
ollowed by pa e n b eaking using
c ea i e echniques
Associa ion heo y combined wi h
blending heo y o explo ing he
po en ial (ho izon al inno a ion)
Sea ch he ho izon o insigh and
knowledge wi hin he domain
pa e ned ou wi h s eng hs and
weaknesses ollowing he Fo ced
Field heo y
Explo a ion o he concep
po en ial by he use o he e -
ical inno a ion p ocess (VIP)
wi h a subsequen desc ip ion
based on Min zbe gs 5Ps
Inc emen al inno a ion wo ks
wi h decision con e gence - i
adical inno a ion ac ion
con e gence appea s p o i able
Suppo o
ac ion in p esen
Possible o ge a o e iew o all
c ea ed inpu by use lis unc ion
Possible o connec inpu s o a
speci ic idea o inspi a ion
Possible o add consciousness-
expanding a eas om ou side
Possible o implemen VIP o ge
holis ic pe spec i e
Inspi a ion ma e s, bu how o
explo e and de elop his?
The con ibu-
ion om ocus
despi e lack o
implemen a ion
in Rose a
In ocus, exis ing knowledge is
p ocessed wi h "wha is". The
known is desc ibed wi h exis ing
cus ome s and cons i u es he
e e ence o he known
Wha is locked pe cep ion, and
wha causes pa alysis? This
de ines wha needs o be
challenged, wha se s a new
(la e al) di ec ion
C ea ion o insigh ega ding wha
is included o ge ideas in new
(la e al) a eas, which c ea i e
echniques make mos sense o
use
Leading o he Inno a ion
Ques ion - he one ha de ines
he di ec ion
La e al ocus is no pa o
p esen Rose a despi e he
ac ha ocus de ines he
amewo k o s a and
wo king wi h Rose a!
The p esen e
added alue
C ea ed inpu s can be added
knowledge, and a iew o all inpu s
can be highligh ed as a lis o pa e n
ecogni ion in inpu
C ea ed ideas can be associa ed
wi h inpu s manually o
inspi a ion and benchma king o
associa ion and blending
Ex e nal esou ces and knowledge
a e implemen ed o he explo a ion
o he eme ging concep , he eby
no el y and IPR
By using he same me hods as
desc ibed in Inpu and Ideas,
e ically p ocessing is enabled
o combina ion o a eas
To con e he manual
ope a ion o sys em-gene- a ed
inspi a ion - and s imu-la ion
in o ma ion o ac ion
The abili y o
ecognize he
alue o new
in o ma ion
Wan o ensu e be e p ocessed
inpu , a en ion o lack o o iginali y
(duplica e) as well as a sou ce o
o iginali y
The c ea i i y and knowledge
om all he c ea ed inpu s has o
be used beyond jus as an ini ia o
o a new idea
The p ocess o combining
knowledge om ex e nal sou ces
such as R&D wi h he pos - ac um
esul s such as IPR
Abso bs holis ic conside a ions
o a concep as he sou ce o
mul iple possible a ian s o he
same concep
Incomp ehensibili y and un-
ce ain y a e sough o be
cla i ied wi h explo a ion and
no by decision
Con ibu ion in
Rose a
"O e iew o all he inpu en e ed in
Rose a o inc ease new alue"
"Inspi a ion om en e ed inpu
and c ea ed ideas"
"Imme sion om ex e nal
esou ces and knowledges"
"Connec ing e ical a eas" Sequen ial wi h e olu ion
p o ide p og essi eness
The ans o ma-
ion o knowled-
ge ans e o im-
p o e nex s ep
Associa ions con ibu e new aspec s
ha canno immedia ely be classi ied
as adding alue, as his will equi e
mo e (nex s ep)
Concep ualized inpu s indica e a
eal alue-c ea ing idea ion-ac i i y
in a na owly conside ed ield bo h
ega ding compa a-bili y and
simila i y in p inciple
Each na ow b eak h ough is
un olded h ough an inclusion
ac i i y o new knowledge and
new conside a ions o explo a ion
o he po en ial
Aspec s o knowledge om widely
b anched a eas in he e ical
chain a e linked o c ea e a
comple e ield o possibili ies in
each chain
The p ocess o p o iding new
knowledge may en ail he isk
o an incomp ensi-ble amoun
o da a wi h in o ma ion and
possibili ies
Obse ed
challenges wi h
exis ing designed
and es ed
Rose a
Lack o o e iew bo h while he
c ea ion o inpu is pe o med and
also a e mo e duplica es han
necessa y a e c ea ed jus like he
inspi a ions sou ce is missing
I is no in ui i e o associa e inpu
wi h ideas, he e o e i only occu s
in mino cases and aluable
knowledge om idea ion is
o e seen
Associa ed a eas o insigh and
knowledge o he blend is limi ed
o he insigh among he eam
membe s
Ve ical Inno a ion P ocess
(VIP) is no ye in eg a ed in o
Rose a bu done as an ac i i y
isola ed om Rose a despi e i
being a cen al pa o he CIS
In o ma ion o associa ing,
blending, and added
knowledge should eme ge
du ing p ocessing o da a
Po olio
ac i i ies and
p esen
implica ions
The mos aluable echniques o
idea ion is chosen (in ocus), bu
sui abili y and e ec i eness ha e no
been p o en
C-Box is ully implemen ed - and
showed e ec despi e i being
implemen ed as an unsuppo ed
p ocess based on he eam
membe s' knowledge
SIP is ully implemen ed and
showed e ec despi e i being
implemen ed as an unsuppo ed
p ocess based on he eam
membe s' knowledge
SIP ully implemen ed a concep
le el based on C-Box a each
elemen , howe e no agg ega ed
a his as a whole o he sum o
elemen s
C-Box and SIP c ea e ans-
pa ency and ha e a s imu-
la ing e ec , and should be
subs an ia ed wi h p oposals
based eg. on SHAP alues
S ep o wa d o
accommoda e
he implica ions
Look up unc ion as an o e iew o
o he equi alen o simila inpu s
while c ea ing and no as a possible
choice a e wa d
An au oma ic lis o inpu and
ideas ha migh a ec he de el-
opmen o he ideas as well as
po en ial simila c ea ed ideas
A eas o app oaches suppo ing
he domain in ol ing ex e nal
sou ces e en no immedia ely
compa able a eas whe e he
p
inci
p
le is used wi h success
Design deals wi h e ical
p ocessing, whe e he same
aspec s as desc ibed unde Inpu -
concep a e used on each o he
e ical elemen s
Each indi idual eco d mus
NOT be p ocessed in isola-
ion, bu he en i e da a se
should be included du ing
p
ocessin
g
Possible a eas o
in e connec ion
o he p epa a-
ion wi h ocus
Inpu is he esul o used c ea i e
echniques de ined in ocus. By
egis a ion and measu emen o
yield o each used echnique, he
e iciency can by op imized
By de ining s uc u e in ocus
de ining wha is a wha o
challenge, C-Box could suppo an
indica ion o he adicali y o he
idea
By de ining s uc u e in ocus
de ining wha o challenge o
whom, SIP could suppo an
indica ion o he adicali y o he
concep
Ve ical p ocessing is o he
p esen manuel implemen ed in
Rose a, bu his mus co e-
spond o he inno a ion ques- ion
amed in he ocus ac i i y
Ini ial ocusing is usually ame
-
de ining and o en mo e ime-
consuming han idea ion - bu
s ill only limi ed explo ed in
Rose a
P ep ocessing Assess he da a o ma . P epa e and
clean he da a. Handle ou lie s and
missing alues.
Sugges a da ase wi h inpu s
De ine/se c i e ias o an idea.
Sugges a da ase wi h ideas.
Suppo da ase wi h inpu and
ideas
De ine/se c i e ias o a concep
wi h knowledge and p inciple.
Se up da a o m o p inciple.
Sugges a da ase wi h concep s
De ine/se c i e ias o sll a eas
in he e ical chains by impli-
men a ion o he no ion om
inpu and ideas in all s eps
The p epa a ion o domain
ep esen a ion in da a and
o e coming cons ain s in he
in o ma ion p ocess
Fea u e
ex ac ion
Explo e he a ailable da a by, e.g.,
gene al ules and s a is ics.
Po en ially expand he da ase om
ea u e enginee ing echniques
Explo e he idea de elopmen
om he inpu s by, e.g., gene al
ules and s a is ics o he ea u es
esul ing in he ideas
Explo e he concep de elopmen
om he ideas by, e.g., p o iding
insigh on he impo ance o each
idea o concep
Explo e he in e connec edness
be ween he c ea ed ideas in he
9 ield o he e ical chain o
c ea e di e en scena ious o
each concep
The inding and usage o
pa e ns be ween ea u es and
a ge s and o e coming
cons ain s in he in o ma ion
p ocess
Model
cons uc ion
Pu sue blindspo ing by, e.g.,
pe o ming sugges ions abou da a
ha may clus e oge he
Pu sue blindspo ing by, e.g.,
pe o ming sugges ions abou inpu
and ideas ha may clus e
oge he . Sugges o p edic
p obabili ies o , e.g., eaching
o iginali y. Pu sue explainabil iy o
gain insigh on he easoning
behind he sugges ions in a model
Pu sue blindspo ing by, e.g.,
pe o ming sugges ions ega ding
a eas whe e he p inciple is sol ed
o handeled. Sugges o p edic
p obabili ies o , e.g., eaching
adicali y. Pu sue explainabil iy o
gain insigh on he easoning
behind he sugges ions in a model
Pu sue blindspo ing by, e.g.,
pe o ming sugges ions in con-
nec ion wi h he e ical chain
o each concep including
p edic p obabili ies o eached
iginali y and adicali y. Pu sue
explainabil iy o gain insigh on
he easoning behind he
sugges ions in a model
The usage o AI models in
di e en sequen ial s eps in he
idea managemen sys em and
o e coming cons ain s in he
in o ma ion p ocess
Po olio So in o ma ion and u ilize pa e ns
in da a o inspi a ion om, e.g.,
blindspo ing
Enable da a-d i en guided
sugges ions o , e.g., assesing
o iginali y o he ideas by aining
an AI model o C-Box
Enable da a-d i en guided
sugges ions by, e.g., aining an AI
model o SIP
De ine he o e all achi ed
adicalli y when measu ing he
p e e ed and pe asi e
adicali y in he majo i y o
ac i i ies ind he e ical chain
A backbone in a s a egic and
sequen ial idea managemen
sys em using AI ha may
enhance he pon en ial o
eaching adical inno a ion
Rose a as an abso p i e
Capaci y me hod o The in e play be ween AI models and Rose aImplica ion om he analysis including he sugges e
d
in e en ions o mee he ba ie s ound
The impac o Rose a as an Idea Mana-
gemen Sys em ha excels especially in
S ep in Rose a p ejec phase
imple ing adical inno a ion p ocessescon e gence c ea ion
Sou ce: Own line-up
Page 18 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
(2) om Findings 6, i can be seen ha adding knowledge and inspi a ion o he p inci-
ple in he concep has signi ican impo ance o pe o mance. Finally (3) i can be seen
om Findings 7 ha seeking and subs an ia ing he e ec o wo k conce ning adicali y
is impo an o he p ocess ega ding pe o ming mo e adicali y (Jakobsen e al., 2008).
These h ee elemen s a e cu en ly manually en e ed based on he exis ing knowledge
(o dina y in elligence) among he eam membe s in o Rose a wi hou he use o sys-
ema iza ion and au oma ed unc ions in he p og am—o else he analysis shows he
e ec . I indica es g ea po en ial o au oma ion and sys ema iza ion o such unc ions.
Table3 illus a es he exempli ica ion o conside a ions o each s ep o p e ious eali-
za ions and he basis o Rose a uni y wi h oppo uni ies using AI o he ou sequences
s ep o he Rose a P e-jec phase, being he inpu -, idea-, concep - and design s ep—bu
also wi h co ela ion o he la e al ocus phase and P o-jec phase o adap he en i e
p ocess (Mikelsone e al., 2022a). The summa y desc ibed in Table3 includes he dis-
cussion o each a ea o he p esen ed Rose e so wa e sys em as an Ide Managemen
sys em, as an abso p i e capaci y me hod wi h implica ions ela ed o he special s uc-
u e, and inally he in o ma ion s eam ela ed o ac ions bo h in he p esen ed and as
conside a ions o a u u e e sion.
I is well-known ha AI ad an ages a e in ol ed in a ious ways in exis ing and u u e
e sions o abso p i e capaci y sys ems o p o ide be e decision suppo , limi unce -
ain y, and a ge i s e o s. The unique ea u e o Rose a, wi h he ex ension o he
p e ious de elopmen o he sys em, is no o limi o minimize unce ain y o c ea e
a basis o decision (O’Conno & Rice, 2013)— a he he opposi e, whe e i is p oposed
o es ablish mo e unknown ac o s and mo e signi ican unce ain y o c ea e a basis
o he associa ion- and blending heo y ha is so c ucial o seeking and wo king wi h
adical inno a ion (Jakobsen & Ge sen, 2022). This is based on he basic assump ion o
wo king wi h a adical inno a ion ha mo e signi ican unce ain y in he p ocess being
ea ed esul s in less unce ain y in he inal esul , o en desc ibed based on hese p o -
e bs, such as “make mis akes, bu do i quickly”, “i ’s no jus abou doing i igh bu
doing he igh hing igh ”, e c.
As a sys ema ic s uc u e, his is a anged h ough an ac ion-based s uc u e. In ha
sense, Rose a p esen s a new amewo k o handling p og ess wi hou decision bu by
implemen ing quali ied handling (ac ing) and achie ing con e gence. Ac ion con e -
gence and decision con e gence a e ela ed concep s bu ocusing on di e en aspec s
o he p ocess o eaching ag eemen . Ac ion con e gence ocuses on he p ocess o
aligning ac ions owa ds and common goals, while decision con e gence ocuses on he
p ocess o eaching a sha ed decision o ag eemen . Theo e ically, ac ion con e gence
is ela ed o play heo y, and decision con e gence is ela ed o game heo y. O e all,
bo h heo ies con ibu e o a use ul amewo k o unde s anding and analyzing social
in e ac ions and decision-making p ocesses. The e o e, he in o ma ion o p o ide
in o ma ion has ano he s uc u e and ano he pu pose han o en seen in he abso p-
i e capaci y as he goal is no , as usual, o con e ge owa ds a decision o owa ds a
solu ion, bu a he o con ibu e wi h in o ma ion ha can lead o inspi a ion, associa-
ion and blending possibili ies wi hin a eas, whe e he amewo k is s ill o en de ined
qui e b oadly (Funnell & Roge s, 2011). When s uc u ing you da a and in o ma ion
app oach, he challenge he e is o ensu e ha he e is no o e low o in o ma ion and
Page 19 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
oppo uni ies despi e he p esence o e en subs an ial amoun s o da a. This is a clas-
sic si ua ion ha is o en sol ed by p o iding me hods and ou ines ha can b ing he
amoun o da a down o a manageable le el. Wi h he cu en sys em (Rose a), he
opposi e is sough , e en wi h supplemen a ion wi h mo e da a and a gumen s o be -
e implemen ac ions. The a is no o limi he Rose a algo i hm, bu o p o ide and
s uc u e da a in such a way ha i appea s ele an o he indi idual ia he communi-
ca ion hey ha e wi h he sys em ( on K ogh, 2018).
