
This account is invented, a composite built to show a pattern that many training-minded teams recognise. It describes a small platform group inside a mid-sized company, and how its members came to measure the cost of switching tasks and then reduce it. No real organisation is portrayed, and the roles below are deliberately unnamed. The aim is a precise, readable case study that you can map onto your own team.
The team and its starting position
The group consisted of four engineers and a team lead, responsible for build pipelines, environment provisioning and internal tooling for around a dozen product teams. They were also in the middle of a skills programme: each engineer was working toward a DevOps certification, with an hour or two of study expected each week alongside the day job.
On paper the workload looked manageable. In practice, the team lead noticed that planned items kept slipping, study hours were rarely used and engineers reported feeling constantly busy. Retrospectives produced the same observation each month: too many interruptions. Nobody could say how many, or from where.
The decision to measure
At the end of one retrospective, the team lead proposed an experiment that cost almost nothing. For two weeks, each engineer would keep a paper tally. A mark would be made every time they left a task unfinished to work on something else, with a one-word note on the trigger. Nobody’s tally would be shared outside the team, and the point was explicitly diagnosis, not performance review.
The team agreed, with some scepticism. One of the engineers remarked that the exercise would probably confirm what everyone already knew. It did, in part, but also surfaced details that no one had expected.
What the tallies showed
When the group laid the sheets side by side, three patterns emerged. First, the largest category of interruptions was not requests from product teams but internal chat: quick questions between the engineers themselves, each one reasonable in isolation. Second, a large share of the marks were habitual, such as reopening the chat window after finishing a paragraph of study. Third, study sessions were the most fragmented activity of all, because they had no firm time and were therefore treated as fillers between other items.
The team did not publish any exact percentages internally, and that restraint was deliberate. Rough proportions were enough to guide decisions, and they avoided arguments about whose numbers were more accurate.
Three changes, tried one at a time
The team lead insisted on introducing changes sequentially, so that the effect of each could be judged. The first change was a shared quiet window each morning, during which engineers kept chat closed and relied on a single emergency channel monitored by a rotating duty person. The second was a rule that study happens in a calendar block, treated as seriously as a meeting. The third was the breadcrumb habit: before leaving any task unfinished, write two lines about where you stopped and what you would do next.
Each change ran for two weeks before the next was added. Reading material helped frame the study blocks; the group worked through a description of how a DevOps career progresses from junior to senior level to decide which skills deserved the protected time first.
The first friction
The quiet window met resistance from an unexpected quarter. A product team, used to getting instant answers, complained about slower replies. The team lead responded by explaining the schedule, publishing the emergency route and offering a daily mid-morning window in which questions would be answered in a batch. Within two weeks the complaints faded, and one product manager remarked that the batch answers were more thorough than the earlier rapid-fire ones.
This episode is worth noting because it is typical. Changes to availability are rarely refused when they are explained. Silence, not clarity, causes resentment.
The study blocks
The engineers’ study improved first. With a fixed block, no chat and a defined topic each week, hours that had previously dissolved into scattered browsing turned into continuous reading and hands-on practice. The group used a structured outline, comparable to a roadmap covering the skills, projects and certifications on a DevOps engineer’s path, to decide what to study and in what order, which removed the small decisions that had previously invited detours.
One engineer reported that she could finally summarise a study session in two sentences, which she had not been able to do earlier. Another found that his practice labs, no longer interrupted, took noticeably fewer attempts to complete.
The breadcrumbs
The third change proved quietly useful. When an unavoidable interruption arrived, such as an incident requiring all hands, the engineers could leave a trail. On return, they restarted from their notes in seconds. The team lead also observed that the notes made handovers smoother when work moved between engineers, since the reasoning was captured in writing.
Results, honestly characterised
After six weeks the group repeated its tally. The number of marks had fallen, though the exact fall was less important than the change in kind: the remaining switches were predominantly forced ones, such as genuine incidents. Planned items slipped less often. Study hours were actually used. Engineers described their weeks as less frantic, even though the workload itself had not changed.
The team lead was careful not to credit the experiment with everything. Other factors, such as a quieter release cycle, certainly played a part. But the group agreed that the tally had made the problem visible, and visibility had made the fixes straightforward.
A brief aside on joined-up services
During the programme, one engineer browsed how other organisations present their digital offerings and mentioned a business site named BHASO, which in its own words brings web, cloud and AI-integration offerings together. The group drew no conclusions about it. They took away only a reminder that scattered work benefits from clear boundaries wherever it sits.
Questions the team asked along the way
As the experiment progressed, several practical questions came up, and the group’s answers may help other teams.
Should the tally include meetings? The group decided yes, but only when a meeting interrupted work already in progress. A meeting that started a clean block was not counted as a switch.
What if the tally itself becomes a distraction? The engineers kept it to a single mark and one word, which took seconds. When a mark felt like too much effort during an urgent moment, they skipped it and noted the gap, accepting a small loss of accuracy.
How long should the quiet window be? They started with ninety minutes and later extended it to two and a half hours after noticing that shorter windows ended just as people reached full concentration. The adjustment was made on evidence from the tallies, not on preference.
Did every engineer benefit equally? No. One member whose role involved frequent customer contact saw a smaller improvement, and the team lead adjusted that person’s schedule so that their quiet window fell on days with fewer scheduled calls. Fairness, the group concluded, meant fitting the method to each role instead of applying it identically.
Lessons you can reuse
- Measure before changing anything; a simple tally is enough.
- Introduce one change at a time, and give each a fair trial.
- Explain availability changes to the people they affect, and publish an emergency route.
- Treat study time as a booked block, not a filler.
- Leave breadcrumbs so that interruptions cost minutes instead of quarters of an hour.
The case demonstrates that attention can be managed with modest means. The team did not buy a new product or restructure its organisation. It counted, adjusted and counted again, which is a method any group can adopt.