Dual Capability EAM for Agility in Business Capability Building : A Systems Theoretical View

This is a sel -a chi ed e sion o an o iginal a icle. This e sion may di e om he o iginal in pagina ion and ypog aphic de ails. Au ho (s): Ti le: Yea : Ve sion: Copy igh : Righ s: Righ s u l: Please ci e he o iginal e sion: CC BY-NC-ND 4.0 h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ Dual Capabili y EAM o Agili y in Business Capabili y Building : A Sys ems Theo e ical View © 2021 SCITEPRESS Accep ed e sion (Final d a ) Pou anen, Jouko; Pulkkinen, Mi ja Pou anen, J., & Pulkkinen, M. (2021). Dual Capabili y EAM o Agili y in Business Capabili y Building : A Sys ems Theo e ical View. In J. Filipe, M. Smialek, A. B odsky, & S. Hammoudi (Eds.), ICEIS 2021 : P oceedings o he 23 d In e na ional Con e ence on En e p ise In o ma ion Sys ems. Volume 2 (pp. 726-734). SCITEPRESS Science And Technology Publica ions. h ps://doi.o g/10.5220/0010455007260734 2021 Dual capabili y EAM o agili y in business capabili y building: A sys ems heo e ical iew Jouko Pou anen, Mi ja Pulkkinen a Facul y o In o ma ion Technology, Uni e si y o Jy askyla, Finland jouko. .pou anen@s uden .jyu. i, mi ja.pulkkinen@jyu. i Keywo ds: En e p ise A chi ec u e Managemen ; Complex Adap i e Sys ems; Business Agili y, dual capabili y EAM Abs ac : Th ough wo cases o IT-enabled business capabili y building in la ge en e p ises, his pape elucida es how he sys ems heo y app oach can explain he en e p ise a chi ec u e managemen (EAM) challenge o suppo business agili y. The obse a ion in bo h o hese cases is ha a legacy EAM app oach does no adap o a business de elopmen scena io in ol ing agili y. This leads o a s udy o he na u e o he challenges in EAM when enabling s a egic business mo es in ol ing new echnologies, a he business uni le el. Fo he ype o p ojec s as in hese cases, we do no ind a i ing pa adigm in he EAM li e a u e. Sugges ed solu ions a e IT bimodali y, o Two Speed IT. Howe e , i s combina ion wi h EAM is sca ce in ea lie esea ch. To be able o p o ide guiding ideas o he u he de elopmen o a dual capabili y EAM app oach, wi h an e iden need, we de elop a sys ems heo e ical s a ing poin o examine he cases. Complex Adap i e Sys em (CAS) cha ac e is ics appea o gi e he necessa y explana ions o build on. Suppo ed by his heo e ical de elopmen , he s udy esul s in p inciples o a dual capabili y EAM, o agile s a egic business capabili y building in ol ing en e p ise e-s uc u ing. 1. INTRODUCTION As en e p ises see dazzling business oppo uni ies wi h new echnologies, hey ace go e nance challenges in he a eas o e.g. isk and secu i y managemen . IT Go e nance as he b oadly accep ed, alue based app oach (Op’ Land e al. 2008), wi h En e p ise A chi ec u e Managemen (EAM) as a ool, poin s o he need o conside he alue o any IT in es men o he en e p ise, as well as di ec and moni o he planning and implemen a ion o he induced business change. This equi es a balanced de elopmen a he co po a e le el, as opposed o indi idual business uni s de eloping hei speci ic solu ions and unning he isk o pa ial op imiza ion, po en ially e en coun e p oduc i e o he co po a e goals (Ahlemann e al. 2012, Pe e son 2004). The EAM p ocess, suppo ing he goal o ci cumspec decision making, joins he echnological iewpoin o he business goals. I a no el IT solu ion needs in eg a ion o he exis ing a chi ec u es, engaging he co po a e IT unc ion is necessa y, and hei esponsibili y con inues wi h he planning and execu ing o he ope a ion, suppo and main enance a h ps://o cid.o g/0000-0002-5313-8666 o he new sys em. Howe e , he co po a e IT and EAM aiming a he alignmen o business and IT de elopmen s wi h an a chi ec u al app oach (Ahlemann e al. 2012) can be ques ioned as a “legacy” app oach, wi h “bu eauc a ic” go e nance models hampe ing agile de elopmen bo h o he business and o IT (D ews e al. 2017). The e o e, EAM p ac ices migh need a e ision o cases o agile IT-d i en business uni le el de elopmen , enabling he swi seizing o new digi al business oppo uni ies. In hei co e ing e iew o o ganiza ional (business and IT) agili y, Tallon e al. (2019) gi e a ho ough unde s anding o he complexi y o he p oblem. While capabili y building and he dynamic capabili ies domina e as heo