As an abso p i e capaci y ocusing on he inno a ion capaci y ealized in he Idea
Managemen Sys em Rose a p o ides wo aspec s:
1: A sys em o ga he ing and p ocess new in o ma ion, being able o unde s and, and
be able o make use o in o ma ion and knowledge pa ly in e nally ( om egis e ed)
and pa ly ex e nally ( om o he sou ces).
2: A sys em assimila ing powe o ensu e ha e ie ed knowledge bo h ia sou ce
(da a) and p ocess (addi ions) con ibu es in an o ganized way o make i possible o
abso b oppo uni ies speci ically app op ia ed o he indi idual (lea ning sys em) and
wi hou an o e low o oppo uni ies.
Implica ions
In i s cu en o m, Rose a may appea as a ela i ely simple egis a ion sys em o
en e ing, adding images and desc ip ions, ela ing inpu o ideas, seeing possibili ies o
concep s, e c. Howe e , in his s udy, we demons a e ha Rose a p o ides a s uc u e
o managing he in o ma ion and, as a esul , p o ides a s a egic ool o inc easing he
abso p i e capaci y o an enhanced chance o eaching adical inno a ion. Mo eo e ,
we b ie ly discuss he p ospec s o implemen ing AI in he nex e sion o Rose a as a
way o o e come he challenges he analysis indica es wi h exis ing Rose a as well as
explo e he manu ac u ed oppo uni ies—and how explainabili y may be suppo ed o
enhance he easoning om he model, hence explainable AI (XAI). Rose a ea s h ee
cen al a eas insc ibed as elemen s (bu no ye un olded) wi h he possibili ies ha lie in
he use o AI:
– Explo e he digi al a ailabili y o he in o ma ion ha may enhance abso p i e capac-
i y.
– To ake ac ion by classi ying adicali ies om C-box and SIP, u ilize a ailable da a o
each speci ic inpu , idea, concep , and design.
– De elop a amewo k ha suppo s he explainabili y o he gi en classi ica ion o
adicali y.
De eloping an AI model is adi ionally an i e a i e p ocess be ween p ep ocess-
ing, ea u e selec ion, and model cons uc ion (Duda e al., 2001), which also applies o
explo ing an AI model o Rose a (see Table3). A es o , e.g., pe o mance may ol-
low hese s eps (Duda e al., 2001). P ep ocessing e e s o deciding on pu pose and p e-
pa ing he da a, e.g., da a o ma s (s uc u ed o uns uc u ed), handling ou lie s, he
impac o a missing alue, and conside ing he sample size ( o a oid o e i ing). Fea-
u e selec ion e e s o deciding on ele an in o ma ion, which can be done manually
(as in Rose a) o wi h he machine’s assis ance ha ela es o classical s a is ics. Da a
Page 20 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
disc e iza ion may also be conside ed o , e.g., simpli ying a model. The da a se is hen
spli in o a aining- and es se (e.g., using 90% o he da a o aining), each con ain-
ing he ea u es o in e es (Duda e al., 2001). Model cons uc ion e e s o bo h model
selec ion and - aining o eaching he decision o ac ion based on knowledge abou he
ea u es, e.g., calcula ing he p obabili y o being a adical idea i knowing ‘Pa en ? = No’,
‘Is_addi ion_ o_exis ing_concep ? = Yes’ and ‘Seen_elsewhe e = No’, e c. The model is only
ained on ou aining da a o lea ning he pa e ns in da a (Duda e al., 2001). When
ained, and depending on a gi en decision h eshold, he AI model now allows a classi-
ica ion based on knowledge p o ided by he ea u es, such as sugges ing ’ he p obabili y
o he idea being adical is 10%. F om a decision h eshold o 50%, he idea is no classi‑
ied as adical’.
While ou sugges ed model uses a ailable in o ma ion o inc ease he abso p i e
capaci y om di e en ea u es o assess adicali y, we may need o assess how well
his new model can be us ed in e ms o classi ying co ec ly. We, he e o e, seek o
alida e he model in, e.g., c oss- alida ion, which p o ides us wi h an es ima e o he
gene alized e o he model pe o ms (Duda and e al., 2001). I a model pe o ms pe -
ec ly, we may need o conside he possibili y o ha ing an o e i ed model—po en ially
om a small sample size. On he o he hand, i he model pe o ms poo ly, we may need
o e isi he di e en s eps o de eloping a model, e.g., e-assessing how we ep esen
he in o ma ion in he da a se , uning a model di e en ly, o deciding on using ano he
model ype. Conclusi ely, we use he es se o simula e he eal wo ld and assess how
well he model may assis ou pu pose. In his con ex , popula me hods o XAI, e.g.,
he SHapley Addi i e exPlana ions (SHAP) (Lundbe g & Lee, 2017) wa e all o o ce
plo s, sha e, o ou knowledge, o e seen undamen al p inciples wi h exis ing s a egic
managemen sys ems, such as o ce ields. The SHAP me hod enables an assessmen o
ea u e con ibu ion o he classi ica ion o adicali y, such as sugges ing ha ‘Seen_else‑
whe e = No’ inc eases he p obabili y o being adical, whe e o ‘Pa en ? = No’ and ‘Is_
addi ion_ o_exis ing_concep ? = Yes’ dec eases he p obabili y o being adical, gi en an
AI model. In all, SHAP suppo s explainabili y by isualizing he ea u e impo ance o
eaching a 10% chance o being adical in he AI model, which we seek o u ilize o
Rose a. Table3 illus a es sequences o he inno a ion p ocess using Rose a and how
he i e a i e s eps o de eloping an AI model may o e lap wi h di e en phases.
