Scrum Successful, but not always by the book Empirical findings, practical conclusions

Transcription

1 Scrum Successful, but not always by the book Empirical findings, practical conclusions Global Scrum Gathering Berlin September 2014 Prof. Dr. Ayelt Komus University of Applied Sciences Koblenz Prof. Dr. Ayelt Komus Updated Version Sept 21, 2014

2 EnteringThe Room 3

3 My Personal Agile Journey (So Far) Business Background (BPM, SAP) Scrum: First idea: Solves BPM-/SAP-Problems! Admiration: How do they do it? Doubts in audiences: Sounds great but not in my organization Status Quo Agile 2012: `pure play is exception rather than the rule! What s an appropriate approach to become more agile? 5

4 Revolution Or Evolution? Scrum Kanban Fully described new organizational system Incremental, evolutionary process. with roles, artefacts and ceremonies 6

5 AGENDA Revolution Or Evolution? Empirical Results: Status Quo Agile 2014 Decision Criteria For Agile Approach Hybrid / Selective Approaches Personal Conclusion 8

6 Studie Status Quo Agile Study on success and forms of usage of agile methods First study in 2012 Online-Survey in Q English and German Cooperation with GPM and IPMA participants, 30+ countries High level of media response 9

7 600+ Participants From 30+ Countries In which field are you working? Software-/ IT-Entwicklung Software-, IT-Development Externer Berater; Dienstleistungs-/IT-Lösungs-Anbieter External Consultant; Service-/IT-Solution-Provider IT-Management Other Sonstiges Product Produktentwicklung Development Process Prozessoptimierung Optimization Administrative Zentralbereich/Stabstelle department, (bspw. central Center of department Excellence) Quality Qualitätssicherung Assurance Higher Education Hochschule Sales Vertrieb Production Produktion Marketing Marketing Procurement Beschaffung Managerial Accounting Controlling Human Resource Management Personalwesen 6% 5% 5% 4% 2% 2% 1% 1% 1% 1% 1% 0% 16% 23% 31% 0% 5% 10% 15% 20% 25% 30% 35% 40% Source: Status Quo Agile

8 When did people start to use agile methods? 15

9 Timeline Most users started using agile methods during the last 4 years. 25% Since when do you personally use agile methods? 20% 21% 15% 14% 14% 12% 10% 9% 8% 6% 6% 5% 3% 3% 3% 1% 0% n=386 (Single Choice, only agile Users) Source: Status Quo Agile

10 Which methods are most popular? 18

11 Relevance Of Specific Agile Methods Scrum and IT-Kanban are the most popular methods Scrum Kanban ( IT-Kanban ) Extreme Programming Feature Driven Development Lean Design Thinking Adaptive SW Development Agile Modeling Usability Driven Development Lean Startup Dynamic System Development Unified Process Crystal How important are the following methods in your area? Source: Status Quo Agile 2014 only users of agile methods 19

12 Other Popular Methods For Scrum-Users Kanban, XP and FDD are the most important methods for Scrum-users Source: Status Quo Agile

13 How successful are agile methods? 23

14 Improvements Due To Agile Methods A great majority experienced improvements with the use of agile methods Did the usage of agile methods lead to improved results and efficiency? n=388 (single choice, only agile users) Source: Status Quo Agile

15 Evaluation Of Practiced Agile Methods Agile users: mostly good or very good results Evaluation of the entire practiced agile methods based on the following criteria. Overall performance of method Gesamte Leistungsfähigkeit der Methode n=1403 Transparency Transparenz n=1396 Customer orientation Kundenorientierung Efficency Effizienz Adherence to Termintreue schedule Teamwork Employee Mitarbeitermotivation n=1367 n=1402 n=1399 n=1405 n=1402 Sehr very gut good Gut good not Ausreichend very good poor Mangelhaft no experience Keine Erfahrung Quality Ergebnisqualität of results n=1407 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (single choice, each method) All agile users The number of answers is larger than the number of respondents because many participants evaluated more than one method (all methods rated A or B ). Source: Status Quo Agile