e ical s ance, wi h some, he ques ion o a chi ec u al modula i y (Tiwana and Konsynski 2010) and decision igh s Tiwana and Kim (2015) poin o simila obse a ions as a e igge ing ou in e es in his s udy, especially he co-loca ion o knowledge and he decision igh s poin ed ou in Tiwana and Kim (2015). Resea ch has e idenced he IT co po a e unc ions ans o ming o Bi-modal o wo-speed IT (Ha ke e al. 2017; Ho lach e al, 2016). In academia, he e m ambidex e i y has been coined o he dual IT unc ion capabili y (Lee e al. 2015). An IT uni wi h his double app oach se es he business no only wi h agile de elopmen , bu also wi h new cul u e- changing app oaches, as De Ops, u he joined wi h he business de elope s o BizDe Ops (G uhn e al. 2015). The duali y means also ensu ing sus ained suppo o balanced long- e m planning and go e nance wi h isk managemen a a ious le els. A ‘ as IT’, howe e , is a solu ion only whe e he de elopmen s a e conduc ed in-house, wi h he en e p ise IT esou ces. In ou case s udy, we obse e he need o a business agili y, whe e IT solu ions new o bo h he business and he IT in he en e p ises a e a s a egic choice as new business capabili y enable s. Bo h cases, howe e , see a co po a e s a egy equi ing an assu ed and well managed EA as well. As no all o he de elopmen esou ce is coming om in-house, a e-o ganiza ion o he IT unc ion alone is no sol ing ou case p oblem. The mo e pe inen ques ion is, how o enable simila , agile s a egic mo es in u u e, as he p esen EAM app oach, in hese en e p ises and in gene al, appea s no o ackle he si ua ion. The ques ion o his s udy is: RQ How can EAM suppo business capabili y de elopmen in an agile manne , when i in ol es he building o a new sys em and a new uni , changing he en e p ise s uc u e? To unde s and he manage ial challenge, on bo h he business and he IT side, and o p opose a solu ion o hese cases, we look a po en ial heo e ical explana ions. Al eady exis ing solu ions a e sc eened in a li e a u e e iew o En e p ise A chi ec u e Managemen and Bimodal o Two Speed IT. Dealing wi h a complex se ing wi h business sys ems embedded in an en i onmen o an en e p ise as a sys em o sys ems, we p opose a sys ems heo e ic explana ion o he si ua ion a hand in Sec ion 2, and discuss he EAM ole. In sec ion 3, we desc ibe wo cases o new echnology induced business de elopmen , and p esen an analysis in Sec ion 4. We e lec he analysis o EAM in sec ion 5. As a esul , we p opose guidelines o a business agili y enabling, dual capabili y EAM in Sec ion 6. 1.1 Ea lie S udies Fo he li e a u e e iew, conduc ed in ea ly 2019, we i s checked bo h he IEEE Xplo e po al and Google Schola and used he sea ch s ings i)“’Bimodal IT’ AND ‘En e p ise A chi ec u e Managemen ’”, and ii)“’Two Speed IT’ AND ‘En e p ise A chi ec u e Managemen ’”. As IEEEXplo e did no yield esul s, Google Schola a ains b oade co e age o his a he no el a ea. A u he gain is o ind non-published schola ly wo k such as heses. The esul s wi h he wo sea ch s ings espec i ely yielded eigh and nine hi s, and a e dele ing o e laps in he wo se s, plus excluding wo k no w i en in English, he ollowing i ems emain: • Th ee MSc Theses: (Boekhol z 2017; Schmid 2018; Na alucci & Manzo i 2016), he las one wi h a gene ic s ance on digi aliza ion. • A disse a ion (Ande sen 2016), ocusing on he echnology dimension o EA, • Fou schola ly pee - e iewed a icles (D ews e al. 2017; Fo mann e al. 2019; Kelle e al. 2018; Legne e al. 2017). • Fu he h ee a icles, published o ano he ield o esea ch, appea ed o be no in he scope. The sea ch was no limi ed wi h ime o publica ion meaning ha his opic only has eme ged. Th ee empi ical s udies, (D ews e al. 2017; Kelle e al. 2018; Legne e al. 2017) p esen elabo a ions o he p oblem a ea. Howe e , i s o hem, as also he MSc heses, p esen digi al-only business cases in he sec o in he on -line o digi aliza ion: banking and inancial. Ano he pape (Fo mann e al. 2019) s udies he de elopmen s o e ime wi hin an IT uni in a digi aliza ion case, and Kelle e al. (2018) collec gene al p ac i ione knowledge om IT-depa men s. Ou s udy, on i s pa , in es iga es wo cases whe e a digi al ool o a new se ice is only a pa o he business po olio in a non-digi al business, and we go beyond he s udies o he IT depa