Conclusion
A he beginning o he s udy, we s aged he ollowing esea ch ques ion: Wha can we
lea n om analyzing da a om 11 adical inno a ion po olios, and wha a e he impli‑
ca ions o inno a ion managemen esea ch om hese indings? Ou explo a ion and
un a eling o hese da a ha e led o ollowing insigh s and indings:
– In abso p i e capaci y, he use o an Idea Managemen Sys em is mos o en seen
as he abili y o a ge and manage in e nally and ex e nally p o ided knowledge
h ough ans o ma i e lea ning (Volbe a e al., 2010; B ix, 2019). Mos o en, his
deals wi h lea ning p ocesses o decision o decision suppo ac i i ies. In con as ,
Rose a p esen s a model wi hou he possibili y o c ea ing a decision, wi hou he
possibili y o me ging ideas, and wi hou he possibili y o any econcilia ion o ideas’
Page 21 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
alue, sui abili y, ele ance, e c. Despi e his app oach o c ea ing inspi a ion and
knowledge o he pu sui o an oppo uni y, he analysis shows ha he numbe
o possibili ies o he s ep-by-s ep p ocess o he c ea ion o adicali y, despi e he
omission o abso p i e capaci y key a eas ega ding decision suppo , con e ge. This
con e gence is a su p ise and unexpec ed (Jakobsen, 2021).
– The e is a ce ain co ela ion be ween he numbe o inpu s c ea ed, and he adical-
i y achie ed, bu p ima ily when wo king wi h mo e han 300 inpu s. On he o he
hand, he analysis shows a signi ican e ec on adicali y when mo e inpu is con-
nec ed o an idea as inspi a ion.
– When wo king on c ea ing mo e adicali y, he e is a s ong indica ion in he p esen
analysis ha he use o inspi a ion ac i i y is ex emely e ec i e.
– Ou analysis indica es ha eam sizes o 3–5/6 migh pe o m bes , while he adical-
i y can dec ease wi h la ge g oup sizes.
– To inc ease he anspa ency and in e p e abili y in Rose a po olio models is
implemen ed. This use o a po olio model show ha he de ined de ini ion ma k-
edly in luences he wo k wi h ideas and concep in he po olio models. Consciously
wo king owa ds he c ea ion o adical inno a ion wi h he mos signi ican possible
po en ial in luences he p ocess o complying wi h his wish.
– Rose a con ibu es o key elemen s as a mo e ecen con ibu ion o exploi ing he
po en ial in he a ea o adap i e capaci y and inno a i e capaci y, whe e unce ain y
is p ocessed wi h he suppo o inspi a ion, new knowledge, o indica ion o po en-
ial oppo uni ies as opposed o emo ing challenges o educe unce ain y.
The exploi a ion o Rose a shows solid indica ions o he implemen a ion o inspi a ion
sou ces ela ed o he c ea ed (in he i s s eps) and ex e nal knowledge o sou ces o inspi-
a ion based on mapping he p inciple o he concep . He e, we see a bene icial use o AI o
handling a ela i ely la ge amoun o da a and he many ela ionships, o ca ego izing he
ield o possibili ies wi hin associa ion possibili ies, blending possibili ies, and he supply o
new knowledge o s imula e ac ion (Fauconnie e al., 2008; Jakobsen & Ge sen, 2022). This
has i s jus i ica ion in he ‘law on a ie y’ conce ning he sen ence ha “only a ie y can
des oy a ie y”. Radical inno a ion con ains in i s de ini ion a signi ican deg ee o unce -
ain y due o many unknown ac o s in i s o m, which is why decisions wi hou knowledge
do no make sense (sabo ages adical inno a ion) bu only allow ealiza ion by emo ing
a ie y by adding new knowledge, new inspi a ion, expe imen ing, and ying o ge ahead.
This is a comple ely di e en discipline ha Rose a p ima ily aims o deal wi h in wo ways:
– As inspi a ion ac i i y in each s ep, con inuously p ocessed p oposals a e p o ided o
simila en ies, inspi ing en ies, challenging en ies, and o he a eas, whe e his has
been sol ed and whe e new inspi a ion o knowledge can be supplemen ed. Ne e -
heless, he ype and o m should be di e en in he di e en s eps as bo h he unce -
ain y and needs a e di e en in he di e en s eps, s a ing wi h c ea ing an o e -
iew (pa e n) o he c ea ed ollowed by mo e and mo e implemen a ion o ex e nal
knowledge, e.g., om pa en s, simila a eas and a eas p ocessing he same p inciple.
– Suppo ed indica ion o adicali y (C-box and SIP) wi h he indica ion o likely place-
men in he po olio model based on he da a desc ibed in he indi idual eco d
Page 22 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
combina ion da a om he ini ial p oblem/challenge ea men (in he CIS model
called La e al ocus). This has no ye been es ed o implemen ed in Rose a bu will
be s ong, as he analysis has shown ha his is pa ly a o able o deepe p ocessing
o he indi idual eco d, and pa s can e en be o g ea impo ance o de e mining
he di ec ion o he na u ally occu ing con e gence.
