ExperimentwithMVPs:theFirst"Startuppuccino"StepstoaLeanEdtechStartup DRONKHANNA,MARCOMONDINI,JEVGENIJAPANTIUCHINA,GIOVANNISTILLITTANOANDXIAOFENGWANG,The Startuppuccino Team Thisexperiencereportreflectsonthefirststepsofasmallsoftwarestartupteam,usingMinimumViableProducts(MVPs)fromtheLeanStartupapproach,blendedwithsomeagilepractices, toexperimentandvalidate their idea.TheStartuppuccinoexperiencereveals thatMVPsplaycrucialandmultiplerolesinastartuplife.TodeveloprightMVPstogetvalidatedlearning,astartupneedstounderstanditsuniquevalueproposition.Neithercouldbeachievedwithoutcontinuousexperimenting,reflectingandlearningfromtheteam.

1. INTRODUCTION

In the past several years, Lean Startup (Ries 2011) has become a prevailing approach applied by manystartups.Atthecoreoftheapproachisusingbuild-measure-learnloopsthatallowastartuptoquicklyvalidateitsbusinessideaandtotestdifferentaspectsoftheirbusinessmodels,inordertoacquirevalidatedlearning.As Ries argues, validated learning is the true measurement of the progress a startup makes (Ries 2011),especially at its early stage.MinimumViable Products (MVPs) enable the team to conduct experiments andlearnaboutthebusinessandcustomervaluewithleasteffort,thereforeitisimportantforstartupstodesignproperMVPsinordertoobtainvalidatedlearning.However,howtodesignMVPssothatastartupcanlearnthevaluablelessonsisachallengeformanyearlystagestartups.

Startuppuccino is an educational platform that intends to help entrepreneurship educators to providestudentswithbetterlearningexperiencewiththeircourses.TheteambehindStartuppuccinointendstobuildan edtech (EducationTechnology) startupdrivenby this vision.The team intends tohelp entrepreneurshipeducators tounderstand if their studentsobtain true learning from their courses,which is theuniquevalueproposition (UVP, Maurya 2012) of Startuppuccino. However to arrive at this UVP has not been an easyjourney.Totestitrequiresevenmoreeffort.InthisprocesstheteamneedstoconstantlyfigureoutwhataretherightMVPstobuild,whatisthevalidatedlearningtheteamwantstoobtainthroughexperimentingwithMVPs,andwheretolookforopportunitiestoconductexperiments.

Thisexperiencereportisyetanotheropportunityfortheteamtoreflectandlearnfromitsfirststepsasasmalledtechstartup,tryingtonavigateamongmanyunclearassumptionsanduncertaintiesandtofigureoutitsownvalidatedlearning.Thekeylessonslearntaresummarizedandreflectedupon.

2. BACKGROUND

The journey of Startuppuccino startedwith the experience and observation of two teammemberswho areuniversity teachers. They teach together on a cross-faculty entrepreneurship course that intends to providehands-onstartupexperiencetostudents.Overseveralyearsofteachingandinteractionwithlocalstartups,thetwoteachershaveobservedthattherewasalackofsupporttoearlystagestartups,especiallystudentstartupteamswhobadlyneedguidanceandsupportonhowtoturntheirideasintoconcretebusiness.Theinitialideaof the teacherswas torecommendgoodsoftware tools to initiateandsupportstartups thataremissingkeyskillsintheirteams(e.g.,design,webdevelopment).

After the discovery of several first-moving competitors and a revealing experience by participating in alocalStartupWeekendevent, theteamdiscoveredthatearlystagestartupsneedtoacquirebasicknowledgeabouthow tobuilda startupbefore they could realizewhat tools theyneedandhow they canbenefit fromthem.Basedon this realization, the teamdecided topivot toprovidingdirect guidance to entrepreneursbyconnecting them to a pool of mentors. However, the team proceeded very slowly towards this identifieddirection.Norealdevelopmentactivitycommenced.Onekeyreasonwasthatnoneoftheteammembershad

Author'saddress:DronKhanna,PiazzaDomenicani3,Bolzano,Italy,39100;email:dron.khanna@unibz.itSecondauthor'saddress:MarcoMondini,PiazzaDomenicani3,Bolzano,Italy,39100;email:mmondini@mondspace.comThirdauthor’saddress:JevgenijaPantiuchina,PiazzaDomenicani3,Bolzano,Italy,39100;email:jevgenija.pantiuchina@unibz.itFourthauthor’saddress:GiovanniStillittano,piazzaUniversità1,Bolzano,Italy,39100;email:stillittano.giovanni@yahoo.itFifthauthor’saddress:XiaofengWang,PiazzaDomenicani3,Bolzano,Italy,39100;email:Xiaofeng.wang@unibz.itCopyright2017isheldbytheauthor(s).

TheFirst“Startuppuccino”StepstowardsaLeanEdtechStartup:Page-2

experienceofcommunitybuilding,andtheteamwasnotinapositiontodoitsinceitrequiredrichconnectionswithlocalecosystemplayers.

Afterseveralbrainstormingsessionsandrealignmentofthevisionsofdifferentmembers,theteamdecidedto focusonone thing that the teamknewwellandknewwhatpotentialpainpoints theycould tackle:bringactive learning to student entrepreneurs through better supporting thework of educators. This is the newdirection that the team has committed to and has been working on since March 2016. Here is where thejourneyofStartuppuccinoreallystarts.

3. JUST-IN-TIMEDEVELOPMENTOFTHEMINIMUMVIABLEPLATFORM

Inthenextseveralmonthsfollowingthediscoveryofthenewdirection,theteamhadintensivebrainstormingsessions to share the new vision and to figure out how to start again in a leanmanner. Customer journeymapping,atoolfromDesignThinking,hasbeenappliedwiththetwoteacherstogettheunderstandingofthescope and initial set of requirements. However, the project was somehow stagnant, and no significant andexcitingprogresshasbeenmade.Theteamneededapushtomakesignificantprogress,andsoonrealizedthattherewas one coming soon. InOctober 2016, the two teacherswould start teaching the new edition of theentrepreneurialcourse.Itwouldbearealandoptimalopportunitytostarttheexperimentofouridea.Atthatmoment,theteambelievedthataconciergeMVP(Ries2011)isacorrectwaytogoforward.Thetwoteacherswould be the chosen customers and the team would gain a better understanding of the problemStartuppuccinointendstotackleaswellasthesolutionitcouldoffer.Figure1isascreenshotofStartuppuccinowebsitethatillustratestheconciergeMVPconcept.

Figure1:TheconciergeMVPofStartuppuccino

ToimplementtheconciergeMVP,theteamneededtobuildaminimumviableplatformtobeabletogatherfeedback fromstudents.Theassumption the teammadewas that,even thougheducatorswere the intendedcustomers,studentshavetoseevalueintheplatformandbewillingtouseitbeforeeducatorscanbenefitfromit. However, the team had to speed up the initial development of the platform in order to start the build-measure-learn loops. There were two choices: 1) develop very intensively during the summer and have arunning platform ready to use in the beginning of the course; or 2) take a more step-wise approach anddeveloptheapplicationalongthecourse.Partiallyduetothelimiteddevelopmentresourcesandtimetheteamhas,butmore importantlydue to theconviction the teamhasondeveloping inanagileand leanmanner toavoidwasteindevelopingsomethingthatisnotgoingtobeusefulfortheusers,theteamdecidedtogowiththesecondchoice.

TheFirst“Startuppuccino”StepstowardsaLeanEdtechStartup:Page-3

Sincethecoursehadweeklysessions,theteamadoptedweeklyiterationstoo.Inordertodecidewhattobedeveloped in each iteration and to prioritize the tasks, the argument of developing MVP has been usedextensively.The teammembersconstantlyremindedeachother thatweweredevelopingaminimumviableplatform that should allow us to understand if students would use such a platform with the support ofeducators.Anyadditionof thenew functionalitywasevaluatedusing thecriteriaofminimumandviable. Inaddition,thedevelopersandtheteachersweresittinginthesameofficespace.Theteachersplayedtheonsitecustomerroleandconstantlydiscussedaboutwhatshouldbedevelopedwiththedevelopers.

Onedecisionmadepurposefullybytheteamthatreflectstheteam’sunderstandingofMVPisnottodevelopthe graphical user interface for educators, despite the shared understanding that theywould be the payingcustomers of Startuppuccino. The teachers from the team modified database tables directly when certainchangeswereneededfromtheperspectiveofeducators.

