Phase 0 – Checking follow-through in retrospectives

You are probably familiar with the 5 phases of a retrospective, as described in Agile Retrospectives by Esther Derby and Diana Larsen. For years now I sometimes have a block before these 5 phases. I call it “Phase 0” and use it to check up on what happened to the action items the team agreed on in the last retrospective.

After all, retrospectives serve a purpose. In the long run, we want to improve and that means trying out things. If all that ever happens is talking and nothing ever changes due to retros, then why do them? Plus, teams quickly learn to resent retros in this case.
Change happens

I got the idea for Phase 0 from a team that was amazing at follow through: Each retro they added all action items and rule changes (we didn’t know about working agreements back then) to a big on-going flip chart. Each item had a “revisit”-date attached to – the date when the team thought they’d be able to judge the effect (usually 2, 4 or 6 weeks). At the beginning of each retro we would go down the list of all open items that had reached the revisit date and inspect them. Did the team do it? Did it work as intended? If yes, rule changes were made permanent and actions crossed off. If not, the items were consciously dropped or changed.

They had continuous improvement down to an art. It was a joy to facilitate their retros. They devoted a huge chunk of time to this process – 20-30 minutes out of 60. That sounds like a lot (it is!) but it worked very well for them. By the time they had analyzed the list, they usually had covered a lot of the things that bugged them.

I’ve never again seen such consistent follow-up. My Phase 0 is very bare bones compared to this: I bring the list of last retro’s agreements and ask what happened with them, boiling it down to 5 minutes.

This accomplishes several things:

  • It lets the team know that someone cares about what happens. (Whenever I remember to, I’ll also ask during the iteration – genuinely curious, not annoyingly!)
  • I can spot problems with follow-up early. And hopefully the team will notice them too

With a mature team, I’ll do this every once in a while. If I think there’s a problematic pattern, I’ll do it more often. I try my damnedest not to be accusing, but if the team consistently does very little of what they agreed to do, that’s indicative of a problem. Phase 0 lets us find this so that we can work on the lack of follow-through.

Curiously enough, I’m not the only one to come up with an extra phase before the 5 phases of lore. At least two other people have developed similar concepts: Marc Löffler and Judith Andresen. I’ve only recently heard about Marc’s ideas. I’m told it’s also something with checking follow-through but I’m relying on hear-say. I’m much more familiar with Judith’s work and her “Intro” is more elaborate than my Phase 0: She does an Intro at the beginning of every retrospective and it consists of the Agenda, restating the Vegas rule & the Prime Directive, and checking follow-through & team rules.

So, there’s at least 3 people who independently arrived at the concept of checking last retro’s agreements at the beginning of the next one. And those are just the Germans! Is anybody else doing this?

PS: Interested in retrospectives? Sign up to the Retromat newsletter to get related news and tricks!

Retromat is not meant for beginners!

Aaaargh, I recently found out that there are indeed people who just use whatever random plan Retromat spits out. Errh mah God! I was in serious denial about this, despite evidence to the contrary.

I never, ever meant for anyone to do this. The random plan was always meant as a starting point from which everyone would merrily click left and right to create a plan that fits their and their team’s needs.

That’s what I meant with the “tweak it” in “Planning your next retrospective? Get started with a random plan, tweak it, print it and share the URL”. It’s a little too subtle. To me it’s obvious that most random combinations will not work well together. It’s obvious to me, because I’ve facilitated retros before and I’m experienced. It’s not obvious for someone new to retros. So here’s a handy note to self:

What is obvious to me is not necessarily obvious to others. In fact, it will be least obvious to the people who need the clarification most.

How could I not realize this for so long? I guess I only get emails from people for whom it works. I don’t hear from those that fail with a random plan or those that “have to pick up the pieces after an inexperienced colleague unleashed a random retro on a team” (actual quote!). I’m so sorry!

I’ll try to find the time to beginner-proof Retromat ASAP. I’ve also thought about the best out-of-the-box, beginner-friendly retro plan I can come up with. It’s practically guaranteed to be better than a random plan.

To reiterate: Retromat is a great source of inspiration for people who know what they’re doing. It’s not a good place to start for people who lack the experience to know whether activities go together well.