16 Evaluation Of Scrum In every single aspect the rating of Scrum is better than the average of agile methods altogether Please evaluate Scrum with the characteristics stated below according to your experiences in your area of the company Overall performance of method Gesamte Leistungsfähigkeit der Methode n=306 Transparency Transparenz n=305 Customer orientation Kundenorientierung Efficency Effizienz Adherence to schedule Termintreue Teamwork Teamwork n=306 n=306 n=306 n=308 very Sehr gut good good Gut not Ausreichend very good poor Mangelhaft no Keine experience Erfahrung Employee motivation Mitarbeitermotivation n=306 Quality of results Ergebnisqualität n=306 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Participants who rated method as follows: A = very important to my area or B = is used in my area next to other methods Single choice for each success criterion Source: Status Quo Agile

17 Success Rate Agile Vs. Classic Methods The evaluation of the success rate of agile processes is significantly better How would you estimate the success rate of projects/development processes performed with agile methods? The success rate of classic project management, rated by the classic user is significantly worse than the success rate of agile methods, rated by agile user (t-test: p<0,001). How would you estimate the success rate of projects/development processes performed with classic methods? 0-9% 0% 0-9% 0% % % 1% 0% agile 10-19% 20-29% 1% 4% classical PM % 1% 30-39% 4% % 3% 40-49% 2% % 9% 50-59% 9% % 8% 60-69% 22% % 26% 70-79% 33% % 28% 80-89% 21% % 24% % 5% 0% 5% 10% 15% 20% 25% 30% 35% n=375 (Nur (single eine choice Antwort agile möglich) users) Median Source: Status Quo Agile % 5% 10% 15% 20% 25% 30% 35% n=82 (single der Teilnehmer, choice user die durchgängig of Classic PM) klassische Methoden Median verwenden (Nur eine Antwort möglich) o 29

18 Comparison - Evaluation Summary all groups 100% Sum of the "very good and "good ratings 90% 80% 70% 60% 50% 40% 30% 20% 10% Scrum Kanban Extreme Programming Feature Driven Develpment Lean Design Thinking Klassisches Classic Project Projektmanagement Management 0% n: see individual evaluations Source: Status Quo Agile

19 In which fields and how are agile methods applied? 32

20 Fields Of Application The use of agile methods is not limited to software development In which areas do you use agile methods/project management? Source: Status Quo Agile 2014 n=387 n=387 (multiple (Mehrfachantworten choice, users of möglich), agile methods) Nutzer agiler Methoden 33

21 Forms of Usage (2014) Scrum `by the book` is the exception rather than the rule Hybrid and selective usage add up to 64% of the sample Projects / development processes are controlled and managed n=601 (Nur (single eine choice, Antwort mandatory möglich, Pflichtangabe) question) Source: Status Quo Agile

22 Variants Of Agile Methods The consistent use of agile methods is the exception Hybrid and selective use represent 62% of the sample How are you using agile methods? Consistently Classic Waterfall Consistently Agile Sometimes Classic, Sometimes Agile varying Mixed Form hybrid Consistently Agile varying n = 236 Mixed Form Consistently Classic Source: Status Quo Agile

23 mmmmhhh. Only about 1 quarter of agile users are `pure play -agile users Doesn t that mean? Most agile users have to / want to combine agile methods with waterfall methods Not to mention many of them are living in a waterfall-environment How do they do that? 36

24 Path of agilization 37

25 Simplified Approach Of Agile Transition There are several possible ways to become agile % of agile techniques applied Pure Play Agile Organization Mix Hybrid BDUF*- Organization % of organization *BDUF Big Desing Up Front ~ Waterfall 38

26 Where Is Your Organization? Which field describes the state of your organization? % of agile techniques applied Pure Play Agile Organization Mix 5 6 Hybrid BDUF*- Organization % of organization *BDUF Big Desing Up Front ~ Waterfall 39

27 Where Is Your Organization? Which field describes the state of your organization? % of agile techniques applied Pure Play Agile Organization Mix 5 6 Hybrid BDUF*- Organization % of organization *BDUF Big Desing Up Front ~ Waterfall 40

28 Agile Techniques Used All Agile User 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 89% 81% 80% 79% 78% 74% Which agile methods do you use? 70% 69% 66% 65% 64% 60% 58% 57% 54% 49% 45% 43% 28% 19% 17% 13% 10% n=386 number of users who stated at least one technique all agile users, multiple choice Source: Status Quo Agile

29 Agile Techniques Used By Form Of Agile Usage 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Which agile methods do you use? Durchgängig Consistently agile n=100 Hybrid n=165 Selektiv Selective n=121 n=121 n = Number of participants who tagged at least one technique Source: Status Quo Agile

30 Agile Techniques Used Scrum Users Only Participants who rated Scrum as follows:* A = very important to my area B = is used in my area next to other methods 100% 90% 80% 94% 86% 86% 85% 86% 80% 75% 76% Which agile methods do you use? 70% 70% 73% 72% 68% 64% 60% 60% 57% 54% 49% 50% 44% 40% 30% 29% 20% 21% 18% 13% 10% 10% 0% Source: Status Quo Agile 2014 n=318 (multiple (Mehrfachantworten choice) möglich) 44

31 Size Of Teams 7 plus/minus 2 people is common in Classic PM, too! Agile Users How many people are typically on a team in your area? Classic PM-Users How many people are typically on a team in your area? 2% 2% 4% 4% 5% 10% 17% % 26% % >25 >25 63% 44% n=388 (Nur (single eine choice) Antwort möglich) n=85 (nur (single eine choice) Antwort möglich) Source: Status Quo Agile

32 Product Owner Only 7% of the Scrum teams have no product owner at all Is there a Product Owner? 7% 7% 29% 57% Source: Status Quo Agile 2014 Yes, Ja, gibt there einen is a Product Owner, Owner, dieser he hat defines das letzte the Wort Sprint bei Backlog der Definition der Sprint Ziele (Sprint Backlog) Ja, Yes, gibt there einen is a Product Owner, Owner, dieser together stimmt die with Sprint the Ziele team (Sprint defines Backlog) out mit the dem Sprint Team Backlog ab Nein, No, there es gibt is keinen no Product definierten Owner Product Owner The Das Team team übernimmt does the die job Rolle of the des Product Owners n=306 (Scrum (Nur eine users, Antwort single möglich) choice) 46

33 Scrum Master About half the teams rely only on a dedicated Scrum Master; 7% have no Scrum Master at all How would you describe the role of the Scrum Master? 21% 53% 7% 19% Es There gibt is einen a Scrum Master, who der eher acts wie like ein a traditional Projektleiter project agiert manager Es There gibt is keinen no Scrum Master Es There gibt is einen a Scrum Master and und a einen traditional Projektleiter project manager There Es gibt is einen a dedicated spezifischen Scrum Scrum Master Master role wie ursprünglich in der Scrum Methodik vorgesehen Source: Status Quo Agile 2014 n=307, Scrum users, single choice 47

34 Duration Of Sprints More than 50% of the Scrum teams have 2-week-sprints How long do your Sprints typically run? 1% 1% 1% 4% 4% 22% 14% 53% 1 Woche week 2 Wochen weeks 3 Wochen weeks 4 Wochen weeks 5 Wochen weeks 6 Wochen weeks more Über 6 than Wochen 6 weeks varying Das variiert (single n=306, choice) Scrum users, single choice Source: Status Quo Agile

35 How successful are those hybrid / selective agile users? 49

36 Pure Play Agile Users Are Even More Successful 35% Wie How hoch would würden you estimate Sie Ihre Erfolgsquote the success rate der of mit projects/development agilen Methoden durchgeführten processes performed Projekte / Entwicklungsprozesse einschätzen? with agile (2014) methods? (2014) Durchgängig Consistently agile n=98 hybrid Hybrid n=161 selektiv Selective n=116 n=116 30% 25% 20% 15% 10% 5% 0% 0-9% % % % % % % % % % Source: Status Quo Agile 2014 Σ 68% Σ 47% Σ 47% 50