Abb e ia ion
AI A i icial In elligence
Acknowledgemen s
Thanks mus go o he Danish Technological Ins i u e, including speci ically o me cen e manage Jø gen Høegh, who
o e a numbe o yea s s ongly in es ed in adical inno a ion as a se ice o o e clien o ganiza ions, including he
de elopmen o ools o suppo adical inno a ion h ough, among o he hings, so wa e based idea managemen
sys em. Wi hou his suppo , de eloping he idea managemen sys em ‘Rose a’ and he di e en e sions o i would
no ha e been possible. Finally, hanks mus go o Associa e P o esso Rune H. Jacobsen a Aa hus Uni e si y, who has
con ibu ed wi h aluable spa ing, especially in ela ion o he un olding o possibili ies and limi a ions when using AI.
Au ho con ibu ions
HSJ is he lead au ho and is esponsible o ge ing da a access. HSJ made he i s d a o he me hodology, da a analy-
sis, and p esen a ion o esul s. RSJ is esponsible o he discussion and AI in eg a ion and made he i s d a o his.
JB is esponsible o he i s d a o he in oduc ion and he i s d a o he heo e ical backg ound. All au ho s ha e
con ibu ed equally o he e ision o he manusc ip .
Funding
No applicable.
A ailabili y o da a and ma e ials
The da a se analyzed du ing he cu en s udy is no publicly a ailable due o con iden iali y ag eemen s made wi h he
companies we collabo a ed wi h.
Decla a ions
Compe ing in e es s
The au ho s decla e ha hey ha e no compe ing in e es s.
Recei ed: 9 Janua y 2023 Accep ed: 11 Janua y 2024
Re e ences
Al shulle , Gen ich (1999). The inno a ion algo i hm: TRIZ, sys ema ic inno a ion and echnical c ea i i y.
Amabile, T. M. (2020). C ea i i y, a i icial in elligence, and a wo ld o su p ises. Academy o Managemen Disco e ies, 6(3),
351–354. h ps:// doi. o g/ 10. 5465/ amd. 2019. 0075
Be e a, M. (2019). Idea selec ion in web-enabled idea ion sys ems. Jou nal o P oduc Inno a ion Managemen , 36(1),
5–23. h ps:// doi. o g/ 10. 1111/ jpim. 12439
B asil, V. C., Sale no, M. S., Egge s, J. P., & Gomes, L. A. D. V. (2021). Boos ing Radical Inno a ion Using Ambidex ous Po o-
lio Managemen : To manage adical inno a ion e ec i ely, companies can build ambidex ous po olio manage-
men sys ems and adop a mul ile el o ganiza ional app oach. Resea ch-Technology Managemen , 64(5), 39–49.
h ps:// doi. o g/ 10. 1080/ 08956 308. 2021. 19476 05
B ix, J. (2019). Ambidex e i y and o ganiza ional lea ning: e isi ing and econnec ing he li e a u es. The Lea ning
O ganiza ion, 26(4), 337-351. h ps://doi.o g/10.1108/TLO-02-2019-0034
B ix, J., & Ho sage , B. (2022). Could we manage ou s a egic inno a ion p ojec s be e ?: Un a eling unexpec ed wai ing imes
as a key p ojec unce ain y. P oceedings om he R&D Managemen Con e ence, June 17-21, Se illa, Spain.
B ix, J., & Jakobsen, H. S. (2013). Co po a e c ea i i y: In oducing he c ea i e idea solu ion amewo k. In e na ional
Jou nal o Inno a ion and Lea ning, 13(4), 388–405. h ps:// doi. o g/ 10. 1504/ IJIL. 2013. 054235
B ix, J., & Jakobsen, H. S. (2014). Business model p e o yping: Explo ing p e-comme cialisa ion oppo uni ies in p ac ice.
In e na ional Jou nal o Inno a ion and Lea ning, 17(1), 98–110. h ps:// doi. o g/ 10. 1504/ IJIL. 2015. 066066
Chesb ough, H. W. (2006). Open inno a ion: The new impe a i e o c ea ing and p o i ing om echnology. Ha a d Business
P ess.
Chil on, Michael A.; Bloodgood, James M. (2007). The Dimensions o Taci & Explici Knowledge: A Desc ip ion and Meas-
u e. 40 h Annual Hawaii In e na ional Con e ence on Sys em Sciences (HICSS’07), 2007, pp. 188a-188a, h ps:// doi. o g/
10. 1109/ HICSS. 2007. 524.
Page 23 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
Ch is ensen, C. M. (1997). The inno a o ’s dilemma: when new echnologies cause g ea i ms o ail. Ha a d Business Re iew
P ess.
Cohen and Le in hal. (1990). Abso p i e capaci y: A new pe spec i e on lea ning and inno a ion. Adminis a i e Science
Qua e ly, 35(1), 128–152. h ps:// doi. o g/ 10. 2307/ 23935 53
Dahlande , L., & Gann, D. M. (2010). How open is inno a ion? Resea ch Policy, 39(6), 699–709. h ps:// doi. o g/ 10. 1016/j.
espol. 2010. 01. 013
De Vaus, D., & de Vaus, D. (2013). Su eys in social esea ch. Rou ledge.
