Are retrospectives an antipattern?

Posted by Doc
Mar 20 2009

My friend Chris Matts stirred me up with the following:

One for you to think about for your Blog.

I consider retrospectives to be an anti-pattern. If you are learning great stuff in your retrospectives, it means that your communication is blocked. They are “batches” of “feedback”. I prefer single piece flow of “feedback”.

Last summer there was a big discussion of retrospectives on the Kanban list. I asked the question….. “Has anyone learnt anything as a result of a retrospective that at least one member of the group did not know about before?” So far, no examples.

Last year, my tech lead wanted to address an issue with one of his developers. He wanted to leave it to the retrospective.

If your team members are saying I’m sorry, I love you, Thank you in the moment rather than leaving it to the following day, then you may find that retrospectives are of little value.

My experience of retrospectives is that they are a place for developers to find their voice, and managers to ignore them. Managers like them because “The developers get it off their chest”.



Needless to say, I agree a bit, and mostly disagree.

Chris and I discussed this on the phone. First qualification from Chris: he is thinking about highly developed teams. Okay – I can see that with some highly developed teams, they may indeed learn to express themselves well on an ongoing basis. Of course, based on the folks I know – even the most highly developed – there are some things that are hard to deal with directly, one-on-one, or in an unstructured group environment like a team room.

Then there’s the idea of retrospectives being an antipattern. I just disagree. I suppose it could be, when used as a crutch or when turned into routine. If a team does something they call retrospectives, but every time they throw up two pieces of flipchart paper and do smiley/frowney or smiley/double-smiley or any other activity/technique every time, then I’d say they are not really using retrospectives to improve their process and team functioning. Is that an antipattern? Or a smell?

Chris and I tentatively agreed that it’s a smell.

Needless to say, I’m all for effective communication on a daily basis, and on not harboring negative feelings any longer than necessary. At the same time, I believe strongly that there’s a place for structured and facilitated activity, no matter how highly developed the team and individuals, to allow for the team to deal with those things that just don’t come out.

Finally, if done well/properly, retrospectives help a team be more effective, which means delivering better software on time and under budget. What manager wouldn’t be happy with that?

Newer :

Older :

21 Responses

  1. So, can looking back ever take you forward?

    My gut says that motion backwards will always end up stalling a meeting rather than keeping the flow.

    I think of anti-patterns as beaver dams. Quite beautiful from a distance, unless you live downstream. And in most cases, everything is downstream. Looking back to review/relive to me seems like a negative at best.

    My vote is anti-pattern.


    The other Doc…

    • Doc says:

      Interesting. Yes, even the word retrospective refers to looking backward.

      So how would you deal with a team that is still growing? What techniques or patterns would you apply to help them grow together?

      What is the forward-looking alternative to retrospectives?

    • Patrick Kua says:

      I think there is still value in looking backwards. Part of implementing change requires people to see a problem that needs solving. Without looking backwards, it’s hard to understand what impact the problem has, how people view it, and often, what the root causes were.

      More importantly doing this as a group is sometimes an essential part to gain a shared understanding of the problem and consequences. Without this, conversations break down into four different solutions as everyone perceives the problem differently.

  2. I personally think that they are an antipattern – especially when we believe that they are required. I’ve been on a good number of agile projects that have succeeded before the retrospectives book swept through the Agile Alliance orthodoxy. Back in the day, most agile teams were young. We didn’t have highly functional agile teams yet. I’ve been on one project that had deep organizational problems due to the naivete of this orthodoxy which led to deeper cultural and behavioral problems that were then only exasperated by the banality of retrospectives.

    Problems should be dealt with when they occur. Kaizen events should happen when appropriate. Facilitation should be necessary when it is. All this should happen just-in-time and with the understanding that festering problems quickly turn into pervasive rot. Keeping problems from festering is a matter of having the experience that feeds the insight that allows them to be seen at their earliest and most subtle signs. They should be dealt with decisively at that point either as organizational or personnel remediation issues, or as kaizen events if appropriate.

    The presence of retrospectives on teams is an indicator that there are looming, subtle, but pervasive problems in the team that are likely far too powerful for retrospectives to solve. Retrospectives and facilitated events aren’t a sufficiently-effective means to make agile teams as communicative as agile development requires. They might be necessary for teams that are in chaos, but the replacement of them with just-in-time handling, insightful management, and kaizen events is a sign that the team is maturing. Institutionalized retrospectives are a sign that the team, its organization, and its culture are suffering from problems that are too subtle for its directors to see.

    • Doc says:

      Thought provoking. So in effect, it seems to me that you’re saying that the tools of retrospective – facilitation, drawing out information, deciding how to move forward – should be used as needed when situations arise, rather than as a scheduled, structured event. Is that right?

      • That’s partly what I’m saying. But specific to this issue of scheduled retrospectives: Scheduled retrospectives are a strong indicator that a team’s management isn’t sufficiently perceptive enough to know exactly when actions like facilitation, drawing out information, deciding how to move forward are needed based on observable phenomena. The schedule can be an excuse to never have to really learn to see need. If the need can’t be seen, the schedule might do nothing to improve this situation, and may yet exasperate it.

      • Doc says:

        Well said. I’ve got to think about how the leader/manager learns to see need and to gain the skills that are needed. I can see that, if exposed to an effective manager, one can learn the skill of seeing the need and addressing the issue. And I suppose the same is true with facilitation.

        The problem I see is that we have way too few managers/leaders who have those skills.

  3. Twitter Comment

    U R teasing me, right? RT @athought: Blogged: Are retrospectives an antipattern? [link to post]

    – Posted using Chat Catcher

  4. Twitter Comment

    @DianaOfPortland Re: [link to post] Clearly retrospectives aren’t the only time to think or talk about improvements… (cont)

    – Posted using Chat Catcher

  5. Twitter Comment

    Are retrospectives an antipattern? [link to post] – First I thought: No. Then it made me think. Then I though: Mostly not.

    – Posted using Chat Catcher

  6. Twitter Comment

    Are retrospectives an antipattern? [link to post]

    – Posted using Chat Catcher

  7. I think there are more forces in play than just the leadership. Nonetheless, when what we’re actually doing is “interventions” but calling them “retrospectives”, it’s time to call much more into question than retrospectives colloquially allow.

  8. The real skills that managers/leaders need are to teach the team and each individual to be leaders and managers of both themselves and others.

    However one must purify ones greatest delusion first.
    Thus managers/leaders must help the team to purify their greatest delusion first.

  9. Twitter Comment

    RT @bellware: be careful of confusing retrospectives with interventions [link to post]

    – Posted using Chat Catcher

  10. Kris Kemper says:

    I think formally having a retrospective as part of the process is good because it reminds us that we should be focusing on constantly improving the process. Retrospection is a critical part of the meta-process that guide agile processes. From the standpoint of a process around a process, it’s really introspection.

    Highly communicative, enlightened teams may not need a meeting because they will be in a constant state of identifying improvements and implementing them, in which case, the retrospective won’t add much.

    For many teams and organizations, such enlightenment won’t be achieved, and they will benefit from a regular checkup on the process. Also, projects that are newly attempting an agile process should have a retrospective phase baked in the initial process blueprint, just to make sure they engage in the activity.

    I would say, that it’s a smell when you do retrospectives too infrequently, say, a month or more (if anyone can’t remember the beginning of the period, too much time has passed). It’s also generally a smell when retrospective take a long time (say, more than an hour). There are either too many people or too many problems.Retrospectives should be frequent and short.

    Somewhat unrelated to you post, I also believe that many retrospectives would benefit from being focused. Instead of the “What was Good/Bad? What should we change?” or similar general questions, a retrospective should ask specific questions. How can we deliver faster, cheaper, higher quality, better for our customers, etc?

  11. marc mcneill says:

    Why wait to do the back slapping and finger pointing to the end of an iteration in this sacred ceremony called a retrospective? Far better to do it informally and continuously, as I suggest here…

    • Doc says:

      Why not do both? I believe that there is a place for immediate response and action, and that there is also a place for calm, retrospective review and discussion.

      Why does it have to be either/or? They’re not mutually exclusive.

  12. […] Are retrospectives an antipattern? | Steven “Doc” List’s Random Musings. […]

  13. Retrospective Antipatterns…

    Are retrospectives an antipattern asks Steven. There are some interesting points raised there, and I certainly can see some antipatterns (some of which I have been guilty of, too). I don’t think it’s the retrospective per se that’s the antipattern, …

  14. […] Are retrospectives an antipattern? | The Doctor Is In (tags: agile antipatterns retrospectives team) Share and Enjoy: […]

  15. […] Anna Forss Leave a comment Go to comments Steven “Doc” List shared his and Chris Matts discussion on retrospectives. A site from Chris Matts: I consider retrospectives to be an anti-pattern. If you […]

Trackback URL for this entry

%d bloggers like this: