Why would you want to improve your test process using TPI or TMMi?
2 likes3,237 views
The document discusses the importance of improving software processes to enhance efficiency and reduce errors, citing historical examples of software failures. It emphasizes principles such as frontloading quality measures and iterative improvements, contrasting model-based approaches with non-model based strategies. The presentation also outlines methodologies for assessing and enhancing test maturity, advocating for a gradual and targeted approach to process improvement.
Why would you want to improve your test process using TPI or TMMi?
- 1. Restricted © Siemens AG 2014. All rights reserved Why Would You Want to Improve Your Process? 13th Exchange of Experience: Testing | Nuremberg, March 13, 2014 Rik Marselis (Sogeti) Geoff Thompson (Experimentus)
- 2. Restricted © Siemens AG 2014. All rights reservedPage 2 March 13, 2014 Rik Marselis, Geoff Thompson Introducing Rik & Geoff NL UKISTQB Passion for quality and for “doing IT right” Enjoy sharing knowledge: presenting and writing Projectleader TPI NEXT 3 daughters 2 daughters 1962 1961 Co Founder TMMi Director of ExperimentusFellow of Sogeti-Labs Happily married Founder ISTQBVice-chairman TestNet
- 3. Restricted © Siemens AG 2014. All rights reservedPage 3 March 13, 2014 Rik Marselis, Geoff Thompson What is “success”? Wikipedia: Achievement of a Goal; the opposite of Failure When is your business successful? Project ready in time? IT Quality meets standards? Business value is created? Clients are happy!!
- 4. Restricted © Siemens AG 2014. All rights reservedPage 4 March 13, 2014 Rik Marselis, Geoff Thompson Why do we need to improve? Because software fails and we need to be as efficient as possible in trying to stop these! London Stock Exchange LA Airport flights grounded AT&T take out ⅓ of US ‘phones London Underground free travel with Oystercard British Passport failure Intel Pentium chip maths division wrong Software failure at Stanstead closed check in Therac 25 radiation overdosing Airbus A380 incompatible software Patriot Missile System – Dhahran Mars Climate Orbiter & Polar Lander
- 5. Restricted © Siemens AG 2014. All rights reservedPage 5 March 13, 2014 Rik Marselis, Geoff Thompson We want to do the right thing in the right way
- 6. Restricted © Siemens AG 2014. All rights reservedPage 6 March 13, 2014 Rik Marselis, Geoff Thompson To do the right thing in the right way can mean... • Saving money both in the short and long term • Better risk management • Accelerated time to market • Reduction in delays to the software development life cycle • Better career opportunities • Improved working life of employees
- 7. Restricted © Siemens AG 2014. All rights reservedPage 7 March 13, 2014 Rik Marselis, Geoff Thompson Business drivers Source - IT priorities for battling the economic slowdown: UK CEO Survey | 2008 – BearingPoint/HP
- 8. Restricted © Siemens AG 2014. All rights reservedPage 8 March 13, 2014 Rik Marselis, Geoff Thompson To avoid going the wrong way!
- 9. Restricted © Siemens AG 2014. All rights reservedPage 9 March 13, 2014 Rik Marselis, Geoff Thompson Opening the gate – getting rid of barriers to quality work
- 10. Restricted © Siemens AG 2014. All rights reservedPage 10 March 13, 2014 Rik Marselis, Geoff Thompson Is your process this sort of Ferrari?
- 11. Restricted © Siemens AG 2014. All rights reservedPage 11 March 13, 2014 Rik Marselis, Geoff Thompson When you would like it to be this sort?
- 12. Restricted © Siemens AG 2014. All rights reservedPage 12 March 13, 2014 Rik Marselis, Geoff Thompson V-model, outdated? • Start/stop model • Create in one go (no increments) • Doesn’t stimulate collaboration
- 13. Restricted © Siemens AG 2014. All rights reservedPage 13 March 13, 2014 Rik Marselis, Geoff Thompson Application Lifecycle model Most people accept “Testing” as a fact of life But often they are not happy about it All activities are relevant, they may be done sequential or in parallel Applies to both Waterfall and Agile
- 14. Restricted © Siemens AG 2014. All rights reservedPage 14 March 13, 2014 Rik Marselis, Geoff Thompson So we performed an assessment to assess the test maturity Conclusion: There’s nothing really wrong with testing The fixing and rework is the real problem! Example; Client said: “Testing costs too much time & money”
- 15. Restricted © Siemens AG 2014. All rights reservedPage 15 March 13, 2014 Rik Marselis, Geoff Thompson We introduced: “the fixing-phase” Now something interesting happened; the stakeholder asked: “Why do we need a fixing phase”? “Can’t they build it right first time”? Fixing phase Source: Jerry Weinberg
- 16. Restricted © Siemens AG 2014. All rights reservedPage 16 March 13, 2014 Rik Marselis, Geoff Thompson Why do we need this fixing? Defect injection 58% of defects originate from early lifecycle activities !!
- 17. Restricted © Siemens AG 2014. All rights reservedPage 17 March 13, 2014 Rik Marselis, Geoff Thompson What often happens Requirements Design Development Testing Acceptance Implementation Maintenance Requirements Design Development Testing Acceptance Implementation Maintenance Defect inserted in activity Defect detected and fixed in activity
- 18. Restricted © Siemens AG 2014. All rights reservedPage 18 March 13, 2014 Rik Marselis, Geoff Thompson The right quality, not too less, not too much Three key principles Frontload the process with quality measures; methods, tools, training, reviews, … People are fallible early reviews And remember: Quality can’t be ‘tested in’ at the end
- 19. Restricted © Siemens AG 2014. All rights reservedPage 19 March 13, 2014 Rik Marselis, Geoff Thompson No faults forward to the next activity Requirements Design Development Testing Acceptance Implementation Maintenance Requirements Design Development Testing Acceptance Implementation Maintenance Defect inserted in activity Defect detected and fixed in activity
- 20. Restricted © Siemens AG 2014. All rights reservedPage 20 March 13, 2014 Rik Marselis, Geoff Thompson When and What to improve When: As soon as you can The next new project Something that is big enough (and important enough) to make a difference Does timing even matter? Everything A bit at a timeWhat someone else tells me needs to change! Use a model to help understand what's needed On the next business critical project Just tinker Something that gives the biggest bang for the buck Once you have a business case Only what you have a business case for What:
- 21. Restricted © Siemens AG 2014. All rights reservedPage 21 March 13, 2014 Rik Marselis, Geoff Thompson To model or not to model that is the question? Non-model based approach involves process improvements by adapting what already exists and augmenting with process components picked up at conferences/training. These processes typically tend to be home grown untried and untested. Model based process improvement incorporates process improvement with international standards and frameworks such as the SEI’s CMMI frameworks, ISO frameworks, ITIL, TPI, TMMi, PCMM and various others. Using such models allows organisations to compare their practices with internationally recognized industry good practices.
- 22. Restricted © Siemens AG 2014. All rights reservedPage 22 March 13, 2014 Rik Marselis, Geoff Thompson Do you feel lucky...
- 23. Restricted © Siemens AG 2014. All rights reservedPage 23 March 13, 2014 Rik Marselis, Geoff Thompson How! Measure Prioritise & Plan Define / Re-define Operate Validate Evolve Initiate • Action Planning • Target Setting • New process definition/refinement • Training • Pilot • Full deployment • Monitoring of results against targets • Evolution of programme approach & targets • Assess Process/ monitor against plan • Obtain agreement to progress
- 24. Restricted © Siemens AG 2014. All rights reservedPage 24 March 13, 2014 Rik Marselis, Geoff Thompson But this success is not just process…. Measure/Prioritise and Plan/ Define and Redefine Operate/Validate Evolve Enthusiasm Maximum Hype Actual Use Disillusionment Practical Benefit RIP
- 25. Restricted © Siemens AG 2014. All rights reservedPage 25 March 13, 2014 Rik Marselis, Geoff Thompson It takes time before the actual improvements are realized When ambitions for change are set high, It takes too much time before the actual savings are realized
- 26. Restricted © Siemens AG 2014. All rights reservedPage 26 March 13, 2014 Rik Marselis, Geoff Thompson Improve gradually parallel and step-by-step improvement, with small but measurable effects
- 27. Restricted © Siemens AG 2014. All rights reservedPage 27 March 13, 2014 Rik Marselis, Geoff Thompson Improve where it helps most Don’t elevate the peaks; start improving by filling the valleys
- 28. Restricted © Siemens AG 2014. All rights reservedPage 28 March 13, 2014 Rik Marselis, Geoff Thompson The basic assumption for process improvement + = People Process Result + + + = = =
- 29. Restricted © Siemens AG 2014. All rights reservedPage 29 March 13, 2014 Rik Marselis, Geoff Thompson So why would you improve? Measure increased success!!
- 30. Restricted © Siemens AG 2014. All rights reservedPage 30 March 13, 2014 Rik Marselis, Geoff Thompson Thank you !!
- 31. Restricted © Siemens AG 2014. All rights reserved Comparing the models Q&A session
- 32. Restricted © Siemens AG 2014. All rights reservedPage 32 March 13, 2014 Rik Marselis, Geoff Thompson The models compared http://www.nl.capgemini.com/bronnen/models-to-improve-your-test-process www.istqb.org
- 33. Restricted © Siemens AG 2014. All rights reservedPage 33 March 13, 2014 Rik Marselis, Geoff Thompson Comparison TMMi TPI
- 34. Restricted © Siemens AG 2014. All rights reservedPage 34 March 13, 2014 Rik Marselis, Geoff Thompson Comparisons TMMi Staged model Advantage of staged: simplicity TPI Continuous and staged model Advantage of continuous: specific improvement focus
- 35. Restricted © Siemens AG 2014. All rights reservedPage 35 March 13, 2014 Rik Marselis, Geoff Thompson Comparison - differences TMMi • Managers Model • Developed by over 500 resources worldwide • Uses ISTQB terminology – which includes TMap terminology • Aligned to CMMI • Accreditation of methods and assessors • Metrics and full lifecycle focussed • Defect Prevention and Quality control focus • Potential improvement recomendations • Free download of model (www.tmmifoundation.org) TPI NEXT • Engineers Model • One organization in different countries • Uses TMap terminology • Independent of (but usable with) SPI- models • Business Driven approach • Improvement suggestions & enablers • Free Excel-tool (www.tpinext.com)
- 36. Restricted © Siemens AG 2014. All rights reservedPage 36 March 13, 2014 Rik Marselis, Geoff Thompson Similarities • Books • Certification • TMMi professional • TPI NEXT certified • Certification of organizations • Referenced in ISTQB Advanced & Expert level • Fit well with CMMI • Encourage structured improvement of test processes
- 37. Restricted © Siemens AG 2014. All rights reservedPage 37 March 13, 2014 Rik Marselis, Geoff Thompson Ask Geoff and Rik (now is your chance!) Q & A