Duan, Y., Wang, W., & Zhou, W. (2020). The mul iple media ion e ec o abso p i e capaci y on he o ganiza ional slack
and inno a ion pe o mance o high- ech manu ac u ing i ms: E idence om Chinese i ms. In e na ional Jou nal o
P oduc ion Economics, 229, 107754. h ps:// doi. o g/ 10. 1016/j. ijpe. 2020. 107754
Duda, R. O., Ha , P. E., & S o k, D. G. (2001). Pa e n Classi ica ion. John Wiley & Sons.
Fab izio, K. R. (2009). Abso p i e capaci y and he sea ch o inno a ion. Resea ch Policy, 38(2), 255–267. h ps:// doi. o g/ 10.
1016/j. espol. 2008. 10. 023
Fauconnie , G., & Tu ne , M. (2008). The way we hink: Concep ual blending and he mind’s hidden complexi ies. Basic books.
Fülle , J., Hu e , K., Wahl, J., Bilg am, V., & Tekic, Z. (2022). How AI e olu ionizes inno a ion managemen –pe cep ions
and implemen a ion p e e ences o AI-based inno a o s. Technological Fo ecas ing and Social Change, 178, 121598.
h ps:// doi. o g/ 10. 1016/j. ech o e. 2022. 121598
Funnell, S. C., & Roge s, P. J. (2011). Pu pose ul p og am heo y—e ec i e use o heo ies o change and logic models.
Jossey-Bass.
Goduschei , R. C., & Faullan , R. (2018). Pa hs owa d adical se ice inno a ion in manu ac u ing companies—a se ice-
dominan logic pe spec i e. Jou nal o P oduc Inno a ion Managemen , 35(5), 701–719. h ps:// doi. o g/ 10. 1111/
jpim. 12461
Go ski, C., & Heinekamp, E. J. (2004). Cap u ing employee ideas o new p oduc s. The PDMA ToolBook 1 o New P oduc
De elopmen , 219.
Hae ne , N., Wincen , J., Pa ida, V., & Gassmann, O. (2021). A i icial in elligence and inno a ion managemen : A e iew,
amewo k, and esea ch agenda✰. Technological Fo ecas ing and Social Change, 162, 120392. h ps:// doi. o g/ 10.
1016/j. ech o e. 2020. 120392
Jakobsen, H.S. (2021) PHD hesis: The CREATIVE IDEA SOLUTION (CIS) P ocess and F amewo k: An inno a ion me hod
wi h a adical and dis up i e app oach, excogi a ed and igo ously de eloped by empi ical explo a ion o i s usage.
Aa hus Uni e si y, Denma k.
Jakobsen, H.S.; Ge sen, F.; Goduschei s, R.C.; Rebsdo . S.O. (2008). Impac o measu emen and e alua ion on inno a ion
le el, 2008; 8 h In e na ional CINe Con e ence, Con inuous Inno a ion - Oppo uni ies and Challenges, Valencia,
Spain.
Jakobsen, Henning Seje ; Ge sen, F ank (2022). Compe ence-based en ep eneu ship based on blending inno a ion
heo y. LUT Scien i ic and Expe ise Publica ions, 978-952-335-694-8
Jakobsen, Henning Seje ; Hansen, Sø en (2007). Expe imen al C ea i i y; CPSI Con e ence, Bu alo, USA.
Jakobsen, Henning Seje ; Rebsdo , Simon Olling (2003). Ideud ikling ed k ea i Inno a ion. Gyldendal, Københa n.
Keding, C. (2021). Unde s anding he in e play o a i icial in elligence and s a egic managemen : Fou decades o
esea ch in e iew. Managemen Re iew Qua e ly, 71(1), 91–134. h ps:// doi. o g/ 10. 1007/ s11301- 020- 00181-x
Keeley, La y; Pikkel, Ryan; Wal e s, Helen; Quinn, B ian (2013). Ten Types o Inno a ion: The Discipline o Building
B eak h oughs.
Knoppen, D., Sa is, W., & Moncaga a, P. (2022). Abso p i e capaci y dimensions and he measu emen o cumula i eness.
Jou nal o Business Resea ch, 139, 312–324. h ps:// doi. o g/ 10. 1016/j. jbus es. 2021. 09. 065
K is iansen, J. N., & Ri ala, P. (2018). Measu ing adical inno a ion p ojec success: Typical me ics don’ wo k. Jou nal o
Business S a egy, 39(4), 34–41. h ps:// doi. o g/ 10. 1108/ JBS- 09- 2017- 0137
Lassen, A. H., Ge sen, F., & Riis, J. O. (2006). The nexus o co po a e en ep eneu ship and adical inno a ion. C ea i i y and
Inno a ion Managemen , 15(4), 359–372. h ps:// doi. o g/ 10. 1111/j. 1467- 8691. 2006. 00406.x
Lei e , R., McDe mo , C.M., O’Conno , G.C., Pe e s, L.S., Rice, M., & Ve yze , R.W. (2000). Radical inno a ion: How ma u e
companies can ou sma ups a s. Ha a d Business P ess.
Lewin, K. (1943). De ining he ield a a gi en ime. Psychological Re iew., 50(3), 292–310. h ps:// doi. o g/ 10. 1037/ h0062
738
Lundbe g, S. M., & Lee, S. I. (2017). A uni ied app oach o in e p e ing model p edic ions. Ad ances in neu al in o ma ion
p ocessing sys ems, 30. h ps:// p oce edings. neu i ps. cc/ pape / 2017/ ile/ 8a20a 86219 78632 d76c4 3d d2 8b677 67-
Pape . pd
Mikelsone, E., Spilbe gs, A., Volko a, T., & Liela, E. (2022a). Idea managemen sys ems in de eloping inno a ion capaci y.
In e na ional Jou nal o Inno a ion and Technology Managemen , 19(3), 2240001. h ps:// doi. o g/ 10. 1142/ S0219 87702
24000 16
Mikelsone, E., U a o a, I., & Sege s, J. P. (2022b). Fou -s ep app oach o idea managemen sequencing: Rede ining o
ein en ing alues in a business model. Jou nal o Inno a ion and En ep eneu ship, 11(1), 1–20. h ps:// doi. o g/ 10.
1186/ s13731- 022- 00236-1
Nei o i, P., Pesce, D., & Ba aglia, D. (2021). Algo i hms o ope a ional decision-making: An abso p i e capaci y pe spec-
i e on he p ocess o con e ing da a in o ele an knowledge. Technological Fo ecas ing and Social Change, 173,
121088. h ps:// doi. o g/ 10. 1016/j. ech o e. 2021. 121088
Nilsson, L., Elg, M., & Be gman, B. (2002). Managing ideas o he de elopmen o new p oduc s. In e na ional Jou nal o
Technology Managemen , 24(5–6), 498–513. h ps:// doi. o g/ 10. 1504/ IJTM. 2002. 003067
Oakley, T., & Pascual, E. (2017). Concep ual blending heo y. Camb idge Uni e si y P ess. h ps:// doi. o g/ 10. 1017/ 97813
16339 732. 027
O’Conno , G. C. (1998). Ma ke lea ning and adical inno a ion: A c oss case compa ison o eigh adical inno a ion p o-
jec s. Jou nal o P oduc Inno a ion Managemen , 15(2), 151–166. h ps:// doi. o g/ 10. 1111/ 1540- 5885. 15201 51
O’Conno , G. C., Lei e , R., Paulson, A., & Pe e s, L.S. (2008). G abbing ligh ning: Building a capabili y o b eak h ough
inno a ion. John Wiley & Sons.
Page 24 o 24
Jakobsene al. Jou nal o Inno a ion and En ep eneu ship (2024) 13:9
O’Conno , G. C., Co be , A. C., & Pe e s, L. S. (2018). Beyond he champion: Ins i u ionalizing inno a ion h ough people.
S an o d Uni e si y P ess.
O’Conno , G. C., & Rice, M. P. (2013). A comp ehensi e model o unce ain y associa ed wi h adical inno a ion. Jou nal o
P oduc Inno a ion Managemen , 30, 2–18. h ps:// doi. o g/ 10. 1111/ jpim. 12060
Sale no, M. S., de Vasconcelos Gomes, L. A., Da Sil a, D. O., Bagno, R. B., & F ei as, S. L. T. U. (2015). Inno a ion p ocesses:
Which p ocess o which p ojec ? Techno a ion, 35, 59–70. h ps:// doi. o g/ 10. 1016/j. echn o a i on. 2014. 07. 012
Scha me , O o C. (2007), Add essing The Blind Spo o Ou Time.
U e back, J. M. (1994). Radical inno a ion and co po a e egene a ion. Resea ch Technology Managemen , 37(4), 10.
Volbe da, H. W., Foss, N. J., & Lyles, M. A. (2010). Pe spec i e—Abso bing he concep o abso p i e capaci y: How o
ealize i s po en ial in he o ganiza ion ield. O ganiza ion Science, 21(4), 931–951. h ps:// doi. o g/ 10. 1287/ o sc. 1090.
0503
on K ogh, G. (2018). A i icial in elligence in o ganiza ions: New oppo uni ies o phenomenon-based heo izing. Acad-
emy o Managemen Disco e ies, 4(4), 404–409. h ps:// doi. o g/ 10. 5465/ amd. 2018. 0084
Yin, R. K. (2013). Case s udy esea ch: Design and me hods. Sage Publishe s.
Zah a, S. A., & Geo ge, G. (2002). Abso p i e capaci y: A e iew, econcep ualiza ion, and ex ension. Academy o Manage-
men Re iew, 27(2), 185–203. h ps:// doi. o g/ 10. 5465/ AMR. 2002. 65879 95
Zobel, A. K. (2017). Bene i ing om open inno a ion: A mul idimensional model o abso p i e capaci y. Jou nal o P oduc
Inno a ion Managemen , 34(3), 269–288. h ps:// doi. o g/ 10. 1111/ jpim. 12361
Zou, T., E ug, G., & Geo ge, G. (2018). The capaci y o inno a e: A me a-analysis o abso p i e capaci y. Inno a ion, 20(2),
87–121. h ps:// doi. o g/ 10. 1080/ 14479 338. 2018. 14281 05
Publishe ’s No e
Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions.
