This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2021-05-19
Channels
- # ask-the-speaker-track-1 (220)
- # ask-the-speaker-track-2 (196)
- # ask-the-speaker-track-3 (323)
- # ask-the-speaker-track-4 (212)
- # bof-arch-engineering-ops (1)
- # bof-covid-19-lessons (1)
- # bof-leadership-culture-learning (1)
- # bof-project-to-product (14)
- # bof-sec-audit-compliance-grc (2)
- # demos (7)
- # discussion-main (1192)
- # discussion-more (15)
- # faq (4)
- # games (69)
- # games-self-tracker (2)
- # gather (5)
- # happy-hour (39)
- # help (79)
- # hiring (10)
- # lean-coffee (13)
- # networking (10)
- # project-to-product (12)
- # psychological-safety (1)
- # summit-info (156)
- # summit-stories (3)
- # xpo-anchore-devsecops (5)
- # xpo-cloudbees (4)
- # xpo-copado (1)
- # xpo-epsagon (1)
- # xpo-gitlab-the-one-devops-platform (13)
- # xpo-harness (1)
- # xpo-hcl-software-devops (9)
- # xpo-ibm (4)
- # xpo-itrevolution (16)
- # xpo-launchdarkly (26)
- # xpo-mirantis-devops (10)
- # xpo-pagerduty (11)
- # xpo-planview-tasktop (10)
- # xpo-redgatesoftware-compliant-database-devops (8)
- # xpo-snyk (3)
- # xpo-sonatype (4)
- # xpo-split (25)
- # xpo-synopsys-sig (4)
- # xpo-tricentis-continuous-testing (4)
@jayne I'm going to jump in here - I'm having difficulty joining agile delivery to ITIL based service delivery. Do you have any pointers around how to address this?
The newest version of ITIL4 has strived to align better with agile software delivery and DevOps. But ITIL by its own admission is still primarily a governance model with checkpoints along the way that some may see as being delays. Any process in particular (e.g. Change) that you are struggling with?
FWIW, our ITIL folks have embraced ITIL4 and supporting push to automated risk assessment that if it is low risk (by a bunch of metrics) then human approval not required
we don't have a well developed and consistent measure of error budget yet but hopefully we will get there too
I think that a key part of the challenge I'm facing is change, and the approach to risk that the service organisation (which is very risk averse) has and the very long decision process (CAB) they have. It's a mindset that I've found difficult to move.
That's good. Take a look at SRE's description of Error Budgets. The approach is really interesting because there is governance in the form of consequences for breach but it does increase flow and is not contrary to the ITIL4 principles.
I think if everyone agrees to SLOs then its less scary to experiment with error budgets as a way of handling standard changes (which are not necessarily small or even extremely low risk). Some of the ITIL teams are having a hard time letting go of the CAB or the need too review most changes instead of pushing the accountability to the developers and operation teams. Its a big paradigm shift but experimentation is a good way to look at marrying both approaches.
Experimentation sounds like a good approach, with an accountability edge - perhaps if your team regularly breaks something during a change then you should expect some additional CAB focus, but if your team generally completes their changes cleanly, then less risk = less focus
Change Management evaluates the risk of a failed change causing a disruption to the business, and attempts to increase the likelihood of a successful change being implemented. CABs were useful in the days where technology change was difficult, complex and long. Today, where a change is self contained, easily reversed, or has no impact on others outside your project, then look for the risk owner and they can approve (e.g. Product Owner). No need for a CAB!
I love the idea of Error Budgets, but that requires a 'trust but verify' environment. Many ITIL based environments are more checkpoint/gateway based. That, along with the common risk aversion, makes the transition very difficult. Experimentation in such an environment can also be awkward. Any key points you could mention that have helped that paradigm shift?
@jayne, what would you say are the key things to take away from Post Mortems if a root cause can’t be found? would it be how to potentially mitigate the problem going forward if it’s high risk?
How do SRE's embed themselves within the team and its processes? - if they are doing stuff outside "the team" on the teams code/service rtc, it may/can cause dependencies.....
It is trust but verify but it is also supporting shared accountability with ops and dev for meeting the SLOs. The budget is meant to be spent but if it or the SLOs are breached, the consequences are that new releases are delayed until dev and ops figure out what's going on and what to do to regain the SLOs. It's like your personal checking account. If you overdraw, you have to figurue out what you spent and how to get back on track. Try it with one service. Research how other enterprises are setting SLOs and error budgets.
In most organizations the ratio of SREs to Agile Teams is not necessarily 1 to 1. But there is no definite formula. I would suggest that SREs work with multiple related product teams to ensure the "service" is reliable, not just the product.
Thanks so much, Jayne! A warm welcome to our next speakers, @saahil and @lindyquick!
Just an FYI - An organisation I worked with made Capacity Management the responsibility of the risk function (rather than IT / SRE), the first time I'd seen this. This was a really innovative approach - capacity was essential to this organisation, they moved huge amounts of realtime market data to currency traders.
@jayne thanks for a really interesting presentation, which has certainly sparked some good thoughts. We're really keen to endorse You Built it, You Run It, with Product (DevOps) teams being on-call, but I am wondering whether an SRE team may help ensure a focus on reducing Toil, a holistic view on supporting complex service interdependencies, as well as aligning the best principles of ITIL with DevOps and agile teams.
SRE Teams are cropping up in many organizations to work with Product teams to focus on reducing toil and focusing on reliability. It's also giving rise to trending topics like Observability. Since SREs are also engineers, this approach is breaking down walls that have existed for a long time.
I like to evolve 'you build it you run it' to 'we build it, we own it' - I feel this reflects distributed authority and autonomy more effectively
@helen.beal Yes I like "We Build It, We Own It!" It's certainly what we're trying to achieve, although sometimes I find we need to remind Product teams that Owning it does mean you should Run and support it out of hours too! @steve.smith unfortunately it's too late for us to steal that for our https://sched.co/ijMc
I agree in principle, and we rarely do need it in reality, but is it safe to have no "insurance policy" at all? I always travel abroad with insurance but have fortunately never had to use it, yet I still continue to buy insurance every time I travel! It would only take one outage where we're loosing £500k per hour before the Exec would say that this model isn't working and we should go back to having a separate Ops team! 😢
@jayne I've not seen your talk yet, I'll go and check it out now and get back to you 👀 In general, I'm really sceptical about SRE teams. In a few companies I've visited, it's essentially just a rebadging of a DevOps team, an Enablement team, call it what you will. When I've worked on such teams, we've incorporated the really great ideas from the original SRE book - like SLOs - without referring to SRE. One example would be John Lewis & Partners, working on the digital platform teams there that report into @simon.skelton. Observability I see as totally different, achieving a sufficiently high cardinality of metrics to be able to form questions, not just collect answers, in real time... that's a real game changer, and I'm only aware of Honeycomb doing that well. I'm hoping to see it at more companies in the next year or so I co-wrote a https://www.equalexperts.com/blog/our-thinking/what-you-should-and-probably-shouldnt-try-from-sre/ a while ago, I'd be interested in your thoughts
@helen.beal > I like to evolve 'you build it you run it' to 'we build it, we own it' - I feel this reflects distributed authority and autonomy more effectively I like You Build It, You Run It, You Own It and other variants. We A/B tested some variants of the phrase at John Lewis & Partners a few years ago, You Build It You Run It (for better or worse) is a phrase that's widely recognised now.
@helen.beal > I guess the goal is to get it to a place where it barely needs any support OOO I'd respectfully disagree with you here. I think that's a little close to "https://twitter.com/SteveSmith_Tech/status/1393847551716175878", or even "ideally we'd have nothing of value that could be stolen". As @simon.skelton suggested ☝️, sometimes we buy insurance for a feeling of safety. That's understandable. At other times, we buy insurance because we genuinely have N items of financial value to protect. In the public sector, it's certainly possible that out of hours there's much less revenue at risk, and much lower costs to avoid. But in the private sector, if repeat if the need for on-call is explicitly tied to revenue at risk (as we've done at John Lewis & Partners) I see on-call as a signal of estimated/actual success - "we've got customers in New York now, not just in London", "more customers are using our services after they finish work themselves", etc. And of course, we all know failures can't be eliminated What are your thoughts?
Still think it's the long-lived product or service (that delivers value to the customer) - not sure I like it being described as a set of steps... although I do see that a lot
"supporting value streams" - convert business hypothesis into a digital enabled solution
In some of our VS, yes we have all 3 and then add the business aspect as well which has been a big challenge too
"respond faster" include legacy systems, so ingrained in how we've done things
We have various sized Value Streams depending on what we are building. Some are smaller from an R&D perspective (10s of people) to others that are building really large systems (1000's of people)
It’s an ongoing challenge - we haven’t solved it completely, but we are making good progress.
would love to hear any key tips or tricks to helping align business!
We have found bringing them into the conversation early makes a big difference.
Also aligning language between business and tech is key. We have heard from some of the key notes as well that the combination of the two is critical - eliminate the silos
its fun to come from an environment where trying to coordinate agile/digital aligning hardware/software and firmware, safety critical env...
Often the business is aligned to the value stream but the Tech department is not
The question is - Do we actually need a separate Tech Department or do we need to integrate the Agile teams in the business departments.
If the business is aligned but not the tech you more than majority work is done to bring stakeholders on board. Now identify the early adopters from tech and bring the business thereof. Do a PoC with them and showcase your success. to start you don't need everyone from business n tech to agree to your idea.
That is indeed the true agile transformation I am seeing. But I love that this is a common approach to do "agile" reorgs
Thanks for the confirmation and confidence booster @siddharth.pareek
how is your approach for embedded different from traditional software?
It’s not drastically different, however with manufacturing processes, the timeline is slower and the investment much steeper.
We learned we have to take it in much smaller pieces and approach it from a Lean perspective more than an Agile perspective as well.
one thing i've noticed is different, the embedded engineers here are extremely well versed in domain and science. I didnt find this in IT developers as much. I think the ceiling could be higher in the embedded space.
interesting observation - I will give that some additional thought as well. We often run into over engineering due to the extensive knowledge.
The irony is that model based dev takes a lot of the heavy lifting away
That's the real magic. Recognizing that you can't change everything all at once and going after the small wins that will add up to big wins
I wonder how in practice you found and trained the people to actually perform the value stream mapping - I imagine it is very time consuming and hard to follow up on the findings
I feel the business is really good at defining the value streams. We often underestimate their ability and knowledge of agile principles.
We have some agile coaches we have brought in as well as leveraged our internal Six-Sigma black belts to help with the mapping. We have different levels of maturity and revisit the mapping at least 2x yearly.
thanks 👍 did Northrop Grumman as a company also do value stream mapping before in manufacturing, or did you start as part of the digital transformation?
We had performed some VSM, however we hadn’t really put all of the pieces together to affect real change
"with our customers and end user in mind..." so easy to lose sight of this
Love this idea of a digital thread! We’ve seen it with both BMW and Cubic Corporation. Both using value stream management to align digital and physical assets to accelerate value flow.
"can't because we're tied to monetary" - need to build the muscle now before the rate of change increases even further
design process to get through the system rather than deliver to customer - omg how true
Thank you! Being able to participate in this learning journey at Northrop Grumman has been an incredible opportunity
I bet. You two are doing a fantastic job at articulating such a complex problem so well. Keep it up! :facepunch:
paired with transformation rather than how we often surprisingly see it - two ships passing in the night!
what do you do when the value stream exposes a multi year lead time? how do you tackle something so large without giving up?
This is a great question. To me this is actually more of a culture transformation issue than a value stream issue. The practice is the same (identify your longest wait times and optimize), but you need a culture of celebrating the small wins and recognizing that any change in the way we work will have compounding impacts
when the improvements are a drop in the overall ocean
multi-year lead times - my time in the defense industry is all coming back to me :rolling_on_the_floor_laughing: so crazy the scale
we aren’t there completely, but it is what we are driving to.
what do you do when the bottleneck lies on an organizational handoff?
we work to eliminate it. Sometimes we can, sometimes we can’t. We have found that by visualizing it and creating transparency around it helps to minimize the impact.
the idea of a digital thread allowed us to look at those bottlenecks and ask whether or not the digital thread actually carried through that handoff or not. So that was a really good place to start in determining if our organizational handoffs were necessary or not
we're working hard to visualize these handoffs, otherwise we just have local optimization
How does VSM and Digital Thread story resonate with your government customers? Are they embracing it, or struggling with it?
It’s largely customer dependent. Some get it some struggle and we spend more time helping the struggling folks get to the Aha moment
gotta love those dependency boards - changing hearts and minds haha
anyone can shine the light on the waste, it's getting rid of the waste that is hard work
Impediments are not in the path, impediments are the path
I just saw this quote" That's incredible. We delivered in two weeks what took three years previously"
when I first heard that, i said "you mean three weeks right? or months?" and the program lead said "NO!"
It was their first sprint too. They are continuing to blow their previous time estimates out of the water
@jonathansmart1 that's a really important point. we were lucky enough to be able to start using the digital threads to highlight that the path was running in the wrong direction!
how long did it take to wind that thread through t the 62 processes
We estimated that it was going to take 3 to 5 years to complete it. We wound through 30 some of the key processes in about 3 months.
They also identified some of the 62 that could be eliminated / retooled and whittled it down to 48 processes which have a thread through them in about 6 months from an MVP perspective
I like the idea of treating software like a part
fixing underlying architecture - was it hard to get that prioritized?
initially yes, but when we were able to visually show that it was their bottle neck, we were able to prioritize quickly.
@saahil / @lindyquick first thing first. I know you would be bombarded with too many ques so I am fine if you respond me later 😄 Ques: VSM exercise is one thing which is viewing the static process flow. The other is VSM system where you see the live data flowing and getting metrics from it. Did you started both at once or was their any transition time by let's say showcasing success VSM exercise then building tool for seeing flow on the go. Apologies for long ques.
It was program dependent. One one of the programs we didn’t have the metrics and had to create them. On another program we were able to incorporate them from the beginning and then refine/address the flow over time. We took what was available and then built out what we needed over time
"what are we trying to achieve" - why - no one was able to articulate - SO COMMON
"what are the use cases" for the platform - I was so shocked to learn that this why is so rarely known
Looking at the use cases for the platform was the catalyst for bringing clarity
I am thinking now about vision/mission, use cases and the importance of communicating (and re-communicating?) these for these reasons
With the digital thread approach, how do you represent/relate customers along the thread
Comms are a challenge, one success pattern we have found is the frequent demos of working systems with the fast feedback. Once the customers became accustomed to seeing incremental delivery they really jumped on board
with these larger value streams, I find that software is rarely the bottleneck. It is often the requirements and validation that take the longest
yes, we are working hard to start visualizing this and how long sits in WIP, contributes to flow time as well as the whole right side of the life cycle. Years of work in backlogs!
the system is doing what it's designed to do
“we didn’t know how to talk about the work we’re doing” 🙂 - and you still were able to convince people to participate!! Very impressive and encouraging
there was definitely a burning platform under their feet, so they were eager for a way out of the madness
really inspired - if can do it under those complex conditions, imagine what is possible!
I was a little scared about this talk at the start seeing the SAFe badges 😈 but love how you approached this talk and your value stream reorgs @saahil @lindyquick
As we return from the break, we look forward to hearing from @me1342 and @ffion!
There is a DOES specific discount for the “People Before Tech: The Importance of Psychological Safety and Teamwork in the Digital Age” book that came out last week which features a Foreword by Prof Dr Amy Edmondson and an interview with Gene Kim! For 25% off People Before Tech, use the discount code DOES2021 at http://www.bloomsbury.com/peoplebeforetech but you can get it off Amazon and some brick&mortar shops as well.
I tried getting the PDF version, but the discount code says it applies yet the price doesn't deviate...
Toyota started by tell employees that no one would lose their job as a result of continuous improvement
Removing fear of job security is definitely a good place to start, especially at the moment
Yes, bit contentious that one in my view as "job safety" and PS are "close but no cigar" level
In healthcare, I found it difficult to change nursing practice until I learned to structure my process change experiments to guarentee patient safety
Healthcare and safety are such big topics. Teaming in the medical field is paramount and hard to get right. One of the things I say is that essential as PS is in the tech industry -and it SO is if you want performance- as compared to aviation or health, if you don't have it, nobody dies. They are just much less awesome.
I won't work for a client who won't promise their people that no one will lose their job as a result of the process improvements they make on my projects.
50k people interviewed 🤐. what a number @me1342, @ffion. your session seems to be interesting as we go by 🙂
I always believe it's most important for our teams to learn from an Incident, not look for blame and punishment, however I'm aware some of the 3rd parties that run some of our Manager Services, fire people who cause Major incidents, which makes me feel very uncomfortable! 😞
Thanks Simon so much for that insight. That kind of punitive consequence is so far the opposite of creating a learning frame and embracing failure which create psychological safety
we've used the blameless retrospective from the DevOps handbook to focus on that what and no the who. By removing the who from the report, people felt safer to be honest
Unfortunately the commercial structure of the contract and the contractual penalties drive that outcome, so we need to think about factoring psyc safety in to contractual negotiations!
There are examples @simon.skelton in the engineering industry and others of contracts that drive collaboration in principle. The truth is that teams/supplier relationships need an agreed method of dealing with conflict from the get go, that includes psychological safety
Interesting paradox for me; I am having to encourage teams to feel safe, prioritizing participation in psych safety over delivery, so they can feel safe enough to improve work and deliveyr
Not a paradox, they've been conditioned for so long that human work is not real work, haven't they? We all have.
It's fascinating isn't it, if we prioritize people first, we get speed and efficiency in delivery - but it doesn't work the other way around! 😉
"Impression Management" is the fear of speaking up or engaging for fear of appearing Incompetent, Ignorant, Negative, Disruptive or Intrusive - the Negative behaviour of Psychological Safety"
"measuring" safety is hard, if everybody says safety is high, it may be actually high, or completely broken
Sure, that's only the case if by "measuring" you mean ask a survey question 🙂
We do need to dig under the surface when teams tell us all is well
Ours stays inside the team bubble and is totally anonymous for that reason
Trust and psychological safety (related but different) are always the foundation stone of excellent, healthy teams
How do you measure it in surveys? And is there. a method for going deeper or validating the survey results?
When I was coaching Lean in healthcare, I repeatedly found myself 'giving cover' to front line employees. They would tell me how to fix the problem (they had told their bosses how to fix the problem, with no result). I told them that I would take the blame if things went south, give them the credit if it worked.
This is one way to measure Psychological Safety. https://engineering.atspotify.com/2014/09/16/squad-health-check-model/. Not as comprehensive as the metrics in this talk though.
When we built the software we incorporated that Spotify model together with Amy Edmondson's PS survey, they formed the basis of the algorithm actually @cncook001
Obvs neither of those were good enough but a solid starting point - and the only ones in those days. Then the task was to get forensic about the behaviours and make sure the algo is robust to spit actionable data or teams won't do it.
My team did the Humor Hackathon last week, it was tremendous fun and I hope helped bring us closer
I shall frame this! 🙂 @christian.kullmann it's one of the plays in our teams' crowdsourced Playbooks
We decided on a theme and brought some memes to the meeting and just laughed and talked through them
It allows people to share their humour which is intensely personal and creates a bond and understanding which build PS
That sounds awesome. In the past we often tried to do stuff by showing local traditions (we have many expats) or just share stories and stuff.
Is there a way to address this with external stakeholders? I feel that our own teams feel rather safe and have lots of laughs and at least do not seem to mind failing once in a while, our external stakeholders often are quite uptight and nervous from what i gather.
Is there a way to address this with external stakeholders? I feel that our own teams feel rather safe and have lots of laughs and at least do not seem to mind failing once in a while, our external stakeholders often are quite uptight and nervous from what i gather.
Interesting question! If you are in a multi org team then yes definitely.
PS applies to any team, ANY. Instant team, old jaded project teams, couples, kids building something. Question is - are external stakeholders part of a team?
Yes and i have a hard time providing a safe environment for them although i try to show them by example how it could be but i guess they get “beaten” in their original teams/orgs quite a lot. Much like that mean barnyard dog that is mean because it gets beaten a lot. Maybe ugly example but cannot think of another right now.
There is a lot of the modeling behavior from higher up in the organization required for safety to emerge. My experience has been it’s relatively easy to get teams to open up within the team itself (unless there’s already a lot of toxicity)…it’s hard to maintain that openness when there isn’t an encouraging response to their candor from external stakeholders.
Senior leaders do need to help lead the way but we do find at the senior level psychological safety can be lower. Psychological safety established really strongly in the team though can withstand a lot of buffering from external parties!
I’d like to hear more about that and how to make it happen. I’ve managed to do that in teams by being a buffer between them and senior leaders, but essentially I’m then creating the safe environment for them. In my experience, it’s frightfully easy for a Senior leader to wreck all the progress that’s made.
That is true. Then it becomes about seeking support and having the courage to challenge that behaviour in the name of creating a collaborative, high performing team environment.
I've noticed impression management when introducing people to slack, especially leaders
It can help to establish openness and trust and rules of behavioural engagement for digital channels too as you set them up
there seems to be culture issues as well. I'm still learning, but our team has a lot of developers from India and I sense they don't feel safe to speak up. a few years ago we did have a command and control leader that wasn't always kind, so that could be part of it too
But I have noticed if you wait them out, they open up
It was the same when I went to pull teams together in China, it took a while to get them comfortable
I disagree as I was saying - it's not cultural if you dig deep to the truly human level. If it's cultural you're not there yet:)
I used to ask a question, and I timed it, it would take 14 seconds to reply
This was their response then too: 🙂 I disagree as I was saying - it's not cultural if you dig deep to the truly human level. If it's cultural you're not there yet:)
With India it’s largely about personal relationships. They will adapt to the culture that’s created, but the default in many organizations and society there is that it is hierarchical. It takes some time for them to move beyond that.
Ha! Glad I'm consistent James 🙂
I genuinely believe all humans can and will be open and flexible and emotionally invested and passionate given half a chance
that's a great insight. I need to be more patient, give more time for answers (even if there is silence) and share that idea with the rest of my team
@saket.kulkarni I completely agree. When I first visited one of our teams in India they were all very 'scared' to say anything, but those barriers can definitely be brought down, but it's much easier to do that and show the human side in person!
here's a play I use. when I speak to other countries I always say "good day". It's a timezone neutral way to greet people. you would be surprised how many people greet me with this today after ten plus years of this.
got tired of hearing my peers say good morning to someone who is up late at night
yes, and we end up saying good morning and good evening. Time zone differences are heard. I like "good day"!
We often found that in interviews with applicants from india or china, espcecially when we asked questions like “Can you tell us about a time you failed? What did you do?” or “Can you tell us about the last time you asked for feedback from a superior or peer?“. For the latter question it’s often an answer along the lines “Yeah, i get regular feedback in our biweeklies” or similar.
Depending on the timezones, I once in a while adopt “Good aftermorning!” just to inject a little humor and highlight the point for other team members that we’re dealing with time differences.
What are peoples experiences of senior leaders in their organisations modelling good behaviours and creating a psychologically safe environment? Or is it practitioners at mid and lower levels that do that?
I would like to see them ask questions that show they value things other than just delivery. Like, "what have you learned this week"?
I personally think modelling is necessary and magical but insufficient -i..e if they show vulnerability it shows org permission but it isn't sufficient to change team dynamics. What we need is Psychological Safety AT THE TOP. No management/leadership structure being a group not a team and having rampant Impression Management,
Yet another reason to get them on the Dashboard @chris.gallivan278 - so they see the value of sharing Learning and Open, Courageous behaviours
in bp i think this is something we've been pretty good at. Right from top down. One of our core values & behaviors is "courage"
That's very cool @eazyd247 - would love to see how that translated into PS
Phenomenal Success
ah gotcha. I'd have to think on specific exmaples. However a lot of it stems from the nature of our work and the safety and environmental implications of getting it wrong, not having a speak up culture. stop the job etc
getting people to respond is a hard problem to solve
@ffion - thanks for an awesome session so far! I am curious, how do you start a conversation with a leader (above you especially) who has a blindspot about their behavior, especially when it’s not making the teams feel psychologically safe?
Thanks Bernard re our talk! Build trust individually with that person first, ask some good open questions so you can then listen to their perspective, reflect back what you hear and then you may find them more open to the feedback. Also as Duena mentioned earlier, speaking to them about how much PS there is in their leadership team can start them on the journey.
That’s a good one - I probably don’t spend enough time with the person I’m thinking of, mainly because of the lack of psychological safety :rolling_on_the_floor_laughing:
Like this: I have noticed everyone prioritizing delivery over tending to our team and personal health - they aren't doing this at random and that should absolutely be reversed
Thanks everybody for an awesome chat!
3Thank you, Duena and Ffion! Coming up next, author @mark.schwartz!!
Amazon and Bloomsbury as well. Next one ought to be ITRev me thinks 🙂
There is a DOES specific discount for the “People Before Tech: The Importance of Psychological Safety and Teamwork in the Digital Age” book that came out last week which features a Foreword by Prof Dr Amy Edmondson and an interview with Gene Kim! For 25% off People Before Tech, use the discount code DOES2021 at http://www.bloomsbury.com/peoplebeforetech but you can get it off Amazon and some brick&mortar shops as well.
Hello everyone - Mark Schwartz here to answer questions on IT DevOps and Corporate Social Responsibility
Thank you @ffion and @me1342 That was a great presentation and very important topic. Laughed it (need to rewatch the video though as i tried following the chat here 😄 )
Data sharing across agencies is unbelievably hard. then you have data standards, quality etc.
Did you get a percentage of housing that requires those docs to start providing housing?
data sharing - yes, hard. but the motivation is there and HUD could help standardize (they do to some extent)
I'm not convinced there is motivation to share data, beyond the big political projects that are forced through. An org usually gets no benefit sharing their data for another purpose, and suffers security worries.
@dacahill7 this is true of Permanent Supportive Housing (PSH) that's managed by the local housing authority. It's usually them that impose the requirement, but of course it comes from local laws, etc.
@mark.schwartz I would love to see Amazon encourage Arlington County VA, where HQ2 is located, to establish a serious ES plan for the county
If Amazon needs a liaison to Arlington County on sustainability topics, I'd love to help.
think about the problem of FINDING the person experiencing homelessness when housing becomnes available
I think apart from the apparent benefit for society in general big companies should definitely think about ESG and CSR to make it attractive to work there! Who wants to work in a company that could change the world, help people etc. but doesn’t use their potential for this?
@gdpowellpe Absolutely agree. We held a roundtable for all the local groups involved in homelessness, including Arlington and DC folks. Plus some of the nonprofits that provide services there (one very active on is N Street Village in DC, which provides shelters for women experiencing homelessness). Some of my info on that documentation problem came from our research with those folks.
Homelessness in Arlington County should be easy to improve by purchasing a few of the empty hotels (as LA is doing) to provide housing.
@sascha.schaerich Totally agree. And since I've been talking these things up within AWS I can see the passion people have for this.
I guess most of these environmental efficiencies are well aligned with saving money, so companies are already trying to optimize for these things
@david.read Oops - just noticed i mentioned David Reed rather than David Read - sorry about that
@david.read Yes, we give general indications of the best instance types to use. Note that it's complicated because CO2 emissions also depend on region, especially when you take Scopes 2 and 3 into consideration
I did wonder if I'd asked something I had forgotten about 🙂
Most big companies try and understand inclusion using staff surveys - measure employee engagement, spot areas of the company with problems, harrassment etc. Does this have any merit - or are there better ways?
Interesting we use diversity and inclusion, but not equity in our people charter. Trying to understand what we are missing without having equity in there.
Probably not a question but more a comment. in my space (Oil and Gas) corporate responsibility is something we've taken very seriously and been at forefront of strategy. Espeically with our expanding to become less reliant on hydrocarbons
@david.read I think they do. My suggestion (oddly, maybe) is to think of it with an engineering mentality. What do you need to "monitor" to see that the processes you've put in place are effective? Maybe more than those surveys?
Love the amazon commitments, but wonder how that ties in with the ban on unions..
@arnab.nandi Yes, I need more on Equity. We're hoping to write a follow-up blog post on the topic. Best thing I say directed toward equity is that point about reporting and then following up to see what is driving anomalies.
@mark.schwartz, you mentioned the 50% sustainability number Amazon aims for... does this include the production of the sold/shipped goods?
@mark.schwartz How can I help Amazon get to Scope 2 net zero for HQ2? Which electrical utility will give you this much green energy? Can Arlington County piggyback on your electrical supply chain?
@stijn.claes I'm sure you understand that that's a too complex for me to take on here. I think we would all agree that employees are critical stakeholders and taking care of the workforce is an important part of ESG.
off course. Just feels like different signals I'm getting from Amazon. But like I said, love the commitments you mentioned in the talk!
@schmark - I’m wondering if you have any tools or techniques to measure/link how bad code creates excess CO2…
@florian.gysin The 50% goal is just for 2030 - the 100% goal is for 2040. The 50% goal is for net-shipping-zero, just from fulfillment center to customer, i believe.
Ok, so basically only for Scope 1 and Scope 2, and not Scope 3?
I think it includes the portion of Scope 3 related to shipping - for example the packaging materials. I should probably double-check before I say that. Happy to do so if you'd like
Consequence Scanning is an approach I've heard of... for product development
@mark.schwartz we have some nice examples at Nike of how we leveraged data and machine learning to minimize the amount of air we ship and thereby the carbon footprint
@gdpowellpe HQ2 is being built to highest standards for green building - I can get you more info on that. Not sure for that building exactly how much of its energy is coming from the different sources, but I might be able to find that out too.
Pass my contact info to the people responsible for achieving Scope 2 zero for HQ2, please; <mailto:gdpowellpe@gmail.com|gdpowellpe@gmail.com>, <tel:7194331739|719 433 1739> Thnks
Feedback for what would help me use AWS better for this: I think being able to calculate how much polution we are using as a direct result of our software and a tool to show what areas are best to reduce that would help me. Based on where my team currently is, I doubt this would be made into a high priority unless we could naturally pair it with other objectives that also help the business.
I think this could as well be additionnal argument to evaluate some ratio usage/pollution of applications/services and maybe accelerate retirement/refactoring...
@mark.schwartz With what Chris said, maybe directly seeing the side effects of having an instance that hasn't been touched in a year would help encourage end of life.
Thank you @mark.schwartz for this different angle talk (and for the inspiring books), still in its infancy but opening some interesting ideas as to how we could contribute to our CSER
@mark.schwartz Deutsche Telekom is also doing a lot in regards to CSR and ESG, if it is sponsoring TV events about the situation of health care workers, moving to sustainable energy and becoming emission neutral, or things like Diversity! There is a lot in that direction, and as I mentioned, it makes me proud to work there.
Thanks everyone for the questions and feedback. As I said, I'm still learning a lot on this, but thought it was important to share what I've learned so far
Everybody has to start somewhere, and it's better than not starting at all. Found it interesting to open up the debate as to how we as technologist can contribute as well,
@mark.schwartz Great talk, putting together all these different threads. I hadn't seen the Scope 1,2,3 breakdown for AWS before; I love the visibility! Now you know what to work on!
If there is volunteer and or open source efforts that work on that homelessness documentation problem, I would be interested in looking at it.
@mark.schwartz overall very glad to hear that Amazon is investing into CSR and ESG, and hope they are serious about it. There is a lot of work to be done.