Pushing Through Friction (SREcon EMEA 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; the person who wrote it left the company five years ago.

As a senior-level engineering leader, 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 actually make a difference?

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 talk about why organizational friction occurs and how to mitigate it. The ability to push through friction will distinguish you throughout your career.

* Talk notes: https://talks.danielna.com/sreconemea-2019/
* SREcon EMEA 2019: https://srecon19emea.sched.com/event/SckQ/pushing-through-friction

Transcript

1. Photo by Joshua Ness on Unsplash
2. Senior Engineer: You know, I've figured out a way to improve our database performance by 75%.
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. 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. 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. 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. Uh... then why not do it? Seems like an obvious win. What gives?
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. 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. 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. Did you ever follow up outside of sending an email?
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. •
14. • •
15. • • •
16. Dan Na Staff Engineer, Squarespace Internationalization Platform
19. talks.danielna.com @dxna
20. Dan Na, Squarespace SREcon19 EMEA | October 04, 2019 | Dublin, Ireland
36. • •
37. • •
38. •
39. • •
40. • • •
41. Friction lies within the gap between how things are and how they should be; your reality and your ideal.
44. Category # of employees < 100 100 - 2,000 2,000 - Google (100,000+)
45. Category # of employees < 100 Mid-sized 2,000 - Google (100,000+)
47. • • •
48. • • • • • •
50. The success criteria moves from individuals shipping code out the door as fast as they can to
58. The normalization of deviance is when deviant behavior becomes the norm.
61. •
62. • •
63. • • •
64. • • • •
65. Organizations and processes incur friction .
66. • •
73. • •
76. • • • •
77. • • • • • • •
78. • • • • • • • • •
79. • • • • • • • • • • • • • •
88. “ “The normalization of deviance in healthcare delivery.” Banja, 2011
89. Photo by Oskar Yildiz on Unsplash
90. 1. 2.
91. 1. Continue towards a failed launch 2.
92. Manager: Hey Dan, I wanted to let you know — we've decided to stop the project.
93. 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?
94. 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.
95. 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.
96. 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. Oh; okay. Well that... is a bummer. Can I take some time to process this? I can't say I'm not disappointed.
97. 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.
98. 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.
100. “
101. “ “Being Glue”, Tanya Reilly
102. Make glue work In fact, make glue work
105. “ “What Google Learned From Its Quest to Build the Perfect Team” NYTimes, 2016.
108. Autonomy Mastery Purpose via “Drive”, Daniel Pink
109. npm install billing-components
110. “Oh, originally we thought we were building a suite of billing components. Over time it became only checkout.”
112. npm install billing-components npm install checkout-ui
113. Making the best engineering decision will often have friction. There’s often the correct path and the easy path.
120. While it is true that deviant behaviors normalize over time, .
121. 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.
122. Hmm… are you sure that team understands what you’re proposing? It doesn’t make sense that they wouldn’t care. 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.
123. Well I sent the email didn’t I?
124. Well I sent the email didn’t I? How many emails do you get per day?
125. Well I sent the email didn’t I? Like… a hundred... How many emails do you get per day?
126. 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?
127. 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?
128. 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?
129. 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?
130. @dxna talks.danielna.com squarespace.com/careers