men /IT unc ion. The echnologies (AI and IoT) a e no el o he en e p ises, bo h p o iling as la ecome digi alise s (Kohli e al. 2011). Fo bo h, he p ojec means new business capabili y de elopmen . A he same ime, due o he de elopmen p ojec , ou cases p esen en e p ises wi h a challenge in he IT go e nance de-/cen aliza ion ques ion (Pe e son 2004), and also de ining he ole o hei EAM eam as pa o he go e nance. The in ol emen o he IT unc ion du ing de elopmen p ojec s low, hus no sugges ing ha a dual capabili y o mul imodal IT unc ion would sol e all o he p oblem. Fu he , he solu ions a e no buil on sys ems al eady ope a ed on, so De Ops, o BizDe Ops al hough in e es ing app oaches, a e no a i ing pa adigm. 2. KEY CONCEPTS 2.1 En e p ises as Sys ems o Sys ems Beyond he p ac ical solu ions, as in oducing agili y o he o ganiza ion o he IT unc ion (e.g. Ha ke e al. 2017), and me hods o agile de elopmen , also ex ended o EA (e.g. D ews e al. 2017), mo e undamen al ques ions o he p oblem ield o EAM ha e been p esen ed (Ab aham e al. 2013; Buckl e al. 2009). To sol e he p oblem o he con o e sial oles o EA o EA managemen : An enable o agili y on one hand, howe e , on he o he , ques ioned as a ‘bu eauc a ic’ and he e o e slow app oach, we s udy he unde lying con o e sy by examining he sys ems na u e o en e p ises. The cases we s udy show he need o an a chi ec u al go e nance o comply wi h he ITG and co po a e go e nance, bu on he o he hand, also he u gency o apidly espond o he business en i onmen change in ol ing echnology. The la e is as compelling om he business side by co po a e go e nance, o sus ain and imp o e he alue c ea ion in he en e p ise. Sys ems heo ies gi e a widely emb aced e en hough no ully u ilized heo e ical scheme, unde lying he EA ield, as a ecen e iew inds (Niemi e al. 2018). Figu e 1 p esen s a simple ske ch o he in e ac ing sys ems in a sys em o sys ems, as an en e p ise can be analy ically seen. We lean on he main idea in he Gene al Sys ems Theo y (GST) and Li ing Sys ems Theo y (LST), bo h p oposing hie a chical le els o sys ems complexi y (Ab aham e al. 2013; Nu mi e al. 2018). Wi hin he en e p ise, comp ising a la ge complex, adap i e sys em (CAS) (Janssen & Kuk 2016), he e a e bo h ully manageable, p edic able echnical sub-sys ems, some mo e complex socio- echnical sub-sys ems, as well as adap i e sub-sys ems (whe e he decision eedom is gi en by he sys em-o -sys ems managemen sys em, con olling he sys ems/sub-sys ems wi hin he en e p ise). A CAS is capable o di ec ing and e- di ec ing hei ac ions and esou ces a hei disposal, acco ding o he signals hey pe cei e om hei en i onmen . The new business capabili y ( ha also can be pe cei ed a new “socio- echnical” sys em), o ms a sub-sub-sys em unde he auspices o a sub- CAS. I s owne ship is no mally wi h a business uni ( he sub-CAS), he whole en e p ise hus seen as a CAS-o -CASs. In a simila ein, Ab aham e al. (2013) ha e in oduced en e p ises as “hie a chical, mul i- le el sys ems”, howe e , lea ing he adap i eness aspec o u he s udy. Capabili y as an “abili y o execu e a de ined and epea able pa e n o ac i i ies” o p oduce a a ge ed ou come in a gi en en i onmen , has been ound a p ac icable uni o analysing he business a chi ec u e as pa o EA (Simon e al. 2014). A business capabili y en ails he espec i e p ocesses, en i ies, o ganiza ions, people, cul u e, and esou ces needed o success ully pe o m an ac i i y o i s a ge ed business ou come. A business capabili y can be dependen on IT (IT-enabled), o imp o ed by IT (IT-enhanced). Figu e 2: En e p ise as a collec ion o mul iple, mul i-le el sys ems a di e en le els o complexi y The concep o capabili y, howe e , highligh s he need o an “ensemble o esou ces” o building a ‘socio- echnical sys em’ om he ‘ echnical sys em’ wi h he business esou ces be o e i can pe o m. Whe e he e a e esou ces, he e is also owne ship, managemen and decision making powe . The decision making le els (Pulkkinen 2006; Pulkkinen & Hi onen 2005) as well as he sys ems heo ies analy ically p esen ing hem (Ab aham e al. 2013); ha e been a subjec o s udy in EAM. We see ha an e olu i e, explo a i e EAM ha allows o pilo ing solu ions a he business uni le el (a.k.a EA segmen , o domain; see B uls e al. 2010; Pulkkinen 2006) is a ques ion o an ins ance o sys ems e olu ion, inducing he e-s uc u ing o he (sub-) sys ems wi hin he en e p ise. Impo an ly, he uni s o he en e p ise use hei domain knowledge (Tiwana and Kim 2015), bo h in cap u ing and in e p e ing signals u ging o esponse om hei en i onmen in en i onmen al scan (Tallon e al. 2019), in oducing he sel - o ganising ai o CAS. As essen ial is also he in e nal domain knowledge, also o he exis ing IT esou ces, managed wi h EA. This is needed o he planning and enac ing he change in he sys ems o espond, e en p oac i ely, o he en i onmen p essu es, as he eme gence ai in CAS: Adap ing (sub)sys ems o he business en i onmen changes. 2.2 En e p ise A chi ec u e Managemen Re lec ing on he heo e ical ame in Figu e 1, he ole o EAM is o a ain awa eness, assu ance and alignmen , ac oss all he sys ems, as well as he need o analy ical ools and planning aid (Ahlemann e al. 2012, Douce e al. 2009; Aie e al. 2009). This is needed o he IT ( echnical sys ems), oge he wi h he business managemen and go e nance (Ahlemann e al. 2012). The de elopmen o new capabili ies equi es awa eness o he exis ing IT (and o he ) esou ces and a chi ec u es, as well as he alignmen o goals and esou ces. Analysis o bo h business and IT a chi ec u es, and suppo o planning as he EA echniques does, can suppo agili y in capabili y building and e-building (Pulkkinen & Hi onen 2005). Me hods o agile EA de elopmen (Boekhol z 2017, Ben e e al. 2012) howe e , do no guide in de eloping a manage ial sys em o o e sigh and con ol (EAM). The EAM as a ool o ITG o bo h exploi a ion and explo a ion should possess mechanisms ha allow i o pu esou ces in conduc ing he en i onmen al scan and subsequen ly, o plan and implemen change as needed, howe e , backed up wi h he co po a e EA p inciples o ensu e compliance, in e ope abili y and balance. The EAM is o be equipped o i s ly, enabling he a chi ec u al lexibili y needed, secondly, he e-s uc u ing o he sys ems, e en e lec ing o en e p ise s uc u es i needed a e new sys em implemen a ion. 3. CASE STUDY Ou wo cases aced a common p oblem, al hough undamen ally di e en : “Alpha” is a la ge public agency, “Be a” a p i a e co po a ion. A s a egic choice in bo h is, o de elop a new business capabili y wi h a echnology new o he en e p ise, he co po a e IT and he business uni s. The s a egic in en en ails a as mo e. Alpha is building a i ual cus ome se ice assis an le e aging AI, and aims a a ‘ i s mo e ’ ad an age wi h his echnology. Be a implemen s a new business se ice concep applying IoT, o suppo he use s o hei physical echnology- in ensi e p oduc cus ome s. Bo h o ganiza ions ha e a co po a e IT depa men , wi h an associa ed EAM unc ion o guide en e p ise IT de elopmen s. In bo h cases alike, a business unc ion speci ic goal, sensed a he le el o he unc ion managemen , igge s a new business capabili y de elopmen . The business unc ion le el echnology scan (al hough no o icially assigned o do his!) in bo h cases spo ed new echnology enable s (AI, IoT) on he ma ke , no cu en ly used in hei o ganiza ion. Fo a apid s a egic mo e, building a new IT capabili y wi hin hei co po a e IT uni s equi ed o AI/IoT would ake oo much ime and is he e o e ejec ed in bo h cases. The ans o ming o he IT uni o bimodali y would no sol e he p oblem. The EAM, on i s pa , is o ensu e compliance and isk managemen as he new echnological capabili y enable s a e in eg a ed in o he en e p ise IT a chi ec u es. Table 1 The da a o he quali a i e analysis Da a sou ces a ailable Case Alpha Case Be a S a egy and plans Yes Yes O ganiza ional guidelines & s anda ds (documen s) 5 4 P ojec plan Yes Yes Numbe o design wo kshops (1-2 hou s each) 21 18 Numbe o wo kshop pa icipan s om o ganiza ion 5-6 2-11 Resea ch me hod. The esea ch da a (see Table 1 abo e) used o he quali a i e con en analysis e e s o he wo k documen s, in e iews and wo kshops du ing case p ojec s. The i s au ho was obse ing he cases du ing he p ojec li e ime: Alpha 6 mon hs, Be a 12 mon hs. The wo kshop pa icipan s we e business p ocess owne s, and membe s o LoBs, business de elopmen and IT depa men s. 4. CASE ANALYSIS As seen in Table 2, bo h cases examined in ol e he implemen a ion o he o ganiza ion s a egy. Howe e , he scanning and sensing