In theory Retromat offers millions of plans for retrospectives. In practice only a fraction of these combinations work well. A random plan is highly unlikely to work out!

When you plan a retrospective with Retromat you have to make sure that you know how the results of one activity will be used in the next activity. That’s what the arrows at the sides are for: To flip through the activities for one that fits to the activities before and after it, as well as your team’s situation.

Retromat needs some experience. Please do not recommend it to beginners! At least not without fair warning. Recommend Agile Retrospectives instead and if it’s urgent, this plan. Thank you!

PS: Interested in retrospectives? Sign up to the Retromat newsletter to get related news and tricks!

Visualize remaining time with TimeTimer

How do you keep track of time when you facilitate a retrospective or other meeting? How do you make sure you all keep short timeboxes? A timer on a smart phone is one way to do it, but for me it lacks visibility. I forget the timebox and only remember it, when it’s used up.


What works beautifully are TimeTimers. With a TimeTimer you set the timebox by pulling out a red disk. As soon as you let go, the red disk slowly starts retreating back below the white parts. That way you always have a pie chart of the remaining time. Elegant, easy to use and it communicates time very effectively!



Mein rechter, rechter Platz

[English summary: Armin Schubert suggested a super nice “Set the stage” activity for Retromat that doesn’t translate well, so I present it in the original German.]

Immer wieder bekomme ich tolle Vorschläge für Retromat, die ich schweren Herzens ablehnen muss, meistens weil es bereits eine sehr ähnliche Aktivität im Retromaten gibt. Bei der folgenden Idee von Armin Schubert war der Grund, dass die Aktivität nur auf Deutsch funktioniert. Aber wozu habe ich ein Blog 😉

Hier kommt also Armins “Mein rechter, rechter Platz”: Diese Aktivität ist für den Anfang einer Retro und läuft wie folgt ab:

Die Teilnehmer sitzen im Kreis und starten mit dem bekannten Kinderreim “Mein rechter, rechter Platz ist leer, ich wünsche mir den $Name her!” mit einer entscheidenden Änderung im Text:

“Mein rechter, rechter Platz ist voll und der $Name, der ist toll!” Dann noch drei positive Eigenschaften des rechten Nachbarn und schon ist derjenige selbst dran.

Die Idee dazu ist in einer Retrospektive entstanden, weil wir einen schnellen aber positiven Einstieg gesucht haben. Das wirkte am Anfang etwas hölzern, war dann aber ein grosser Erfolg, auf dem im Nachgang immer wieder referenziert wurde.

(Falls jemand den (neuen) Kollegen rechts von sich nicht kennt, kann er gerne die anderen Anwesenden um Hilfe bitten. Hat bei uns mehrfach super funktioniert!)

Thanks for sharing with us, Armin!

PS: Interested in retrospectives? Sign up to the Retromat newsletter to get related news and tricks!

Distributed Retrospectives – Interview with Philipp

People ask me: “How do you best run a remote retrospective with a distributed team?” and I have no idea. I’ve only ever worked with co-located teams. That’s why I started to ask people who actually do run distributed retrospectives. After the interviews with Christoph and Frank I present to you:

Philipp Flenker, Product Owner from Münster

tl;dr A) Most online retro tools are bad. Just try something simple like wikis, Google Docs, etc. before wasting time with research on specialized tools. B) There are activities that don’t work remotely, e.g. anything with movement or anonymity*.

Screen Shot 2016-05-25 at 00.02.10

Full Interview

What’s the situation?

Our team consists of 5 people including me. We cover 3 time zones (3 people in Germany, 1 US east coast, 1 US west coast) and 2 languages. We speak English at work.

I’m the PO but since we don’t have a Scrum Master (don’t ask …) I facilitate the retros for our team.

Sprints used to last a month. Last month we switched to 1-week-sprints.

We meet every 6 months to see each other in person. That really helps with “individuals and interactions over processes”.

Do you prepare differently for a remote retro than a co-located one?

Yes, there are a lot of activities that don’t work in our setting:
Anything with metaphor, because of the language barrier. For instance, when using Speedboat, the US colleagues would use all the nautical terms and us Germans couldn’t follow.

Anything with movement is out.

And nothing is anonymous. This one I miss the most. Anonymity is great to have, e.g. for ESVP and we can’t really achieve that. Each input has a name attached to it.*

One activity that works well is Learning Matrix. It’s easy to learn and fits a wiki page.
About twice per year we meet in person. I plan these retros very differently! I plan for more honest and more difficult subjects. Participants tend to be too polite in remote retros. There’s very little healthy conflict.

What’s your setup?

First, everybody writes notes on their own computers. We discuss everybody’s notes and then we all paste our notes into a wiki page that I prepare in advance.

We can hear each other, but we don’t have video. We just can’t get a good video feed across 4 company VPNs. Hangouts don’t work at all in the company network. Skype is not very good. [At a former employer we had one remote guy and used Lync. That worked okay.]
The audio works most of the time. We only lose someone’s audio connection once in about 10 retros.

Our biggest challenge is the different time zones. We hold the retro at 5pm German time. One of the German guys is an early riser, so by that time he just wants to be done and head home. In California it’s 8am, in Colorado 9am. And the Colorado guy likes to sleep in. The bio rhythms don’t work out here. Concentration is difficult to maintain. The retro needs to be short, 90 minutes max.

When you’re co-located you can easily spot when someone is wool gathering. In a remote setting you can’t. Especially if you only have audio and people frequently mute themselves.
Retros are still valuable to us. Small changes are better than stagnation.

Do remote retrospectives have any advantage over co-located ones?

No, remote retros are not better than normal ones in any way. But they are better than losing a team member all together by not allowing remote work.

Any tipps for new facilitators of remote retrospectives?

Most tools suck. There are a lot of retro tools out there and I haven’t found any good ones. You just waste time trying them out to assess what they can and cannot do. The wiki page we ended up with, was better than most. Next retro we’ll try out Google Docs. In a Google Doc you can see the others’ cursors and it’s a fair guess that the cursor is where people are reading. So we’ll have more information than with the wiki page.

Thank you very much, Philipp!

You can follow Philipp on Twitter!

* Professional Scrum Trainer Jason Knight has found a workaround to achieve anonymity: Create a Google doc (or spreadsheet) with a private link that anyone can use to edit it. The team members open the link in a browser in incognito mode.

PS: Interested in retrospectives? Sign up to the Retromat newsletter to get related news and tricks!

Distributed Retrospectives – Interview with Frank

People ask me: “How do you best run a remote retrospective with a distributed team?” and I have no idea. I’ve only ever worked with co-located teams. That’s why I started to ask people who actually run distributed retrospectives. After the initial interview with Christoph, I present to you:

Frank Deberle, Developer/Coordinator, working in Mainz

frank-deberle tl;dr 1) Don’t fret. Remote retrospectives are not as bad as it may seem. Just try to run one and you’ll see. 2) works well for us

Full Interview What’s the situation?

I’ve been facilitating retrospectives for 2 years now. For the last 6 months these retrospectives have been remote – every 3 weeks, 60-90 minutes. There’s 3 of us in Mainz and 2 in Stuttgart. We all know each other face to face too, which makes it easier to work together remotely. Two of the team member are immigrants, but they both speak German, so the language barrier is low.

We all work for an agency that in turn works in a big project for another company. I coordinate everyone from our side working on that project and I facilitate the retrospectives in that capacity. That is to say, we are probably a special case, because our retrospective is not the whole team working on that project, but only with the people from our agency, working on that project. They are part of two different Scrum teams (both working on the same project). Phew, that was a little complicated.

Anyway, at first we were all together in Mainz but then we started an office in Stuttgart and suddenly we were a distributed team. In the beginning I was convinced that retrospectives couldn’t possibly work if we were not all in the same room. I was kind of waiting / hoping for the perfect solution to come along. But then we realized we needed to do retrospectives again. We tried it and it just worked. There was no need for me to be so worried about it! Of course, it’s different, but at least you get to do a retrospective at all!

Remote retrospectives? At least you get to do a retrospective at all!
– Frank Deberle (@fdeberle)

How is a remote retrospective different from a co-located one?

It’s very hard to feel everyone’s vibe. In a co-located retrospectives it’s much easier to pick up nuances in voice and mimic and thus read the team’s general mood accurately.

Also everything seems to take a little longer than when co-located. Some part of it is the occasional lag or that Mac microphones’ sensitivity settings spontaneously self-lower. The bigger part is that it seems more chewy in general. Because feedback is less direct, people tend to explain in greater detail. And all of that together leads to slightly longer retrospectives.

What’s your setup? 

We use video chat. It’s super easy to set up. Once you’ve installed a Chrome plugin all you have to do is send around a link, no special code or password required. The quality is well enough, certainly better than Skype. We’ve never tried Hangouts.

Ideally we use 1 laptop in Stuttgart (for 2 people), and 2 laptops in Mainz (1 for the whiteboard, 1 for 3 people).

Sometimes we enhance this setup with our agency’s bluetooth speaker and standing mic. That improves the sound quality, but we only use it, if it’s already set up.

Do you prepare differently for a remote retro than a co-located one?

Not really. That is, I don’t plan differently, but I noticed that in the distributed retrospectives we tend to do fewer activities. I think it’s because of the slower pace and more explicit explanations I mentioned before. 

Any tipps for new facilitators of remote retrospectives?

Just try it out! It’s really not that big a deal! Oh, and vary what you do. Otherwise it’ll get boring soon. Retromat is cool for that! Okay, that last one was more of a general tipp 😉

Thank you very much, Frank!

PS: Interested in retrospectives? Sign up to the Retromat newsletter to get related news and tricks!

Distributed Retrospectives – Interview with Christoph

Because of Retromat people assume I’m knowedgable about retrospectives and ask me questions. Which is fine! I am knowedgable about retros and I am happy to answer questions. Unfortunately I could never answer the single most asked question: How do you best run a remote retrospective with a distributed team? I have been fortunate enough to only have worked with co-located teams, so I have no idea. Of course that’s not helpful for the people who ask.

In order to remedy my ignorance in that area I decided to ask people who actually run distributed retrospectives to share their insights with me (and thereby you :))

Meet Christoph Sperle, Scrummaster from Basel

I’m very grateful that Christoph agreed to answer my questions. Without further ado 

tl;dr 1) You don’t need to plan a remote retro differently from a co-located one. 2) WebEx works well. 3) Use the inbuilt audio.

Full Interview

What’s the situation?

I have facilitated retrospectives for the last 18 months and remote retrospectives for 9 months.

The remote team consists of several people in Basel and 3 remote team members that are in Poland, UK and the US respectively. All remote people have at one point been to Basel and met the Basel team, but not the other “remotes”.

The different time zones (US) and the language barrier add additional layers of difficulty.

Outside of the retrospective the team communicates mostly by chat.

Do you prepare differently for a remote retro than a co-located one?

I used to plan them differently, for example doing brainstorming and a Learning Matrix in Trello. But this approach completely killed the vibe of a retrospective. Everybody was just staring into their computers. No real exchange.

Today I plan remote retrospectives the same way I plan co-located ones.

What’s your setup?

We use WebEx to have all remote team members with us in the room on a big screen. The remotes see our whiteboard on their screens. We do all activities as you would normally do. Whenever the remotes share their stickies, I write them down and put the on the board. That’s the most stressful part for me.

We tried lots of different things with audio. The telephone was not a good option. Low quality and it was confusing that the voice did not come from the screen. Now we just use the WebEx audio and the speakers from the TV screen that shows our colleagues. That works well for us.

Do remote retrospectives have any advantage over co-located ones?

None that I can think of. Technology is always hassle and makes it more difficult to address things. Our team members are familiar with each other by now. For fresh teams it’s harder to speak their minds.

 Thank you very much, Christoph!

PS: Interested in retrospectives? Sign up to the Retromat newsletter to get related news and tricks!

118 activities in Retromat

Loads of great new activities in Retromat! Check them out:

100 Activities in Retromat!

Rejoice with me for Retromat now features 100 activities!

Screen Shot 2015-04-07 at 9.33.04 PM

It’s come a long way since its humble launch with 16 activities 3 years ago (May 2012). It started with activities from Diana Larsen and Esther Derby‘s “Agile Retrospectives” and in an attempt to come full circle activity #100 is by Diana Larsen again.

Huge thanks to everybody who made that happen: People who thought of activities, suggested them, translated, sent photos, supported the print version, spread the word, … All you of you: Thank you!