Eating Your Own Dog Food - Agile and Beyond 2013

Transcript

1. 01 eating your own dog food Matt Van Vleet & DJ Daugherty Friday, March 8, 13
2. 02 who we are Matt van vleet COO / PRODUCT OWNER dj daugherty POLYGLOT Friday, March 8, 13
3. 03 ABOUT US OUR PHILOSOPHY WHAT WE THINK We are different... we are not just interested in cutting edge technology. We are interested in using technology to solve problems and deliver business value. OUR vision WHAT WE SEE Deliver the most value... in the shortest amount of time... with the biggest return on investment. do we eat our own dog food? HELL YEAH!!! Without constraints, we are learning even more. Delivering exponentially faster than we have ever before. Friday, March 8, 13
4. 04 format? STORY TIME W/ MATT Friday, March 8, 13
5. 05 let’s go ! Friday, March 8, 13
6. 06 as matt and DJ, we want to talk about agile practices we use everyday, so that... others can see the value in our craziness ! Friday, March 8, 13
7. 07 anyone hungry? Friday, March 8, 13
8. 08 as product owner, I want to fail fast if the current practices are not delivering a valuable solution, so that... I can change and deliver value quickly. Friday, March 8, 13
9. 09 fail fast FAILING FAST IS... GOOD COMPLETE FAILURE is... BAD Friday, March 8, 13
10. 10 fail fast failed 2 times before project NEW LANGUAGES (GROOVY/GRAILS & RUBY) 1. TOO MANY PRODUCT OWNERS 2. LACK OF UNDERSTANDING OF DATA MODEL 3. TREATED AS LEARNING TOOL FOR INTERNS 4. Friday, March 8, 13
11. 11 fail fast success this time... Groovy / Grails 1. more focus 2. engaged product owner 3. overcome data model challenges 4. disposed of previous code 5. Friday, March 8, 13
12. 12 as product owner, I need a way to capture the features required by the customer, so that... i can deliver a valuable product quickly. Friday, March 8, 13
13. 13 value stories Friday, March 8, 13
14. 14 value story IS THE value story solved? move on ! Friday, March 8, 13
15. 15 Learn Friday, March 8, 13
16. 16 HOW ABOUT NOW? Friday, March 8, 13
17. 17 as product owner, I need a way to to keep track of all of the features and stories... so that, I can effectively deliver a valuable product. Friday, March 8, 13
18. 18 card wall Friday, March 8, 13
19. 19 CARD WALL MAKE IS simple MAKE IT WORK thrown away regularly IT IS EASIER TO RE-PRIORITIZE WHEN EVERYTHING IS OUT OF THE WAY 2. nothing fancy WE CURRENTLY USE A SHARED GOOGLE SPREADSHEET 1. constant re-prioritization OUR CARD WALL IS ALWAYS CHANGING... WHATEVER IS AT THE TOP OF THE LIST... IS THE NEXT PRIORITY. 3. Friday, March 8, 13
20. 20 as product owner, I want ‘things’ delivered quickly Friday, March 8, 13
21. 21 do less ! Friday, March 8, 13
22. 22 do less ! Why? Friday, March 8, 13
23. 23 “PEOPLE LIE... ALTHOUGH, more charitably, you might say they don’t know what they want.” - TV’s Dr. Gregory House Friday, March 8, 13
24. 24 do less faster feedback FEEDBACK IS MUCH FASTER WHEN CODE IS RELEASED TO PRODUCTION MORE FREQUENTLY 1. validate ‘things’ are getting done MORE CONSISTENT WORKFLOW 2. prevent large batches into production TESTING IN PRODUCTION IS MUCH FASTER WHEN SMALL ‘THINGS’ NEED TESTED 3. constant improvement YOU ARE NEVER TOO FAR OFF IN THE WEEDS 4. ACTIONABLE TASKS BASED ON REAL USAGE... WITHOUT ASKING ! 5. Friday, March 8, 13
25. 25 as product owner, i need to solve my business problem, so that... Friday, March 8, 13
26. 26 perfection is not the goal Friday, March 8, 13
27. 27 form function over Friday, March 8, 13
28. 28 as product owner, I want my application in production, so that... I can use it. Friday, March 8, 13
29. 29 EARLY AND OFTEN deploy early and often Friday, March 8, 13
30. 30 “if you are not embarrassed by [the first version of your product], you’ve launched too late!” - Reid Hoffman (co founder / chairman LinkedIn) Friday, March 8, 13
31. 31 How ? Friday, March 8, 13
32. 32 early and often continuous integration Friday, March 8, 13
33. 33 early and often continuous delivery Friday, March 8, 13
34. 34 early and often continuous DEPLOYMENT Friday, March 8, 13
35. 35 early and often production ALL committed code is ready ! Friday, March 8, 13
36. 36 continuous integration ALL TESTS ARE EXECUTED PRIOR TO ARTIFACT CREATION 1. PROMOTION TO PROD ONLY ARTIFACTS ARE MOVED 2. AUTOMATIC DELIVERY IF ARTIFACT CREATED... SHIP TO PRODUCTION 3. DEFINITION OF DONE WHAT IS IT? I DO NOT KNOW... ASK THE PRODUCT OWNER! 4. early and often Friday, March 8, 13
37. 37 How about now? Friday, March 8, 13
38. 38 as product owner, I need a way to release partially complete features to a subset of userS... so that I can gather feedback on a more timely basis. Friday, March 8, 13
39. 39 feature toggles Friday, March 8, 13
40. 40 “right now, out on facebook.com, is all the code for every major thing we’re gonna do in the next six months and beyond.” - Chuck Rossi (Facebook engineer) feature toggles Friday, March 8, 13
41. 41 feature toggles the key to ‘FAST’ DELIVERY Friday, March 8, 13
42. 42 feature toggles the key to delivering value Friday, March 8, 13
43. 43 feature toggle decouples deployment from act of feature live Friday, March 8, 13
44. 44 feature toggle there is a common complaint... Friday, March 8, 13
45. 45 feature toggle too much effort it takes Friday, March 8, 13
46. 46 feature toggle first class WE TREAT FEATURE TOGGLES AS FIRST CLASS CITIZENS 1. first consideration WE CONSIDER FEATURE TOGGLES WHEN FIRST DISCUSSING STORIES 2. THOUGHT CHANGE FEATURES TOGGLES ‘DO’ CHANGE THE WAY YOU CREATE STORIES 3. Friday, March 8, 13
47. 47 feature toggle AFTER YOU HAVE TOGGLES... Friday, March 8, 13
48. 48 feature toggle PARTIALLY COMPLETE WE CAN RELEASE PARTIALLY COMPLETE FEATURES TO A SUBSET OF USERS 1. DEGRADE WE CAN ALSO DEGRADE FEATURES FOR WHATEVER REASON 2. ROLL BACK... NOT !!! WE DO NOT ROLLBACK... DISABLE AND FIX 3. Friday, March 8, 13
49. 49 as Product owner, i need to deliver my product... on time and on budget, so that... I will meet the expectations of my customer. Friday, March 8, 13
50. 50 low cost of change Friday, March 8, 13
51. 51 low cost of change Test driven development ALL STORIES START WITH UNIT TESTS... INTEGRATION TESTS ! 1. CONTINUAL REFACTORING IN ORDER TO DELIVER VALUE... BOTH THE DELIVERY OF BUSINESS AND TECH STORIES ARE REQUIRED 2. small stories IF IT IS WRONG... EASIER TO FIX 3. Friday, March 8, 13
52. 52 as a team member, i need to trust EACH MEMBER OF MY TEAM, so that... I can deliver a valuable product. Friday, March 8, 13
53. 53 trust Friday, March 8, 13
54. 54 TRUST lose the ego THERE ARE NO EGOS... WE ARE ALL TRYING TO SOLVE THE SAME PROBLEM 1. success factors NO BLAME GAME 2. THIS ISN’T THE THIRD GRADE !!! ADMIT THE UNKNOWN WHEN SOMETHING IS NOT KNOWN... BIG AND VISIBLE RESOLVES THE ISSUE MOST QUICKLY 3. WHAT NOT HOW PRODUCT OWNER FOCUSES ON PRODUCT, NOT IMPLEMENTATION 4. TRUST TO SUGGEST HOW PRODUCT OWNER SHOULD FEEL COMFORTABLE ENOUGH TO SUGGEST HOW 5. Friday, March 8, 13
55. 55 TRUST TECH DEBT EQUAL TO BUSINESS STORY, REQUIRED TO DELIVER VALUE 1. VALUE STORY DEFECTS 2. DEFECTS DISCOVERED FROM ‘REAL’ USAGE AND FEEDBACK USER STORIES DETERMINED BASED ON CURRENT VALUE STORY 3. REFACTORING PRIORITIZED THE SAME AS BUSINESS STORIES 4. Friday, March 8, 13
56. 56 last chance! Friday, March 8, 13
57. 57 THANKS as matt and DJ, we want to thank each of you for coming to listen to our craziness ! Friday, March 8, 13
58. 58 FOLLOW US WE ARE ON THE WEB. follow us if you are interested in what we are currently interested in... working on... thinking about ! matt’s Information www.pillartechnology.com mvanvleet@pillartechnology.com @vleet dj’s Information ddaugherty@pillartechnology.com @ddaugher Friday, March 8, 13
59. 59 WEB RESOURCES FIND OUT SLIDES ON THE WEB github repository https://github.com/ddaugher/eatingYourOwnDogfood https://speakerdeck.com/u/ddaugher/p/eating-your-own-dog-food speakerdeck Friday, March 8, 13
60. 60 QUESTIONS & ANSWERS Q& A Ask ‘em if you got ‘em Friday, March 8, 13