SAFe is designed to help businesses continuously and more efficiently deliver value on a regular and predictable schedule. It provides a knowledge base of proven, integrated principles and practices to support enterprise agility.
As an RTE, I want to have a healthy and smart Scrum Master community, because I count on them to help me run the train.
In my current ART, I am working with fifteen amazing scrum masters from development, design, documentation, testing, performance, and globalization. As you can feel, they have unique skill sets, but they all possess the quality of an excellent agile leader.
Hand-picked content for Scrum Masters
One of the things I decided to do, so I can help them grow is to prepare a weekly dose of inspiration in the form of a newsletter.
I know it’s mainstream, but so far, I got a few great feedback entries encouraging me to continue with I do.
I always read the articles before sending them out, and I am picking up content that applies to solve the problems we know we have in our small world.
Also, I want to keep it simple and entertaining because I know this is one of the few thousands of e-mails they receive. I don’t want my e-mail to go to /dev/null immediately :)
This article is a part of series of articles on building a great Community of Practice (a.k.a CoP). I plan to write on that topic almost every week. If you have the same intent as me – helping your Scrum Masters grow – keep an eye here.
Scrum of Scrums or SoS (not as in … — …) is a scaled agile technique for facilitating the coordination between all the teams working on the same objectives.
If you do Scrum already you can think about that as a large standup for people representing each team, where they do the same – answering the holy questions:
What has your team done since we last met?
What will your team do before we meet again?
Is anything slowing your team down or getting in their way?
Are you about to put something in another team’s way?
SoS is not a status meeting for the management; it is a meeting for working together as a team of teams to handle dependencies and joint problems.
I will now share how I run that meeting for maximum efficiency.
I intend to keep in short, no more than 45 min. the time is divided and planned:
First 5 min, the facilitator talks about any critical updates.
30 min for risks and dependencies management. We go team by team, and the scrum master talks shortly about the topics.
The last 10 min we save for Q/A (see below for what exactly)
Keeping track on what we decided and discussed during the Scrum of Scrums
I create a table for every meeting in our wiki for every Sprint, that looks like this:
The team name
Nothing to add here, it should be self-explanatory, but sometimes you could change it to “I wish my team name was…” and leave it to the imagination of the scrum masters.
A link to the Scrum or kanban board. If anyone wants to explore more, they should have a quick way to navigate to the board.
The name of the person that represents the team in that meeting and the point of contact of the team if you need anything from them after the meeting.
The Sprint goal usually is defined as a high-level summary of the target the product owner would like to accomplish during a sprint.
I have added it here because the team needs to know what they are willing to achieve in that Sprint.
How will the team know if they have achieved the Sprint Goal?
On how may epics(as in Jira epics), the team works at the same time. Sometimes the teams work on many items at the same time without considering the impact of that.
Little’s Law—the fundamental Law of queuing theory—tells us that the average wait time for service from a system equals the ratio of the average queue length divided by the average processing rate. (While this might sound complicated, even the line at Starbucks illustrates that.) Therefore, assuming any average processing rate, the longer the queue, the longer the wait. Simply reducing queue length decreases delays, reduces waste, increases flow, and improves the predictability of outcomes.
Again this should be used as a conversation starter, not as a blame mechanism.
As in, how are we progressing towards the goal?
Are there any open risks you would like to call out. It can be 3-rd party dependency that you can’t handle or some people leaving, and you haven’t planned for that or anything that is a risk to the delivery of the goal.
Does your team own any dependencies? We use this as a conversation started. That’s the most important part of the meeting.
Open up the Scrum of Scrums meeting
Usually, the meeting is a closed event – one person from the team is allowed to attend. I have opened that meeting to everyone by sending the invite to every team member that contributes to the product. The condition is that they can listen only, and if they have some questions, we will use the last 10 -15 min from every meeting for such a discussion.
By using this “hack,” I was able to give visibility to everyone to understand what the teams will be working on in the current Sprint. Usually, we have people from the support, field teams, and management listening to what the teams plan to do, without turning this into a status meeting.
I decided to experiment with rotating facilitators of the Scrum of Scrums. Habitually, the RTE is the leading facilitator. I believe in teamwork, and that’s why every week, there will be a different facilitator – one from the Scrum Masters. I think this boosts ownership and supports transparency.
Take that step
If you are willing to use one of my approaches, feel free and then share the results. I am also curious about how you hold your Scrum of Scrum meetings.
PI Planning is a high-priced event. You need to book a room, prepare the materials, order food, plan a social event, flight people over, pay for hotel and transportation and some meals, + even more costs.
The total price can go up to tens of thousands of euros and, in some cases, maybe more.
If you add on top of that the time people spend focusing on the PI Planning process – ~100 people for a few days, you get the whole idea.
That’s why it is very crucial to get feedback on:
How to improve the experience on the attendees
How to cut some costs
How not to repeat mistakes we did already
How to make the event more efficient
What SAFe recommends on collecting feedback
SAFe recommends having a retrospective-like event at the end of the Pi Planning, where all the attendees can put some stickies on the wall, and the facilitator will help to combine them in affinity notes, and then the RTE will create follow-up items in the backlog.
I see a bit of a problem in that since we tried it first:
The people were super tired and didn’t have the motivation to do yet another exercise.
They had feedback on their minds, but they didn’t like the attention to stand in front of everyone and put their stickies on the wall and explain why.
Some of the attendees shared verbal feedback with me during the day, but forgot to add it on the boards, thinking that the RTE will do something anyway.
What difference did I make?
I was reading a great case-study from Eik Thyrsted Brandsgård and Henrik Knibergand and based on something that they did, I decided to go even further and create a Feedback Meter:A scale from +2 (meaning this is super great, and I want this to continue) to -2 (meaning consider changing that in the future).
The board was there for the duration of the whole PI planning event, and the attendees could put their stickies on that scale. Also, I needed to communicate from time to time that the feedback board is there, and it can be used at any time.
What I have accomplished with this continuous feedback approach?
A mood meter: A quick view on the mood in the room at all time – you can see from that picture that we need some improvements :)
Act on time: A see something – do something attitude – if something needs improvement, don’t wait till the end of the Planning, but act now and add a sticky
Participation: An excellent way to read the feedback from others and to add your own to theirs.
Continuing after the PI Planning
There are 2 groups of attendees that you would miss if you stick to what SAFe recommends as a retrospective practice – those that are attending remotely and those people that love to share anonymous feedback for some reason.
I know you would tell me that this is behavior that I should not encourage, but in my case, I received surpassing feedback from that particular crowd.
To solve that problem, I have created a Podio survey and engaged anyone who could go and add even more feedback online.
What is your opinion on continuous feedback?
I am eager to learn more about how you deal with problems like that in the comments. Remember to start with +2 :)
Let me start with a bit of theory: There is an Agile framework called SAFe ® that allows enterprises to scale and deliver value faster by optimizing how it flows through different systems and sub-systems, using the principles of Lean-Agile.
One of the primary artifacts of that framework is something called a Big Room planning where all of the teams working on one program get together for a while and discuss open items, groom stories, and plan the objectives for the next 5 sprints.
I will not get into the details explaining what a sprint is or what grooming is, I assume, if you clicked on my boring title, you should know that already. If not, well, feel free to DuckDuckGo it and find out.
Let me get back to the topic. During that planning event, I, as a Release Train Engineer or say as a facilitator, am responsible for explaining some of the metrics so anyone can understand it. One of those metrics is the PPM.
I have a piece of good news for you, I don’t expect you to know what that is, because is SAFe specific and I’ll try to explain you in by copy and pasting something from the SAFe website (yes I want to make a point here)
“Each team’s planned vs. actual business value is rolled up to create the program predictability measure”.
To decode that. I’ll share with you that each team defines their own objectives during the PI planning, and the Business Owners are assigning the value to each one of them, based on the assumption of the amount of value that the team will deliver in the next few sprints.
What does it mean to be predictable in this case? (as in Program Predictability Measure)
I will not talk about how to change the ways you define and measure your PPM now, but I would definitely write another article about that.
Reduced time-to-market that can be elevated to faster value delivery.
Increased flexibility for changes in scope with minimal cost.
Value delivered in the way the customer expects it.
If we drill down to the specifics, this could go all the way to how much time yous pend on a defect and how efficient is your CI/CD pipeline or how well defined are your Acceptance Criteria.
Per “Blogagility”. The effectiveness of the business outcome of our collaboration and alignment either builds Trust, or it destroys it.
I totally agree with that statement. That’s why I focused my story on how trustable we are in the eyes of our Business Owners.
Why Trust matters.
I used a storytelling approach to explain why it matters and tried to connect it to our own software development world.
Here comes the story:
It was one Friday night, and after a hard day at the office, I went home and met some Balkan friends. We did some things that Balkan people do, drinking some liquids that Balkans people drink and when we started seeing Balkan dictators at some point there was this voice behind me:
My son: Dad, where is the pizza.
Me: grabbed my phone, opened the app I used for home delivery and clicked on my order.
The app: Time for delivery: 18:45; Status: Preparing
Me: looks at my Ikea clock on the wall: It was 20:00.
Me: clicks on the contact us icon and enters my order id.
The app: Order not Found
Me: uses my so-amazing Czech language and checks the status on the phone
The app support: Oh, we apologize, the supplier forgot about your order, here’s your money back.
The time: You lost 40 min of your life.
The app: Your order has arrived.
So we ended up with pizza (not so warm) and money in my pocket. You could way that this is a win-win situation, but I mostly care about the experience, not about the money.
Let’s go back to the Trust.
This happened twice, so my Trust with the process of that company went down.
I really don’t care about who’s fault it is.
is it the cooks (developers) that were too slow?
or it’s the delivery guy (the CICD pipeline)?
or it’s the misalignment (program manager) between different departments?
or something completely different?
At the end of the day, the Trust in your product matters, and your product is a complex system that should work well together.
I can imagine that this problem doesn’t happen just to me, and it happens to other people, and they complain differently. Then the company receives that feedback, and they can say what their trust score as we can say as a team of teams (a Program) what our PPM is.
I saw some people nodding their heads in agreement with me this time, so I believe that it was one small step forward to explaining some of the terms in a way that makes sense.