Pushing Through Friction (Web Unleashed 2019)

Things are broken. The deployment pipeline is painfully slow, your engineering team has doubled in the last year and there’s a lack of sufficient process and management. You git blame a file that’s used everywhere but nobody understands it and the person who wrote it left the company five years ago.

Experience tells you things could be better. You see the gaps; if only the company adopted policy A or dumped technology B, everyone would benefit, but there’s so much inertia, the company has always used B. You are frustrated. Can you make an actual difference here?

Yes. You are encountering organizational friction, and learning to identify, accept and push through friction is a key skill of engineering leaders. In this talk Dan will discuss why organizational friction occurs and some actionable steps on how to mitigate it. Why no matter what role you have – manager or individual contributor – the ability to push through friction will distinguish you throughout your career.

Dan tweeted a popular thread about pushing through friction in March 2019 (https://twitter.com/dxna/status/1103683631376154624) , this talk expands on the contents of that tweet.

Delivered on September 14, 2019 at Web Unleashed 2019 in Toronto, CA: https://fitc.ca/presentation/pushing-through-friction/

* Talk notes: https://talks.danielna.com/webunleashed-2019/

Transcript

2. 2 Senior Engineer: You know, I've figured out a way to improve our database performance by 75%.
3. 3 Senior Engineer: You know, I've figured out a way to improve our database performance by 75%. me: Yeah? Wow, that sounds amazing.
4. 4 Senior Engineer: You know, I've figured out a way to improve our database performance by 75%. me: Yeah? Wow, that sounds amazing. "Yeah, except it’s not going to happen."
5. 5 Senior Engineer: You know, I've figured out a way to improve our database performance by 75%. me: Yeah? Wow, that sounds amazing. "Yeah, except it’s not going to happen." Weird. Is it because it’s a lot of work?
6. 6 Senior Engineer: You know, I've figured out a way to improve our database performance by 75%. me: Yeah? Wow, that sounds amazing. "Yeah, except it’s not going to happen." Weird. Is it because it’s a lot of work? No, it’d take maybe two weeks to complete.
7. 7 Uh... then why not do it? Seems like an obvious win. What gives?
8. 8 Uh... then why not do it? Seems like an obvious win. What gives? Yeah, I emailed the head of that team. Shared my idea with proof that it'd work. But they're in the middle of an 18 month project and I didn't get a response. I guess my idea just isn't important enough.
9. 9 Wait, what? Your idea improves database performance by 75% across the entire product, takes two weeks, and is tabled for working on an 18 month project?
10. 10 Wait, what? Your idea improves database performance by 75% across the entire product, takes two weeks, and is tabled for working on an 18 month project? Yeah, I'm frustrated.
11. 11 Did you ever follow up outside of sending an email?
12. 12 Did you ever follow up outside of sending an email? Look man, I did my part. I looked into a hard problem and found a solution and offered it to the relevant team. And at the end of the day nobody cared. So I'm frustrated.
13. Outcomes • A senior engineer at a company developed a brilliant idea to dramatically increase product performance at minimal engineering cost. 13
14. Outcomes • A senior engineer at a company developed a brilliant idea to dramatically increase product performance at minimal engineering cost. • That idea will not be executed, and as a result, customers will suffer. 14
15. Outcomes • A senior engineer at a company developed a brilliant idea to dramatically increase product performance at minimal engineering cost. • That idea will not be executed, and as a result, customers will suffer. • That senior engineer feels like their ideas, no matter how brilliant, do not have an impact. 15
16. Dan Na Staff Engineer, Squarespace Internationalization Platform
19. talks.danielna.com @dxna 19
20. Dan Na, Squarespace Web Unleashed 2019 | September 14, 2019 | Toronto, Canada Pushing Through Friction
26. Agenda 1. What causes friction? 2. How can organizations and individuals overcome friction? 26
27. What is friction?
36. More friction • You git blame a file that's used everywhere but nobody understands it and the person who wrote it left the company five years ago. • You begin work to update an old, untouched but critical part of the codebase, only to realize there are no tests. 36
37. More friction • You git blame a file that's used everywhere but nobody understands it and the person who wrote it left the company five years ago. • You begin work to update an old, untouched but critical part of the codebase, only to realize there are no tests. 37
38. Organizational friction • New hires often feel lost, disengaged and unhappy. 38
39. Organizational friction • New hires often feel lost, disengaged and unhappy. • Teams are building redundant or non-complementary features. 39
40. Organizational friction • New hires often feel lost, disengaged and unhappy. • Teams are building redundant or non-complementary features. • You look around the lunch room and everyone looks the same. 40
41. Friction lies within the gap between how things are and how they should be; your reality and your ideal. Friction is all the work it’s going to take to get there. 41
42. You've promised to ship a thing on a deadline and you unexpectedly discover that the underlying system is wildly inaccurate. 42
43. A single point of failure gives her two weeks notice. 43
44. Why does friction occur?
46. “ When I started at Uber, we were almost 1,000 employees and were doubling the headcount every six months. An old-timer summarized their experience as: "We're growing so quickly that every six months we're a new company." A bystander quickly added a corollary: "Which means our process is always six months behind our head count." 46 “An Elegant Puzzle: Systems of Engineering Management” Will Larson, then Senior Engineering Manager, Uber
47. Tech company categories by headcount Category # of employees Startups < 100 Mid-sized 100 - 2,000 Large 2,000 - Google (100,000+) 47
48. Tech company categories by headcount Category # of employees Startups < 100 Mid-sized 100 - 2,000 Large 2,000 - Google (100,000+) 48
50. Startup Priorities • Ship fast • Figure out the product • Find customers 50
51. Startup Priorities • Ship fast • Figure out the product • Find customers 51 Mid-sized Priorities • Scale the product • Scale the organization • Don’t implode
53. The success criteria moves from individuals shipping code out the door as fast as they can to teams executing, in concert, with operational excellence. 53
54. Aviation Safety
56. 1. There are five checklists and mandatory control checks that are required to be run prior to flying. The pilots never ran any of them. 56
57. 1. There are five checklists and mandatory control checks that are required to be run prior to flying. The pilots never ran any of them. 2. Pilot interviews and voice recordings also showed that mandatory flight control checks were not performed on this flight, nor were they ever performed. 57
58. 1. There are five checklists and mandatory control checks that are required to be run prior to flying. The pilots never ran any of them. 2. Pilot interviews and voice recordings also showed that mandatory flight control checks were not performed on this flight, nor were they ever performed. 3. The crew received an alert message indicating that the rudder’s load limiter had activated. This is abnormal. The crew saw the alert. We know this because it was verbalized. Action taken? None. 58
59. 4. The pilot was unable to push the power levers far enough forward to achieve takeoff thrust. Worse, he actually verbalized that he wasn’t able to get full power, yet continued the takeoff anyway. 59
60. 4. The pilot was unable to push the power levers far enough forward to achieve takeoff thrust. Worse, he actually verbalized that he wasn’t able to get full power, yet continued the takeoff anyway. 5. The pilot realized the gust lock was engaged, and said so verbally several times. At this point, the aircraft had about 5,000 feet of runway remaining, which afforded plenty of time to abort the takeoff. They chose to continue anyway. 60
61. The Normalization of Deviance http:/ /danluu.com/wat/
62. The normalization of deviance is when deviant behavior becomes the norm. To anyone outside of your organization it's obvious that what you're doing doesn't make sense, but to those inside the organization it's normal and standard procedure. 62
65. Real Onboarding WTFs • We don't have a staging environment? 65
66. Real Onboarding WTFs • We don't have a staging environment? • How do we deploy to prod? Ask the CTO to do it. 66
67. Real Onboarding WTFs • We don't have a staging environment? • How do we deploy to prod? Ask the CTO to do it. • I didn’t have a 1-on-1 with my manager until a year after my start date. 67
68. Real Onboarding WTFs • We don't have a staging environment? • How do we deploy to prod? Ask the CTO to do it. • I didn’t have a 1-on-1 with my manager until a year after my start date. • There was a hard limit on how much code we could deploy, so we switched our indentation from spaces to tabs and bought ourselves a few more months. 68
69. Organizations and processes incur friction slowly. 69
70. Summary so far • Friction is an inevitable outcome of company growth, which we can’t control. • But friction is often made worse by the normalization of deviance, which we can control. 70
71. How do we fix it?
72. Organizations
73. Organizations Short-term: Discrete Fixes 73
74. 74 ORGANIZATIONS: SHORT-TERM DISCRETE FIX #1 Document single sources of truth and keep them updated
77. Single points of failure • Leads to burn-out • When those individuals leave the company is screwed 77
78. Effective docs criteria • Single source of truth • Updating docs is part of the acceptance criteria for shipping new work. 78
79. ORGANIZATIONS: SHORT-TERM DISCRETE FIX #2 Adopt processes to vet technology decisions 79
80. Goal: Major bump a dependency 80
81. • ES5 • ES6 • CoffeeScript • IcedCoffeeScript Goal: Major bump a dependency 81
82. • ES5 • ES6 • CoffeeScript • IcedCoffeeScript Goal: Major bump a dependency 82 • Semi-colons • No semi-colons
83. • ES5 • ES6 • CoffeeScript • IcedCoffeeScript Goal: Major bump a dependency 83 • Semi-colons • No semi-colons • Prettier • eslint
84. • ES5 • ES6 • CoffeeScript • IcedCoffeeScript Goal: Major bump a dependency 84 • Semi-colons • No semi-colons • Prettier • eslint • Jest • Jasmine • Mocha • No-tests
87. ORGANIZATIONS: SHORT-TERM DISCRETE FIX #3 Solicit the WTF of new hires 87
91. Organizations Long-term cultural behaviors 91
92. ORGANIZATIONS: LONG-TERM CULTURAL BEHAVIOR #1 Address hard truths -- kindly 92
94. “ Dr. Smith’s penmanship is frequently illegible, but he becomes very testy and sometimes downright insulting when a nurse asks him to clarify what he’s written down. So, rather than ask him, the annoyed nurse will proceed to the nurse’s station, consult with another nurse or two, and collectively try to decipher Dr. Smith’s scrawl. 94 “The normalization of deviance in healthcare delivery.” Banja, 2011
95. 95 Photo by Oskar Yildiz on Unsplash
96. 96 1. Continue towards a failed launch OR 2. Stop the project.
97. 97 1. Continue towards a failed launch OR 2. Stop the project.
98. 98 Manager: Hey Dan, I wanted to let you know — we've decided to stop the project.
99. 99 Manager: Hey Dan, I wanted to let you know — we've decided to stop the project. me: What?! Why? Do you know how hard we've worked on this?
100. M: Yeah, we do. I want to make a few things clear. Number one: we recognize your team’s effort on this. We think you've done amazing work in response to a really hard ask. We also think you've learned a ton and we know that everything you've learned will only help you ship harder and more visible projects moving forward. We are super proud of all of the work you've done. 100
101. But the data on the project isn't lining up. We miscalculated and it's our fault, not yours. We've received new data that proves our original forecasts were wrong, and instead of having you spend the next four months working on something we don't think will work, we'd rather put you on a project that we think is better positioned for success. 101
102. But the data on the project isn't lining up. We miscalculated and it's our fault, not yours. We've received new data that proves our original forecasts were wrong, and instead of having you spend the next four months working on something we don't think will work, we'd rather put you on a project that we think is better positioned for success. 102 Oh; okay. Well that... is a bummer. Can I take some time to process this? I can't say I'm not disappointed.
103. Yeah, absolutely. Take some PTO and get your mind right. We just want to make sure we're honest with you because we value your trust. 103
104. Okay. Well, thanks for letting me know. Yeah, absolutely. Take some PTO and get your mind right. We just want to make sure we're honest with you because we value your trust. 104
105. ORGANIZATIONS: LONG-TERM CULTURAL BEHAVIOR #2 Celebrate the glue work 105 https:/ /noidea.dog/glue
106. “ Of course coding is an important skill in a software engineering team. But there are a ton of other skills that we need to bring to work every day. Skills that can mean the difference between a project that succeeds and one that fails. 106
107. “ 107 Like noticing when other people in the team are blocked and helping them out. Or reviewing design documents and noticing what's being handwaved or what's inconsistent. Or onboarding the new people and making them productive faster. Or improving processes to make customers happy. I call all of this glue work. “Being Glue”, Tanya Reilly
108. Make glue work promotable work. In fact, make glue work mandatory for promotion. 108
109. ORGANIZATIONS: LONG-TERM CULTURAL BEHAVIOR #3 Make psychological safety paramount 109
111. “ Psychological safety is ‘‘a sense of confidence that the team will not embarrass, reject or punish someone for speaking up... It describes a team climate characterized by interpersonal trust and mutual respect in which people are comfortable being themselves.’’ 111 “What Google Learned From Its Quest to Build the Perfect Team” NYTimes, 2016.
112. Individuals
113. INDIVIDUALS Develop and own your sense of agency 113
114. 114 Autonomy Mastery Purpose Intrinsic Motivation via “Drive”, Daniel Pink
115. 115 npm install billing-components
116. 116 “Oh, originally we thought we were building a suite of billing components. Over time it became only checkout.”
118. 118 npm install billing-components npm install checkout-ui
119. Pragmatic optimism to see what realistically could be made better; Courage to acknowledge what needs to be fixed and be accountable; Persistence to stick to it. 119
120. Making the best engineering decision will often have friction. There’s often the correct path and the easy path. Take the correct path, even if it’s hard, because it ultimately leads to better outcomes. 120
121. MAJOR CAVEAT Being a hero, or an asshole, doesn’t scale. 121
122. INDIVIDUALS Strategies 122
123. INDIVIDUALS STRATEGY #1 Have important discussions face to face 123
124. INDIVIDUALS STRATEGY #2 Get to know other people on other teams and in other orgs 124
125. INDIVIDUALS STRATEGY #3 New idea? Try it once. 125
126. This is the job.
127. 127 Look man, I did my part. I looked into a hard problem and found a solution and offered it to the relevant team. And at the end of the day nobody cared. So I'm frustrated.
128. 128 Hmm, it doesn’t make logical sense to block this work. The improvement is massive and quick! Look man, I did my part. I looked into a hard problem and found a solution and offered it to the relevant team. And at the end of the day nobody cared. So I'm frustrated.
129. 129 Hmm, it doesn’t make logical sense to block this work. The improvement is massive and quick! Look man, I did my part. I looked into a hard problem and found a solution and offered it to the relevant team. And at the end of the day nobody cared. So I'm frustrated. I know!
130. 130 Are you sure that team understands what you’re proposing? It doesn’t make sense that they wouldn’t care.
131. 131 Are you sure that team understands what you’re proposing? It doesn’t make sense that they wouldn’t care. Well I sent the email didn’t I?
132. 132 Are you sure that team understands what you’re proposing? It doesn’t make sense that they wouldn’t care. Well I sent the email didn’t I? How many emails do you get per day?
133. 133 Are you sure that team understands what you’re proposing? It doesn’t make sense that they wouldn’t care. Well I sent the email didn’t I? Like… a hundred... How many emails do you get per day?
134. 134 Did you try talking to them in person? Did you try writing up a formal RFC that outlines your findings and circulating it for advocates?
135. 135 No... Did you try talking to them in person? Did you try writing up a formal RFC that outlines your findings and circulating it for advocates?
136. 136 No... That could be the best way to get this done. That team is probably swamped too. Did you try talking to them in person? Did you try writing up a formal RFC that outlines your findings and circulating it for advocates?
137. 137 No... Okay, I’ll give it a shot. That could be the best way to get this done. That team is probably swamped too. Did you try talking to them in person? Did you try writing up a formal RFC that outlines your findings and circulating it for advocates?
138. @dxna talks.danielna.com squarespace.com/careers Thank you.