The team at Harbourlight Flying Club, a made-up school on the edge of a regional airfield, had a problem that many small flight schools recognise. Enrolment was growing, the schedule was creaking, and every possible fix looked expensive. Priya, the chief flight instructor, wanted to expand the fleet. Tomasz, the operations manager, wanted to rebuild the booking process. Ana, who handled student accounts, wanted to change how students paid. Each plan was sensible. Together they would have committed the school to a lot of money and a lot of change in a single season. The account below traces how the team ranked the plans by how easily each could be reversed, and what that ordering did for training at the field.
The meeting that changed the order of work
At the end of a long Friday, Priya drew a line down the whiteboard. On the left she wrote “can undo this month.” On the right she wrote “cannot undo this year.” She asked each person to place their proposal. The results surprised everyone. Ana’s payment change belonged on the left, because it could be switched back by the next billing cycle. Tomasz’s booking redesign was mostly on the left, apart from a data migration that needed care. Priya’s fleet expansion sat firmly on the right, since a purchase or long lease could not be reversed once signed.
The conclusion was not that the fleet expansion was a bad idea. It was that it should be the last decision, taken with the evidence the other two changes would produce. That sequence became the school’s working rule: easy-to-reverse moves go first and generate the facts; hard-to-reverse moves wait for those facts.
Change one: how students pay
Harbourlight had been selling large prepaid blocks of lessons. Students liked the discount; the school liked the cash flow; but several students had paused training for work or family reasons and wanted refunds that the terms did not clearly allow. Ana proposed offering smaller blocks and a pay-as-you-fly option alongside the existing packages. If the experiment failed, the school could withdraw the new options at the next price review. It did not fail. Students who were unsure about their long-term plans chose the smaller blocks, and several of them later moved to larger ones once they felt settled. The school learned something useful about how its students actually behaved, at almost no risk.
Students comparing options at other schools often ask how to keep costs under control, and the school linked its new page to an existing read on approaches to trimming the cost of flight lessons. Staff noticed that students who had read it asked sharper questions about terms.
Change two: the booking process
Tomasz’s frustration was a paper diary and a shared spreadsheet that regularly disagreed with each other. He proposed one new scheduling tool. Rather than switching everything overnight, he ran the new tool for a single aircraft and a single instructor’s students, keeping the old diary as a backup for several weeks. When a problem appeared, such as a confusing cancellation screen, they fixed it without anyone being stranded. When the tool proved reliable, they added the next aircraft. At no point did the school depend on something it had not yet tested.
Tomasz also made one decision that he later called the most valuable of the project: he exported all booking data in a plain format every week and stored copies. If the tool ever failed or the provider changed its terms, the school could move elsewhere. The cost was ten minutes a week, and it made the whole change reversible.
Change three: the fleet
By the time the team returned to Priya’s expansion plan, they had several months of data. They knew which hours were consistently overbooked and which were quiet. They knew that a fair number of students were asking for a different aircraft type for later training. The article on how the type of aircraft shapes training costs helped them frame the discussion honestly, since the cheapest aircraft to run is not always the one that suits every stage of training.
Even then, they did not leap to a purchase. They first arranged access to an additional aircraft on a short-term basis, which let them see whether demand held up across a full season. Only when the demand proved real did they discuss a longer commitment. Priya noted that the short-term arrangement cost slightly more per hour but prevented a much larger mistake, namely owning an aircraft that suited a month of demand rather than a year of it.
The objections along the way
Not everyone was convinced at first. Priya worried that waiting would lose momentum, because a competing school nearby was advertising a new aircraft. Ana worried that offering smaller blocks would reduce income. Tomasz worried that running two systems at once would double his workload.
Each worry was real, and each was handled by shrinking the question. For momentum, the team set a firm review date three months out, so waiting had an end. For income, Ana tracked revenue weekly and agreed in advance the level at which she would withdraw the experiment; it was never reached. For workload, Tomasz limited the parallel period to a few weeks and wrote a short list of what had to be true before the old diary could retire.
Notice that none of these answers involved a promise to be right. They involved a plan for what to do if the team turned out to be wrong, which is a much more dependable foundation. Teams that can describe their retreat calmly in advance tend to move faster, because nobody is afraid of being trapped. At Harbourlight, the whiteboard with its two columns stayed on the wall for months afterwards, and new proposals were still placed on it before anyone argued about their merits.
What the team learned
- Sorting proposals by how easily they can be undone changed the order of work without cancelling anything.
- Small, reversible changes produce evidence that makes the big decisions easier and safer.
- A backup route, such as an exported record or a retained old process, turns a risky change into a manageable one.
- Some decisions really are one-way, and those deserve patience, more data and a second look at the contract.
A parallel from another industry
Tomasz’s brother works in software, and he described the same approach in his own field: staged releases, easy rollbacks and extra caution for irreversible actions. He mentioned that organisations such as DevOpsSchool Netherlands, an Amsterdam provider of DevOps, SRE and cloud consulting and training teach engineering teams to work this way. The flying club had never thought of itself as an operations team, but the resemblance was hard to miss, and the vocabulary of staged change and fast fallback proved handy in their own meetings.
Where Harbourlight is now
A year later, the school has an extra aircraft, a scheduling tool the staff trust and a payment structure that fits different kinds of students. None of those results came from a single bold move. Each came from a series of steps, most of which could have been reversed if the evidence had gone the other way. Students noticed the difference mainly as fewer double-booked aircraft and clearer terms. Instructors noticed it as a calmer Friday afternoon.
For schools and students alike, the message is simple. Treat training decisions, whether they concern fees, scheduling or equipment, as a mix of two kinds: those you can walk back and those you cannot. Make the first kind quickly and learn from them. Make the second kind slowly, once you have something better than a hunch. Students planning their own route into the air may also find the scholarship overview for flight training students useful, since funding is one of the decisions where keeping options open pays off most.

