Sounds like alignment to the flow of value

It's about as close an alignment as we were able to do at the time. I think for our business the flow of value actually cuts across the business divisions. So true value stream alignment would require a reorganisation of the entire company.

That change has been a massive unlocker of value for us. Seeing high performing teams disbanded at the end of a project was just such waste.

When Joshua Arnold, Allan Kelly, and I first started talking about https://twitter.com/joshuajames/status/341152035716071424, I referenced this explosion by Allan Kelly on project funded teams https://www.allankellyassociates.co.uk/archives/690/corporate-psychopathy/

How is that measured?

Sorry I'll need to speak to an accountant 🤯

Interested to understand how future cashflows are estimated 🙂

Depends on the initiative. Each piece of work we take on has a finance partner who helps to create a financial assessment - what do we predict the 'hard' benefit will be of doing this piece of work. That might come from cost savings, increased revenue etc. We can then compare the prediction to the realised benefit as the work is delivered. (obviously the trick is to do this at the right level of granularity. Initiatives can't be too big or too small).

Interesting. With a link to OKRs? (where the KRs might include cost or revenue as a lagging indicator)?

I asked a finance person and they sent me this https://www.investopedia.com/ask/answers/032615/what-formula-calculating-net-present-value-npv.asp

we'd like to say it's linked to OKRs however we're not as disciplined as we need to be with those IMHO

Thanks 🙂 I'm familiar with how to PV future cashflows, curious as to the prediction of future cashflows

BTW, great talk, thank you!

basically my life right now

At the moment we're just in Tech however will need to see what this move takes us and who knows what the future holds....

👍 thanks for the clarity... that is definitely a further step and one that is largely left separate in other organisations

We're working towards enabling more collaboration between business and tech. My personal vision would be that eventually tech gets embedded into the business... But that can only happen when Tech is able to thrive in a distributed environment. Mature Communities of Practice etc. I don't think we're quite ready for that yet.

No question that a few years back Tech at Sainsbury's were seen as holding things back and the last to move with the times; now, we are leading the way on org iteration, op model change and cost optimisation - from back of the pack to the front, so I think we're well placed to try and take some of these principles outside of Tech

yes, this resonates with my experiences, without collaboration platforms it is hard to share outside the bubbles

or too many collaboration platforms, people do not know which one to use for what!

open space, CoPs, conferences and wow awards help connecting the disconnected information and learning bubbles

Very true David. In the award winner's story it helped to accelerate a cultural journey they were already on. Some very relevant lessons (providing we truly reflect on them) from Covid on how we come together to make bolder change 🙂

I've seen some great examples of how covid helped focus the mind. Things like payment holidays, bounceback loans were implemented very quickly because people were focussed on a clear outcome and not working on multiple priorities. WiP was limited.

never waste a good crisis 🙂

some people live for a crisis....

That's where I am

Coming right up...

Too impatient! :)

Looking forward to your talk, Steve!

Thanks Vlad, hope you're well

This talk is about the stuff I was working on, when we last saw each other

Ah, I cannot give away the plot too early... listen on!

Please hold @patrick.anderson 😆 , we'll come back to this thread if you wish - in about 10 mins

Haha!

I shall wait 😁

@simon.skelton @steve.smith very cool. great work 👍

Thanks @patrick.anderson, a lot of thought has gone into it. We don't mention in the talk that we also track the time it takes to complete an improvement task from a post-incident review or Chaos Day. That's a trailing indicator from an incident but we also treat it as a leading indicator of adaptive capacity for future incident management

I wrote a general blog on the topic a while ago - https://www.stevesmith.tech/blog/operability-measures/

Thanks @steve.smith!

You're welcome, happy to answer any questions you've got

do you have a product owner that goes from search all the way to shipment?

@steve.smith @simon.skelton - happy to follow up after the conference as well

We have Product Owners for different areas such as Search, but they are part of a Product Domain, with an overarching Product Manager too.

Each Product team also has OKRs which filter up into high level OKR's

Are you using something associated with a perfect digital order as the top most OKR?

We wouldn't have that but something Net Promotor Score related.

very interesting 🙏 , how do you incorporate the "voice of the consumer" into your OKRs?

The majority of our features are run as A/B test experiments and therefore we can see if customers like them, but as part of that experiment we also take inputs from any customer feedback too.

awesome! Great story and congrats on the hard work / success thus far

@nasello.scott thanks! I've seen Nike present previously and been inspired by that, so I think sharing what works well, and what doesn't, is really important, so thanks to you too!!

I'll probably follow up with you all after the conference

Incident Managers are part of a separate Service Management team, but they are just there to provide support to those on call by managing Incident Comms and ensuring other teams outside Digital are called in if required.

so Service Management is outside the products? We currently try to integrate this as well on a product level. So that are close to the products. There responsiblity is as well the comms and coordination only

I agree that's a great way to go. We've just had a reorg which established a new Service Management function with a 3d party, so one of our current priorities is working with them to bring Service Management much closer into the Digital Product teams.

So yes, Service Management is outside Products at the moment!

We had the same discussion if we go with a separate team or not. We now decided to do have them to the product team, let's see how this works out 🙂

I think the main thing is to try it and learn from it! Every organisation has multiple factors which influence what will, or will not, work well, so I don't think there's one perfect template that will work for everyone. I'll be very interest to hear how it works out for you! 🤞

Happy to share it! If it works out then it is definitely an option for a talk next year 🙂

I'll definitely look forward to that! It's great to see how each organisation is continuing on it's journey! The TUI talk was great yesterday!

@jonathansmart1 It's always an interesting thing to balance the level of insurance we want, but we'll always keep this under review and continue to adapt as necessary. Fortunately at the moment we're not seeing many major incidents which means we get more challenge as to whether a call rota is required, but that's a good problem to have!

you mean, the real thing? 😉

Either!

@jonathansmart1 who doesn't! I'd say the PlayStation 5 launch is a good example! 😂

Any intentionally planned Chaos Days without telling anyone?!

Unintentional = far fewer in 2021 than happened in 2018 👍

Intentional = we always tell people

I assume was a bit planned for ps5... like a prime day at amazon or black friday

At the moment real Chaos Days are all planned and communicated, however we're starting to give less and less information/notification about them, but we need to balance not impacting teams plans too much!

or say foreseen (sorry not natural born english speaker)

@christian.lefevre We certainly knew PS5 was being launched, but we were not expecting 25x our highest Black Friday peak!!

I just bought a new sofa from http://www.johnlewis.com, @simon.skelton is always watching 🙂

@simon.skelton if I give you some feedback now, could you get it live before 5pm? (kidding!) 🙂

@jonathansmart1 we'll that would be possible but I'm not sure our Product Owners would prioritise your request! 😄

Hi Andy! We have 100s of build teams across our organization, so it ultimately depends on the maturity of the team and complexity of the test.

Thank you. I'll probably have many more questions.

Ahh, so yes, linked in Jira. Cool!

For more mature systems, yes, that's something that we would expect to see that! Our Partners and Beyond team have a really great pipeline that checks for Jira progress, automated tests had a passing score, and Github PR approvals were manually done.

The Git PR approval functions as real eyes-on-code peer review and using the time to upskill and rope other resources into the story.

That's fantastic

We would look to see what other mitigating controls are in place. For example, is there monitoring or a review to identify instances where the automated deployment process has been bypassed?

This might be too AWS specific, but would something like CloudTrail fill that gap with logs about who used their credentials to do what?

That could be one solution. Keeping in mind that logging alone is not a control. There needs to be a review or monitoring of the logs.

Fair enough! But monitoring who accesses does not guarantee that this group of privileged users is not actually bypassing the process, right? In teams where we have more and more users in this scenario that navigate between development (pipe as a code and support to devs) and production, the review of this ammount of logs is not a simple question.

@genek it was our pleasure! We'd love to connect with @cornelia!

Excellent! Looking forward to connecting with @jwillis and @john_z_rzeszotarski!

In fact, we are getting close o being code complete if the second version of the 2019 DAG paper as part of the 2021 Forum Papers.

We are looking at starting a working group to discuss this under the DevOps Mutalization SIG in FINOS https://landscape.finos.org/?selected=dev-ops-mutualization

Also https://github.com/finos/devops-mutualization

I could not make it to this talk and I will try to catch up but its so interesting to see this work.

I am not sure where we are at in my team. We started some work towards continuous compliance and governance and would love to learn more and exchange notes