ha igge s he de elopmen comes om a lowe echelon. The u gency elemen in bo h poin s o business agili y. Ini ia i e owne ship: The lead in he explo a i e de elopmen is wi h he line o business (LoB). They ha e he equi ed knowledge o he cus ome needs and a ision o he new se ice, i.e. he unde s anding wha he echnology can do o he business. The LoB is also esponsible o he end esul : In case Alpha, an imp o ed cus ome se ice, and o Be a, a new e enue c ea ing a e -sales se ice. The LoB as a decision uni needs o ha e he necessa y decision eedom o he e olu ion o he sys ems i has owne ship on. Poin ing o a business capabili y pe spec i e, i was c ucial o app oach he de elopmen om he cus ome and p ocess pe spec i e, i.e. including all business capabili y elemen s, iden i ying new p ocesses and oles, inally c ea ing an eme ging socio- echnical sys em. Au oma ed se ice equi es a new le el o unde s anding he cus ome needs. The AI componen s and con en managemen equi e new asks, skills and ooling. Especially, o he AI solu ion, managing he co pus and he g ound u h aining and es ing we e no el addi ions o Alpha capabili ies. Role o EA: The alue o quali y EA a e ac s is e iden in p o iding he con ex and he new solu ion con en , as well as unde s anding he e ec on exis ing p ocesses. In case Alpha, EA p o ided he necessa y unde s anding o he con ex o he pilo ed solu ion and was he basis o he design o he o-be a chi ec u e. In case Be a, i p o ided he in o ma ion Table 2 The Case Analysis: The Eme ging o a New Capabili y as a Sub-Subsys em Case a ibu e Alpha Be a 1 Business d i e S a egy deploymen , cus ome se ice imp o emen S a egy deploymen , g ow h gene a ion 2 Capabili y de eloped New AI-based i ual assis an se ice channel o cus ome s New business concep : A e -sales p oduc se ice wi h IoT suppo 3 Business goals Se ice quali y imp o emen Cos sa ings Fi s agency o deploy AI New e enue om no el se ice business Cus ome commi men o p oduc 4 Key echnological goal o de elop en e p ise IT AI adop ion in a pilo se ice a ea o u he deploymen IoT pla o m deploymen Senso da a analy ics adop ion 5 Ini ia i e and p ojec owne ship Cus ome Se ice De elopmen Uni – o be handed o e o cus ome channel managemen Business De elopmen Uni – o be handed o e o a new uni 6 No el y o he solu ion High (no p io AI implemen a ions) High (no p io implemen a ions o IoT / SDA) 7 Type o solu ion Pilo implemen a ion P oduc ion quali y Table 3 The ole o EAM p io o he p ojec Role o EAM Alpha Be a Focus o he EAM eam Business sys ems, Adminis a i e sys ems Adminis a i e sys ems EAM ole in he p ojec In o med Consul ed Pe cei ed ole o EAM Slow, no alue Slow, limi ed alue EAM ole in pos - implemen a ion phase S anda diza ion o he solu ion C ea ed EA knowledge e en ion S anda diza ion o he solu ion C ea ed EA knowledge e en ion Case a ibu e Alpha Be a Business d i e S a egy deploymen , cus ome se ice imp o emen S a egy deploymen , g ow h gene a ion Capabili y de eloped New se ice channel o cus ome se ice New business concep (A e -sales p oduc suppo se ice) o he in eg a ion equi emen s and da a models. In bo h cases, he solu ion a chi ec u e will e ol e o be pa o he new business and IT a chi ec u es. Role o EAM ini ially (depic ed in Table 3) e ol es o enabling he echnology ans e o di use he solu ion in he o ganiza ion, suppo ing o gene alize i o o he use cases. The legacy ole o he EAM eam, a ‘ egula o y’ one as pe cei ed in bo h cases in he beginning, is he e o e challenged. EAM is ex ended o a suppo i e ole o business and sys ems e olu ion. T adi ionally, EAM in gene al has a ole o con ol and ensu e con o mi y wi h s anda ds. Fo as p og ess, sou cing mus be ex ended beyond own esou ces. In he case o no el echnology, an o ganiza ion migh no ha e he equisi e IT knowledge, nei he ime o de elop i . Sou cing managemen can be suppo ed by EAM as well. Finally, inco po a ing he esul s and he new knowledge o he EA o u he use is, howe e , essen ial. In agile business de elopmen si ua ions, EA is a ool o an agile bu balanced managemen , wi h sound bu lexible, dual capabili y EAM. The EAM may ha e a consul a i e ole du ing he de elopmen , o ake on only a e wa ds. A e he p ojec , he EAM eam e ains a chi ec u al knowledge and manages he u he deploymen o he de eloped solu ions. Bo h cases highligh he ask o EAM o main ain and upda e he EA ela ed in o ma ion (as an example o “ca og aphy”, [32]) o u he de elopmen s. F om a CAS poin o iew, he EAM suppo s he eme gence o he no el socio- echnical sys em, isola ed o de elopmen ime and a e wa ds being adjus ed in o he new EA baseline, and a new o ganiza ion s uc u e eme ging wi h he de elopmen . We discuss his nex . 