Scrum and Anonymous Retros
Scrum is often treated as a religion. I’m reminded of a telling comment made by a professional Scrum consultant. When asked about whether or not some section of the Scrum Guide shed light on a hypothetical under discussion, this consultant commented something to the effect that “It might be in the Scrum Guide, I’m not sure; I don’t read it as much as I should…I really should read it more often.” It was the sort of slightly ashamed comment regularly made by Midwest kids in church youth groups about how often they read their Bibles. “I don’t read it as much as I should…I really should read it more often.”
It’s pretty clear to me that Scrum isn’t a magic bullet for your engineering org’s problems. But it’s a decent place to start if 1) you don’t already have a good philosophy for running an engineering org and 2) you resist the urge to get religious about Scrum. You can resist the religious impulse by making it part of your engineering culture to think about problems in your processes, form hypotheses about what might be better, conduct experiments based on those hypotheses, and reflect on how those experiments turned out.
To specifics…
The problem I’m observing is that we don’t have a lot of trust in our engineering org right now. There’s not necessarily a dramatic reason for this. We have a lot of junior developers. Junior developers have less confidence and are less likely to take the risk of being vulnerable (i.e. express opinions that others might disagree with). Vulnerability builds trust. Also, we just went through a reorganization. We naturally have less trust with people we don’t know.
How to fix this problem and build trust?
Sprint retrospectives can be a great place to build trust at the team level. You have lively conversations around potentially controversial (but often insightful!) cards added to the retro board. But what if you’re not getting any cards submitted at retro? That’s our current situation. We don’t get many cards, and the ones we have usually play it pretty safe.
My hypothesis is that our lack of trust and lack of retro cards are connected as a chicken-and-egg problem. We don’t have the trust needed for good retros, and we can’t have good retros without trust.
The experiment I want to conduct is to make retro cards fully anonymous. I guess that might seem counter-intuitive: doesn’t anonymity go against trust? We should feel comfortable owning our opinions in front of others! Well, trust is a great value to have, but it doesn’t just appear from scratch. You have to build it one person at a time. Until you have that trust, you probably won’t have people eager to put up their observations on a retro board.
I’m hoping “anonymous ideation”, as the jargon goes, will have a flywheel effect. In a low-trust environment, it will allow the team to offer observations and engage in lively discussions without putting themselves at risk. Those discussions, if we can keep good faith in the best intentions behind the observations, should lead to more trust. Eventually we should be able to turn off the anonymous mode, like taking off training wheels, but the flywheel of trust will keep spinning.
My hypothesis will be falsified if this doesn’t lead to more cards during retros, or if it leads to negativity and destructive criticism during retros. I don’t think the latter is very likely just due to basic professionalism in our organization, but it’s worth keeping an eye on.
We’ll see how it goes.