I have a folder on my phone called “Retros.” It’s just photos of whiteboards. I started taking them six months ago because something felt off, and I couldn’t put my finger on it.
Last Tuesday I scrolled back through them and found the exact same lime-green sticky note in five consecutive photos. “Fix flaky integration tests.” Same handwriting — Sarah’s neat little all-caps. Every two weeks she dutifully rewrote it, because the old one had been taken down, and nobody remembered it existed.
That’s when the word graveyard popped into my head, and it wouldn’t leave.
The thing is, Sarah wasn’t lazy. None of us were. Our retros looked great. We had the columns, the dot-voting, the little round stickers with frowny faces. Someone always volunteered to “own” the action item. Sarah would write her name next to it and give a small smile that said I know this is hopeless but I’ll carry it anyway. Then we’d close the retro and sprint headfirst into the next two weeks of feature work, where the flaky tests would fail again, and someone would sigh and hit rerun, and nobody would mention the lime-green sticky note until the next retro rolled around and we pretended it was a new problem.
We weren’t improving. We were performing improvement.
The lowest point came during a retro where a junior developer — two months in, still too new to know which things you’re supposed to say and which you’re supposed to swallow — asked, “Why do we write down things we never fix?” The room went quiet. I watched our scrum master’s face cycle through three expressions in half a second. Then someone said, “That’s a great point, let’s make sure we circle back on it next time.” I wanted to scream.
That night I couldn’t sleep. Not because of the question, but because of the silence that followed it. That silence was the sound of a team that had stopped believing its own rituals. We’d turned retros into a pressure-release valve — a safe space to vent, nod, and walk away feeling mildly better, without ever having to solve anything. And the cost wasn’t just the un-fixed bugs. The cost was that every one of us was learning, every two weeks, that raising a problem was the end of the conversation, not the beginning.
So I did something awkward. At the next retro, I brought photos of the last five boards printed out on actual paper. I spread them across the table like a detective laying out evidence. “Look,” I said. “This exact sticky note appears in every single one. Sarah’s been volunteering to fix it since March. It’s July. Nobody’s given her a single hour to do it. The action item column isn’t a to-do list. It’s a cemetery.”
There was a long pause. Then Sarah laughed — not a happy laugh, the kind of laugh you make when someone finally says the thing everyone’s been thinking. “I stopped writing it in May,” she said. “I just copy it from the photo on my phone because I’m embarrassed to admit it’s still broken.”
We made a rule that day, scribbled on a fresh sticky note with a Sharpie so it would stick: If a problem shows up in three retros, it stops being an action item and becomes a blocker. No new feature work until it’s dead or downgraded to a conscious decision to ignore it forever. The flaky tests got fixed in four days. Not because we suddenly knew how to fix them, but because we finally gave ourselves permission to treat improvement as real work.
I still take photos of our retros. But now I take them because the board actually changes. And when I see a sticky note that’s been there too long, I don’t feel dread. I feel like it’s about to meet its end.
Top comments (0)