5. DISCUSSION En e p ise A chi ec u e and EA managemen ela e o he sys ems na u e o en e p ises (Ab aham e al. 2013). Fo EAM, ea lie s udies sugges explaining models (Ab aham e al. 2013; Buckl e al. 2009) bu ocusing on he EAM i sel , lea ing he en e p ise as a sys em o sys ems in he backg ound. Theo e ical explana ions gi e an analy ical ool o unde s and he p oblem ield and u he consequences: We looked a he esea ch ques ion: How can EAM suppo business capabili y de elopmen in an agile manne , when i in ol es he building o a new sys em and a new uni , changing he en e p ise s uc u e? The poin o iew o he EAM is an embedded go e nance sys em o he sys em o sys ems. Looking a he cases, we see ha he incu ing phenomena can be explained as a “sys emic e olu ion” o he en e p ise as a sys em o sys ems. I sel being a CAS (o CASs), he en e p ise con ains a numbe o sub-CASs. The ocal poin in ou s udy a e he business line managemen in cha ge o business pe o mance on hei own a ea, hus in cha ge o a sub-CAS. Wi hin he decision eedom gi en in he en e p ise s a egy hey de elop he LoB s a egically. Po en ial changes in he decision making s uc u es, and he e-s uc u ing o he uni s is a delica e ma e , also in he case o he new capabili y building. As new manage ial asks and oles may open wi h he new business capabili y, and he e-o ganiza ion o he uni s may be necessa y, a nega i e de elopmen may ollow in o he pa s. This calls o enabling he e olu ion and he bi h o a new sub-CAS (a uni ) on one hand (sel -o ganising and eme gence cha ac e is ic o CAS), and on he o he hand, he e ision o he en e p ise s uc u es and lines o decision making, as needed. IT de elopmen s no mally a e moni o ed by he EAM unc ion o deploy he cu en en e p ise s a egy ( om he poin o iew o he business), and o ollow he se EA p inciples and s anda ds ( om he IT manage ial iewpoin ). The moni o ing and con ol o compliance equi es o mal p ocesses, akes ime, and in oduces igidi y o de elopmen p ojec s. A pilo p ojec means expe imen al lea ning, and he s a egic choice in ou cases was o lea e he de elopmen in o he hands o he LoBs (in bo h cases unning a empo a y p ojec o ganiza ion). In his si ua ion, i.e. an o ganiza ional change (o ganiza ional e-design), an in e im, de elopmen - ime s uc u e (Pulkkinen 2006) is keeping he p ojec as a “de elopmen ime EA segmen ” (o domain), managed by he empo a y eam ha is de eloping he new capabili y. Fo de elopmen ime, i is isola ed om he es o he en e p ise s uc u es and EA. The eam engages bo h en e p ise and ou sou ced capaci ies (p o ide expe s). This enables he new capabili y building and es ing p io o in eg a ing i o he es o he en e p ise and i s EA, as well as a managed ansi ion o a new o ganiza ional s uc u e. This p o ides a me hod o implemen eme gen beha iou o a CAS in a con olled manne . Fo he IT o ganiza ion and EAM, in bo h cases, he solu ion o choice was no o es ablish a new “agile IT” uni , wi h, o wi hou EAM, bu o employ esou ces as needed, which esembles mo e he BizDe Ops – ype (G uhn e al. 2015) o o ganizing in a empo a y eam. Howe e , he De Ops mode was no a possibili y, since he e we e no sys em ope a o s o be in ol ed, as he de elopmen was no on exis ing sys ems. The de elope s we e o a la ge ex en hi ed. The en e p ise s a egic managemen suppo s in bo h cases he agile pilo ing, jus i ied by he impo o knowledge (use o IT p o ide s) o injec knowledge and o os e o ganiza ional lea ning and de elopmen , since he s a egy sees o u he deploymen o he chosen echnology. The o ganiza ion and EA design c ea ed in he de elopmen domain can be cap u ed as a new e sion o segmen business a chi ec u e and, wi h he knowledge c ea ed in he pilo p ojec , eplica ed as e e ence a chi ec u e o o he segmen s o his en e p ise o hei capabili y de elopmen . As o he explaining heo ies, changes in he EA segmen s uc u es mean ha he sys ems s uc u e o he en e p ise e ol es. New echnology is he co e echnical sys em, en ailing a new socio- echnical sys em o eme ge, wi h among o he hings a new business p ocess o be designed as he co e o a new capabili y o be es ablished. New EA segmen s uc u e means e olu ion o he sub-CASs wi hin he en e p ise. The s a ing poin o an e olu ion s ep can be a sub-CAS (a business uni ) pe cei ing i sel u ging signals om i s en i onmen , and – wi h he managemen sys em, i.e. he op echelon, allowing – can ac upon hem, inally leading o modi ica ions, o “sys emic e olu ion”, in he s uc u e o sub-CASs wi hin he en e p ise. The signal igge ing a new capabili y de elopmen a ises om he business en i onmen . Fo he en e p ise go e nance, i may be no clea which LoBs a e a ec ed. Fo a gi en signal, he LoB whose domain knowledge unde s ands he use case bes , is also amilia wi h he speci ic equi emen s, such as legisla ion and indus y s anda ds. The sel - o ganizing beha iou acco ding o he heo y on CAS is enabled by he lexible EAM, ha conside s bo h he po en ial in achie ing goals (i.e. business pe o mance) and he isks in ol ed. In addi ion, when he LoB is allowed o inno a e, new use ul pa e ns migh eme ge, ha a e no pa o he cu en EA ( equisi e a ie y), bu can be po en ially use ul o u he uni s in he en e p ise. This is one o he known bene i s o a managed EA, and he segmen a ion o EA as an aspec o i , allowing o agile pilo ing – a capaci y o he EA wi h de ined segmen s. 6. CONCLUSIONS Concluding om he obse a ions and hei analysis, an eme gen model o a dual capabili y EAM o building new capabili ies can be summa ized as ollows:  A e conside a ion o business po en ial and isk, a empo a y eam is o med, led by he LoB ha owns he case (as a sel -o ganizing CAS)  Rep esen a i es om he a ec ed LoBs (co- e olu ion) a e engaged in he eam, and he equisi e a chi ec u al and echnical skills a e ensu ed wi h a necessa y le el o ou sou cing.  The con ex o he solu ion is s udied, e lec ing on he cu en EA, necessa ily including he business a chi ec u e.  The scope o he pilo is educed o co e only he no el elemen s, as a de elopmen - ime EA segmen . Fo echnology de elopmen , a empo a y de elopmen en i onmen is c ea ed; Cloud and i ualiza ion echnologies a e he agili y enable s he e.  Du ing he expe imen a ion phase, necessa y changes o he business p ocesses, owne ships (“decision uni ”) and oles a e designed and assigned, o a leas d a ed. This means a e-w i ing o pa s o he business a chi ec u e, o ming also he new o modi ied EA-segmen ing s uc u es: The new o ganiza ion.  A he IT a chi ec u e le el, he equi ed sys ems in eg a ions a all le els o sys ems and he cons ain s like s anda ds o be applied a e iden i ied  Wo k and p og ess a e measu ed wi h minimized managemen and EAM con ol.  The iabili y o he solu ion and he p oposed alue is e alua ed du ing and a he end o he p ojec .  I he ou come is posi i e and alue- o - business can c ea ed, he solu ion will be consolida ed, and i will be s anda dized wi h he help o he EAM eam o eplica ion in u he segmen s wi h simila use cases. This b ings in he mo e adi ional EAM ole o ensu e cohe ency and consis ency o e he whole en e p ise IT. IT is in bo h cases a igge o he s a egic de elopmen case, leading o changes in he business po olios, business a chi ec u e s uc u es, oles and esponsibili ies, and inally po en ially he decision uni s, i.e. en e p ise s uc u e. Inducing his ype o o ganiza ional change is di icul , and he empo a ily isola ed de elopmen ime segmen could gi e he en e p ise change managemen he necessa y space and ime o adjus men s. Thus he EAM allowing o he de elopmen - ime segmen enables a smoo h ansi ion, o he use o a new ope a ional echnical sys em, as well as an eme ging ope a ional socio- echnical sys em. Di e en om ea lie cases we do no sugges a pe manen “ as IT EAM” o enable eme gen capabili y de elopmen . Po en ially, wi h ime, such eam also uns in o simila p oblems as poin ed ou wi h he bimodali y o IT (Ha ke e al. 2017). Ins ead, he exis ing EAM p o isions EA in o ma ion (baseline), and enables he wo k o he empo a y eam esou ced wi h EA skills. The use o a empo a y eam o ms he co e o his dual EAM capabili y. A e he p ojec , he s abiliza ion and s anda disa ion o he solu ions a e gi en o e o he EAM eam. Bo h cases conce ned building no el capabili ies ha ha e clea and limi ed in eg a ion poin s o he exis ing a chi ec u e. When upg ading exis ing sys ems, he e o could be mo e demanding, due o exis ing in e connec ions be ween sys em elemen s. The dependencies would limi he eedom. Wi h only wo cases in s udy, he gene alizabili y o he esul migh be limi ed. On he o he hand, he o ganiza ions we e undamen ally di e en : Thei owne ship (public / p i a e), he na u e o hei business (se ice / manu ac u ing). Also, he solu ions used di e en echnology (AI / IoT). In spi e o hese di e ences, he new echnology p ojec o ganiza ion and ex ending he ole o EAM o a dual capabili y app oach in i we e simila . Fu he s udy is needed, i e.g. he no el solu ion would be mo e in e connec ed wi hin i s con ex , and o cou se wi h u he ypes o o ganiza ions. REFERENCES Ab aham, R., T ibole , J., Win e , R., T ans o ma ion o Mul i-le el Sys ems – Theo e ical G ounding and Consequences o En e p ise A chi ec u e Managemen , in: P ope , H., A ei o, D., Gaaloul, K. (Eds.), Ad ances in En e p ise Enginee ing VII, Sp inge Be lin Heidelbe g, 2013, pp. 73-87., h p://dx.doi.o g/10.1007/978-3-642-38117-1_6 Ahlemann, F., S e ine , E., Messe schmid , M., & Legne , C. (Eds.). (2012). S a egic en- e p ise a chi ec u e managemen : challenges, bes p ac ices, and u u e de elopmen s. Sp inge Science & Business Media. Aie , S., Ku pjuwei , S., Saa , J., & Win e , R. (2009). En e p ise a chi ec u e design as an enginee ing discipline. AIS T ansac ions on En e p ise Sys ems, 1(1), 36-43. Ande sen, P. (2016). Managing he IT A chi ec u e: A Mul iple Case S udy (Doc o al disse a ion, Aa hus Uni e si y). Boekhol z, V. C. M. (2017). Imp o ing En e p ise A chi ec u e a Agile Wo king Financial O ganisa ions (Mas e 's hesis). U ech Uni e si y he Ne he lands Ben e, S., Bombosch, U., & Langade, S. (2012). Collabo a i e en e p ise a chi ec u e: en iching EA wi h lean, agile, and en e p ise 2.0 p ac ices. Newnes. B uls, W. A., an S eenbe gen, M., Foo huis, R., Bos, R., & B inkkempe , S. (2010). Domain a chi ec u es as an ins umen o e ine en e p ise a chi ec u e. CAIS, 27, 27. Buckl, S., Ma hes, F., & Schweda, C. M. (2009, Oc obe ). A iable sys em pe spec i e on en e p ise a chi ec u e managemen . In 2009 IEEE In e na ional Con e ence on Sys ems, Man and Cybe ne ics (pp. 1483-1488). IEEE. Douce , G., Gø ze, J., Saha, P., & Be na d, S. A. (2009). Cohe ency managemen : Using en e p ise a chi ec u e o alignmen , agili y, and assu ance. Jou nal o En e p ise A chi ec u e, 4(2). D ews, P., Schi me , I., Ho lach, B., & Tekaa , C. (2017, Oc obe ). Bimodal en e p ise a chi ec u e managemen : The eme gence o a New EAM unc ion o a BizDe Ops-based as IT. In 2017 IEEE 21s In e na ional En e p ise Dis ibu ed Objec Compu ing Wo kshop (EDOCW) (pp. 57-64). IEEE. Fo mann, L., Ha ke, I., & Benlian, A. (2019). Na iga ing Th ough Digi al T ans o ma ion Using Bimodal IT: How Changing IT O ganiza ions Facili a es he Digi al T ans o ma ion Jou ney a Deu sche Bahn Ve ieb GmbH. In Digi aliza ion Cases (pp. 393-410). Sp inge , Cham. Eisenha d , K. M., & G aebne , M. E. (2007). Theo y building om cases: Oppo uni ies and challenges. Academy o managemen jou nal, 50(1), 25-32. G uhn, V., & Schä e , C. (2015, Sep embe ). BizDe Ops: because De Ops is no he end o he s o y. In In e na ional Con e ence on In elligen So wa e Me hodologies, Tools, and Techniques (pp. 388-398). Sp inge , Cham. Ha ke, I., Kalgo as, B., & Benlian, A. (2017). Op ions o T ans o ming he IT Func ion Using Bimodal IT. MIS Qua e ly Execu i e, 16(2). Ho lach, B., D ews, P., & Schi me , I. (2016). Bimodal IT: Business-IT alignmen in he age o digi al ans o ma ion. Mul ikon e enz Wi scha sin o ma ik (MKWI), 1417-1428. Janssen, M., Kuk, G. (2006) A Complex Adap i e Sys em Pe spec i e o En e p ise A chi- ec u e in Elec onic Go e nmen , P oceedings o he 39 h Hawaii In e na ional Con e ence on Sys em Sciences. Kelle , T., Baumga ne , T., & B ucke -Kley, E. (2018). Towa ds a mul i-modal IT. e-Socie y 2018, 209. Kohli, R., & Johnson, S. (2011). Digi al T ans o ma ion in La ecome Indus ies: CIO and CEO Leade ship Lessons om Encana Oil & Gas (USA) Inc. MIS Qua e ly Execu i e, 10(4).