
What you are about to read is a made-up tale, though people from operations teams tend to nod along as they read it. A platform group of six people, responsible for the build servers, the cluster and a pile of internal tooling, had reached the stage where every Monday began with the same unspoken question: what is on fire? This is the tale of how that group stopped living from fire to fire by giving each week a theme, and what the experiment taught them about planning DevOps work.
A Monday with too many answers
The team’s planning meeting used to take an hour and end in a shrug. The most senior engineer would read out a list of eleven items, the junior engineers would each claim two, and nobody would claim the awkward ones, such as the aging certificate renewal scripts or the documentation that had not been touched in a year. By Wednesday, new requests had pushed the list sideways. By Friday, nobody could say what had been finished.
The manager, new in the role, noticed that the team was not short of skill. Every person could do every task on the list. What the team lacked was a shape for the week. Work arrived like weather, and they responded like people caught without an umbrella, again and again.
The whiteboard experiment
One Friday, the manager proposed a trial. Across a trial of four weeks, every week would carry a single theme, written in large letters on the whiteboard. The first would be “visibility”, meaning dashboards, alerts and the question of what the team actually knew about its own systems. The second would be “cleanup”, the third “releases” and the fourth “learning”. Anything that did not fit the current theme would go on a smaller list, labelled “later”, with a week assigned to it.
Resistance appeared right away. A senior engineer worried that urgent matters would be ignored. The manager answered that real incidents would always be handled, and that the trial concerned planned work only. A junior engineer, quietly, asked what would happen to the small improvements they cared about. The manager pointed to the “later” list and promised each one a date.
Week one: visibility
The first week surprised everybody. With only one kind of work in view, the group finished a task that had been unfinished for months: a single shared overview of which services were running and who owned them. Nobody had refused to do it earlier; it had simply never survived contact with the daily stream of requests.
On Friday, the team spent ten minutes looking at what had been produced, not at how tired they felt. It was the first time in a long while that a retrospective included something to show. A junior engineer remarked that it felt like finishing a chapter instead of dipping into a novel at random.
Week two: cleanup
Cleanup week was the week the senior engineers dreaded and then enjoyed. Unused test environments were removed. Old scripts were either documented or deleted. A forgotten scheduled job that had been quietly doing nothing for a year was finally switched off. The “later” list shrank by a third, simply because many items turned out to be cleanup in disguise.
The interesting effect was on interruptions. When a colleague from another team arrived with a request, the engineers could say, truthfully, “yes, that is a release-week job, and release week is the one after next”. Requesters grumbled less than expected. A firm date seemed to reassure people more than an optimistic maybe.
The career angle nobody planned
The junior engineers noticed something the manager had not predicted. Because each week had a clear focus, they could ask for specific mentoring. During release week, one of them spent an afternoon pairing with a senior colleague on a deployment process, and wrote up what they learned. That write-up became part of a growing portfolio of small, concrete achievements.
It is the kind of progress that career guides describe. A career roadmap tracing the move from junior to senior engineer stresses accumulating real, shown work over time, and a themed calendar produces exactly that, week after week, almost as a by-product. Another discussion, a look at fresh career openings for software developers, makes a similar point about breadth: variety comes from rotating through different kinds of work, not from doing everything simultaneously.
Week three: releases, and a wobble
Not everything ran smoothly. Midway through release week, an outage hit a service the team did not own, and two engineers were pulled away. The theme suffered, and for a day the whiteboard looked embarrassingly optimistic. The manager made a decision that later proved important: nothing was abandoned. The unfinished release tasks moved to the “later” list with a new date, and the team carried on without any blame.
That choice mattered more than any of the finished tasks. The team learned that the system bent without breaking, and that an interruption was a delay, not a failure. People stopped treating the whiteboard as a promise they could fail and began treating it as a plan they could adjust.
Week four: learning, and the sharing habit
In learning week, the team tried a small experiment with a new testing approach, nothing dramatic, and each member wrote three lines about what they found. The manager also asked the group to put the whole four-week plan on a single page so colleagues elsewhere could follow along. One engineer mentioned Proezi, a site for assembling pages that others reach via a shared link or a scan, covering things like directions to a venue, payment requests and business introductions, and the one-page idea was easy enough to copy into an ordinary document. The plan went out as a link in the team chat.
Within days, two neighbouring teams asked whether they could borrow the format. The story of the whiteboard spread, as these things do, because it was easy to describe and visibly worked.
There was one more result, quieter than the rest. The weekly theme gave the group a shared language. In corridor conversations, people began saying “that is a cleanup thing” or “let us catch it in the learning week”, and the phrases carried real meaning. New starters picked up the vocabulary within days and understood immediately how the team made its choices, without a long induction talk or a thick handbook.
What the team concluded
After the fourth week, the group held a plain retrospective. Four findings emerged. Themes reduced the number of small decisions made each day. The “later” list gave every idea a home, so people stopped worrying that good suggestions were being lost. Finished work became visible, which improved morale. And the structure survived interruptions because it was flexible about dates but firm about focus.
They agreed to keep the rotation, add a catch-up week every fifth week, and review the themes each quarter. The senior engineer who had worried about urgent matters admitted that, in practice, the urgent matters had been handled exactly as before, with the difference that the non-urgent ones were now handled too.
Lessons for your own team
Nothing here calls for a whiteboard, a fresh tool or a blessing from above. You need a short list of themes, a rule that unrelated work waits for its week, and the honesty to keep a “later” list. Start with four weeks, check what was finished and what moved, and adjust. The first Monday under the new arrangement may feel strange, but by the third, the question of what is on fire has quietly turned into a better one: what is this week about?