Thisjust-in-timebutintensivemannerofdevelopmentputmuchpressureonthedeveloperstodelivertheworking functionality needed for each week. However, the benefit is visible too. The team get immediatefeedback from the students and other intended users (e.g., mentors that support student projects) on thedevelopedfunctionality,andifandwhattodevelopmoreinthefollowingweek.

4. CONTINUOUSEXPERIMENTATIONGUIDEDBYTHEUNIQUEVALUEPROPOSITION

The intensive development of the minimum viable product gave the team a sense of achievement andsatisfaction.However, italsoobscuredtheteam’svisionofamorecrucialaspectofthestartup:whatarethevaluepropositions (Osterwalder andPigneur2010) that theplatformprovides to theusers and customers?How unique are they (Maurya 2012)? How are we different from other existing platforms to supportentrepreneurshipcourses?

InthemiddleofthecourseinNovember2016,aftertheprototypedevelopmentcameintoastablestage,the team had the opportunity to support another course in another university, even though it was not anentrepreneurship course per se. The team failed to support this course due to the limited capacity at thatmoment. However, a big debate happened within the team as to whether we should support non-entrepreneurshipcourses,whichturnedouttobecrucialfortheteamtoquestionwhatweretheuniquevaluepropositions of Startuppuccino. Several brainstorming sessions were dedicated to answer this question.Eventuallytheteamagreedthatprovidingthevisibilityofthelearningofthestudentteamstotheeducatorsisthe unique value proposition that we should focus on. This UVP became the compass that guided thesubsequentdevelopmentofStartuppuccino.

Withthisnewunderstanding,theteamrealizedthatweneededtodesignnewMVPstofindtheanswerstotwokeyquestions:1)howtocapturethelearningfromstudents,and2)howtomakeitvisibletoeducators.

Right from the beginning the team knew that, for entrepreneurship courses which generally have aprevalenceofpracticalactivities,a"learningbydoing"style,conventionalintermediateandfinalexamsarenottheappropriatedevices to capture the true learning that students couldachieve.The teachersof the courseborrowedthepracticeofretrospectivesfromagilemethods.Asinthepreviouseditionsofthecourse,betweenNovemberandDecembereachstudentteamhasbeenrequiredtoattendaprivateretrospectivesessionwiththeteachingstaff,andanothersessionattheendofthecourse.Originallythemaingoaloftheretrospectiveswasfortheteacherstogetaclearideaonthestatusofallprojects.SoontheStartuppuccinoteamrealizedthatit could be a good occasion for the teachers to understand if the students have learnt properly. However,severaldrawbacksof thisapproachalsoemerged.Sincetherewereonlytworetrospectivesessionsovertheperiod of a semester (about four months), the students tended to forget what happened to their projects.Meanwhile,itwastimeconsumingandresourceintensivefortheteacherstorunretrospectivesessionswithallstudentteams(10to15teams),eveniftheyhappenedonlytwicepersemester.Basedonthisobservation,theStartuppuccinoteamrealizedthatafeaturethatcansupportcontinuousretrospectivesofstudentteamsbythemselveswould achieve the goal of capturing learningmore accurately, and save the time and energy ofeducators.

Inordertoactquickly(halfsemesterhasalreadypassed),theteamappliedtheconceptof“piecemealMVP”fromtheLeanStartupapproach,Ratherthandevelopinganativeformintheplatform,whichtakesmoretime,theteamquicklydesignedaGoogleform,whichallowedustotestwhatshouldbecollectedfromthestudentsandhowtocollect them.This formhasbeenourMVPtovalidate theriskyassumptions that "educatorscanevaluate the learning of students with the information collected by the form", and "students are able tointerpretandprovidevaluableinformationthroughthequestionsaskedintheform".Afterbeinginstructedonhow to use the form, the students had not been forced to fill the form. In the end we collected only few

TheFirst“Startuppuccino”StepstowardsaLeanEdtechStartup:Page-4

answers,andevenlesswerethosewithvaluablecontentthatcangiveagoodvisibilityofthestudents’learningto educators. The experiment result with the Google form MVP was not positive, but it still offered somevalidatedlearning.Theteamrealizedthat,inordertogatherthelearningfromstudentsregularly,aformonlymaynotbesufficient.

Meantime,tounderstandhowtovisualizethelearninghasproventobeevenmoredifficult.Theteamwentbacktotheretrospectivechartsdrawnbythestudentsfromthepreviouscourses.However,sincetheteachershavebeengivingstudentsinstructionsonhowtodrawtheretrospectivecharts(e.g.,drawatimelineofwhathappened), the team was not sure if this was really the correct way to visualize the learning. In order toexperimentonthevisualizationpart,whenthecurrenteditionof thecourseended in late January2017, theStartuppuccinoteamdidfinalretrospectiveswitheachstudentteam,andleftopentothestudentstoillustratetheir learning on flipcharts. After all charts were drawn, the team analyzed them collectively to identifycommonelementsandtoevaluatetheireffectivenessofrepresentinglearning.Onemainvalidationfromthisexperimentwasthatthemajorityofthestudentteamsdidactuallyfollowatimelineflowevenwhentheywereleftopen todraw their learning inany format.Wealsoasked themembersof theStartuppuccino team thatwerelessexposedtothecoursetointerpretthecharts,toseewhichoneswereeasierforthemtounderstandtheprojectandlearningpoints.Thechartsthatfollowedatimelinewereeasiertointerpret.

At thispoint, the teamstarted todesign the learning chart and its components thatwouldbeeventuallyimplementedintheStartuppuccinoplatform.Inordertoimplementthispartinaleanmanner,weneedednewopportunities to experiment the solution. Since we were running out of time and students, we decided toconsider our own startup experience with Startuppuccino in order to test our design of “collection andvisualizationofthelearning”.MockupMVPisthedeviceweusedtoexperimentthesolution.Weusedflipchartwithpost-itnotestosimulatetheinterfaceofonlinelearningchart.Asaresult,wecreatedourownlearningchart(seeFigure2).

Note:Post-itnoteswithgreentext–Whathappened;Post-itnoteswithredtext–Whatwehavelearnt

TheFirst“Startuppuccino”StepstowardsaLeanEdtechStartup:Page-5

Figure2:TheStartuppuccinoLearningChartAt the end of this experiment all the teammemberswere satisfiedwith the experience and the general

feelingwasahappysurprisefromtheworkandachievementweobtained.Thefeelingofbeingsurprisedwasduetothememoryofmanycrucialeventsforgottenbyanysingleteammember.Theexperimentonourselvesshowedusthevalueofhavingaplace,inthiscasetheA1-sizedflipchart,tocollectalltheinformation,andtheneedforcollaborationofalltheteammemberstobuildthelearningchart.ItalsoledustotherealizationthattheGoogleformMVPcouldbeusefulonlyinthecontextofteamretrospectives.

However,wewereawareofthelimitationofthisexperimentusingourownexperience,andthepossibilityofbiases.Moreover,itwasaone-offratherthancontinuousretrospectiveintendedbyoursolution.Thiswasthereasonwhywelookedformoreopportunitiestorunthefullexperiment.Anewopportunitycamealongwhentwoteammemberswereinvitedtorunaone-weekintensiveLeanStartupcourseinalargeuniversityinearlyMarch2017.Attheendofeachday,thestudentswereaskedtoconducttheretrospectiveandprovidethelearningusingtheMockupMVPweprovided.WeobservedhowthestudentsusedtheMockupMVPtocreatetheir learning chart, if they followed our instruction, and collected the feedback from them. The resultexperience was positive, and allowed us to learn about more detailed aspects of our solution. One mainvalidation came from two studentswhowereprofessors in another institute.They found the learning chartwas very goodway to understand students’ learning, and told us that theywould go back to trywith theirstudents!

BasedonthelearningfromthissetofexperimentsguidedbytheUVP,theStartuppuccinoteamwillfocusondeliveringtheonlinelearningchartinthecomingmonths.

5. WHATWELEARNED

InsightA:Focusonwhattheteamknowsthebest,andacquirethenewknowledgealongtheway.Originallywewanted to build a community supportedby an online platform,which could connect early

stagestartupcompaniesandentrepreneurswiththeresourcesandguidancetheyneed.However,communitybuildingdemandstheexpertisethattheteamdidnothaveandwerenotfamiliarwith.Asaconsequence,theteammoved slowly and lackedmomentum tomake real progress. After pivoting to the idea of building anactivelearningenvironmenttosupporteducatorstobetterservestudententrepreneurs,theteamfoundtheirhomeground.Twokeymembersoftheteamareeducatorsthemselves.Theyrunentrepreneurialcoursesandknowwhatstudentslikeanddonotlikeaboutthecourses.Thepainpointsthattheteamtargetsatarethepainpointsfeltbythemdirectly.Startingwithwhattheteamknowsbest,wecouldquicklyidentifywhereandhowtoproceedinamoreconfidentmanner.InsightB:Needforastrongbusinessdriver.

TheexperienceofdevelopingtheminimumviableplatformdescribedinSection3madeusrealizetheneedandeffectivenessofhavingastrongbusinessdriver.Eventhoughwedecidedandstartedwithwhatweknewthebest, thespeedof thedevelopmentstillsuffereduntilwehad identifiedonebusinesscase,whichwastosupport the upcoming entrepreneurship course. The course provided a sense of real business and urgency,whichputtheteamintorealactionandmotivatedustoproducetangibleresults.Thetwoeducatorsactedassuper early adopter of the idea and onsite customer of the solution development, and provided continuousdrivefortheteamtopushtheideaforward.InsightC:Obtainuniquevaluepropositionalongtheway.

In retrospect, themain purpose of the conciergeMVPwe createdwas to start the learning loop. Itwasdevelopedwithoutagoodunderstandingofwhattheuniquevaluepropositionofourstartupwas.However,itdoesnotmean theeffortofdeveloping theminimumviableplatformwasawaste.Wewouldnothavebeenable to achieve the understanding of our UVP without doing it. In this sense, you get your unique valuepropositionnotupfront,butalongthewayofcontinuousexperimentation.AclearlyidentifiedUVPprovideagoodtargetandguidanceforexperimentswithMVPs.AfterweunderstoodwhatourUVPis,wehadbeenabletoutilizeallthepossibleopportunitieswecouldgrabtoconductasetofexperiments,withasetofMVPs.InsightD:GrabanyopportunityorcreateityourselftoexperimentwithMVPs.

WeutilizeddifferentMVPs to conduct the experiments that allowus toobtainvalidated learningonourbusiness idea before investing significantly to develop the full solution.We startedwith the conciergeMVPwith thesupportofaminimumviableplatformthatwedeveloped ina just-in-timemannerwith thehelpofseveralagilepractices.Welearntthevalueinpiecemealdeliverythatallowedustoobtainconstantfeedbackandadjustourcourseaccordingly.Inaddition,asasmallstartupwithlimitedresources,weneedtoseizeeverypossibleopportunity,andevenproactivelycreatenewopportunities,inordertoconducttheexperimentswith

TheFirst“Startuppuccino”StepstowardsaLeanEdtechStartup:Page-6

MVPs that allow us to obtain validated learning.We need to experiment, reflect and learn in a continuousmanner.Validatedlearningisthetruemeasurementoftheprogressofaearlystagestartup,andyouwouldnotbeabletoobtainvalidatedlearningwithoutexperimentingandreflectingconstantly!

ACKNOWLEDGEMENT

We would like to thank the shepherd of our experience report, Lise Hvatum, for her careful review, valuable recommendations and timely push. Without her help, we couldn’t have made to the end of this learning journey. REFERENCES[Maurya2012]Maurya,Ash“WhyLeanCanvasvsBusinessModelCanvas?”https://blog.leanstack.com/why-lean-canvas-vs-business-model-canvas-af62c0f250f0[lastaccessdate:08-04-2017][OsterwalderandPigneur2010]Osterwalder,AlexanderandPigneur,Yves“BusinessModelGeneration:AHandbookforVisionaries,GameChangers,andChallengers.Ahandbookforvisionaries,gamechangers,andchallengers.”JohnWiley&Sons,Inc.,2010.[Ries2011]Ries,Eric“TheLeanStartup:HowToday'sEntrepreneursUseContinuousInnovationtoCreateRadicallySuccessfulBusinesses.”CrownPublishingGroup(USA),2011.