37 Evaluation Practiced Agile Methods (1/2) Pure play agile users are the most content in most aspect Quality of results Teamwork Consistently Durchgängig agile n=465 Hybrid n=556 Selective Selektiv n=386 Consistently classic Durchgängig project management klassisch n=84 0% 20% 40% 60% 80% 100% very Sehr good gut good Gut Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung Employee motivation Consistently Durchgängig agile n=467 Hybrid n=554 Selective Selektiv n=384 Consistently classic Durchgängig project management klassisch n=83 0% 20% 40% 60% 80% 100% Sehr very good gut Gut good Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung Adherence to schedule Durchgängig Consistently agile n=465 Hybrid n=551 Selective Selektiv n=383 Consistently classic Durchgängig project management klassisch n=84 0% 20% 40% 60% 80% 100% Sehr very good gut Gut good Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung Consistently Durchgängig agile n=464 Hybrid n=551 Selective Selektiv n=384 Consistently classic Durchgängig project management klassisch n=84 0% 20% 40% 60% 80% 100% Sehr very good gut Gut good Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung The difference betweeen all agile groups / consistently classic and the goups consistently agile / consistently classic are significant for each criterion (chi-square-test: p<0,001). Source: Status Quo Agile

38 Evaluation Practiced Agile Methods (2/2) Pure play agile users are the most content in most aspect Efficency Transparency Consistently Durchgängig agile n=465 Hybrid n=551 Selective Selektiv n=384 Consistently classic Durchgängig project management klassisch n=84 0% 20% 40% 60% 80% 100% very Sehr good gut good Gut Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung Consistently Durchgängig agile n=464 Hybrid n=550 Selective Selektiv n=382 Consistently classic Durchgängig project management klassisch n=84 0% 20% 40% 60% 80% 100% Sehr very good gut Gut good Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung Customer orientation Durchgängig Consistently agile n=463 Hybrid n=555 Selective Selektiv n=404 Consistently classic Durchgängig project management klassisch n=84 0% 20% 40% 60% 80% 100% Sehr very good gut Gut good Ausreichend not very good Mangelhaft poor Keine no experience Erfahrung The difference betweeen all agile groups / consistently classic and the goups consistently agile / consistently classic are significant for each criterion (chisquare-test: p<0,001). Source: Status Quo Agile

39 Company s Success Agile Vs. Classic PM Success in general correlates with agility (statistically not significant) 100% 90% 80% How do you estimate the success of your company over the last three years? classic agile PM 70% 60% 50% 40% 30% 20% significantly less successful than other deutlich weniger erfolgreich als andere companies Unternehmen in der the Branche industry weniger erfolgreich als andere Unternehmen the der Branche industry ähnlich erfolgreich wie andere Unternehmen industry der Branche erfolgreicher als andere Unternehmen der the industry Branche deutlich erfolgreicher als andere companies Unternehmen in der the Branche industry less successful than other companies in as successful as other companies in the more successful than other companies in significantly more successful than other 10% 0% Consistently Durchgängig agile n=93 Mischformn Hybrid n=153 n=153 Sowohl Selective als auch n=116 Durchgängig Consistently klassisch classic n=84 The difference betweeen all agile groups / consistently classic and the goups consistently agile / consistently classic are not significant (chi-square-test: p<0,272 respectively p<0,125). n=446 (Nur (single eine choice) Antwort möglich) Source: Status Quo Agile

40 Any hints when to use which agile elements or when to use agile at all? 54

41 Focusing On Tasks: Complicated And Complex Challenges Requirements Far from agreement Chaos Complex Complicated Close to agreement Simple Close to Certainty Far from Certainty Technology Source: Ralph D. Stacey: Strategic Management And Organisational Dynamics: The Challenge of Complexity 55

42 Complicated, Complex? 56

43 Focusing On The Single Person/Team: Shu-Ha-Ri Shu: Follow precisely Ha: Understand principle. Learn from other master. Integrate. Ri: Learn from own practice. Adapt and extend. 57

44 Business Value As A Factor To Decide For / Against Agile How valuable are the incrementals delivered while activities are still running? Value Agile BDUF (Waterfall) Actual Delivery (Waterfall) t 58

45 Risk As A Factor To Decide For / Against Agile How valuable are the incrementals delivered while activities are still running? Risk BDUF (Waterfall) Agile Actual Delivery (Waterfall) t 59

46 Your Point Of View? Criteria Agile / Classsical PM Which criteria do you use? Which criteria should be used? Which critieria have proven to be important? 60

47 Factors Supporting Agile Pro Agile Deficits of BDUF are felt (Waterfall-Pains) Importance of results Focus on results Highly motivated teams Transparency Product structure supports small increments (value, breakdown of task to match team size) Team Culture Users willing to participate Pure Play 64

48 Factors Contradicting Agile Contra Agile Fix Price Culture/system demands precise plans 9-to-5 culture Status is important Part-time task Just one shot at a certain date (no real-life learning possible) Labor law makes subcontracting difficult if product isn t defined beforehand Industrialization is regarded desireable by overall strategy Hybrid (Cherry Picking) 65

49 Reasons For Failure Have any of your recent agile projects failed? If so, what was the main reason? Source: Swiss Agile Study

50 Situational Parameters Suitable For Agile Methods In which situation do you consider agile methods particularly suitable? internal orientation budget of less than 1 million Euro project team of 5-9 people often to constantly recurring activities only roughly defined budgets results required that are not described in very much detail duration of 3-9 month Source: Status Quo Agile 2014 (answers picked most often and in the in the median) 67

51 What could that look like in real life? 68

52 Water-Scrum-Fall Is Common 4 Weeks 4 Weeks 4 Weeks A A B Packaging to Releases User Requirement A B C Functional Specification Realization B Functional Specification C Realization Functional Specification. A B C Rel. X.y A B C 69

53 Water-Scrum-Fall Is Common User Requirement A B C A 4 Weeks 4 Weeks 4 Weeks Functional Specification A Realization B Functional Specification B C Realization Functional Specification. Packaging to Releases A B C Rel. X.y WATER SCRUM FALL A B C 70

54 Agile Techniques For Hybrid Approaches Reduce size of deliverables, increase number of deliveries Real-life testing Time Boxing Install product ownership Check development priorities on a regular basis Improve estimation methods Impediment orientation impediment backlog Diversity in teams Daily Scrum Kanban Boards (overall / team perspective) Retrospectives Definition of Done 71

55 What s The Starting Point Of Your Agile Transition? Most of our customer s are patients that are very sick; they are in severe pain. My senior management won t have it. But I see the advantages We ve got 50 Scrum teams up and running. But we have to change the whole organization fundamentally or we won t be able to master the challenges of the future We don t do that kind of stuff: We just talk and use our heads Just another bandwagon to jump on 72

56 What s The Starting Point Of Your Agile Transition? Most of our customer s are patients that are very sick; they are in severe pain. My senior management won t have it. But I see the advantages Every organization, We don t do that kind of stuff: We just talk and use our heads We ve got 50 Scrum teams up and running. But we have to change the whole organization fundamentally or we won t be able to master the challenges of the future every task has it s own story and needs its own approach Just another bandwagon to jump on 73

57 More Often Than Not Agile Elements Can Be Included Situational Parameter (Cultural, Task, Resources ) Agile Element Table: Source Master Thesis Andreas Baulig 74

58 More Often Than Not Agile Elements Can Be Included Situational Parameter (Cultural, Task, Resources ) Agile Element Table: Source Master Thesis Andreas Baulig 75

59 More Often Than Not Agile Elements Can Be Included Agile Element Situational Parameter (Cultural, Task, Resources ) For reasons of acceptance and feasibility it s often a good idea to create an individual mix of agile elements to start with. Sometimes it s a good idea not to mention agile at all! Table: Source Master Thesis Andreas Baulig 76

60 Simplified Approach Of Agile Transition There are several possible ways to become agile % of agile techniques applied Pure Play - Growth Agile Organization Mix Hybrid - Growth BDUF*- Organization % of organization *BDUF Big Desing Up Front ~ Waterfall 77

61 Procedural Model To Be Adapted During Realization Under Cover/ First Steps Pilots Roll-Out Selective Use Agile Journey Single agile elements Training of single techniques, Coaching, Retrospectives No agile wording Elements like PO, Backlog, Dailies, Visualization Training of Scrum, Kanban, Coaching, Retrospectives First pilots running in Scrum-/Kanban-mode Training, Coaching, Retrospectives Roll-out for appropriate tasks Initialization of activities combining several agile teams Training, Coaching, Retrospectives Training, Coaching, Retrospectives?????? (Your own agile path) 78

62 My Personal Conclusion Hybrid or selective approaches are common Every organization, every task has it s own story and needs its own approach Good News: Many agile elements can be combined and used separately (Cafeteria approach) Actively strive for a good fit between the - organization/task and - the right portfolio (and language) of agile elements as a critical success factor abg

63 Questions? Thank you for your attention Contact: Prof. Dr. Ayelt Komus Mobil: