Case: the sponsor changes priorities every week

When a sponsor asks for a different new thing every week, the requests are rarely the point. Two different ideas in two weeks usually point to one worry, and each idea is a guess at how to fix it.

At Klasna, Leo asked for a mobile preview (families live on their phones) and then group classes (cheaper for families). He says both help reach 120. That's the clue: he is worried about reaching 120 learners by 28 February, the result criterion he signed.

Naming the worry changes the conversation. Instead of arguing about two features, you can answer it with what the project already has: the waiting list review, the landing page, the January ads. And when you do discuss the features, you judge each one by the same question: does it move the 120 more than it risks the 1 February date?

Assuming bad faith (he doesn't trust the team, he forgot the brief) closes that door.

A lead who answers a sponsor's idea with "no" is taking a decision that isn't theirs. Adding to the scope is the sponsor's call, and the brief says Leo decides changes in writing within two working days. The lead's job is to put the cost in front of him.

Each question about a new request has its own owner:

• Do we add it? Leo, the sponsor, in writing.
• How many days would it take? The owner of that work: Danylo for the preview, Vira for groups.
• Does it meet the quality bar? Vira, the expert: groups need their own pacing and their own demo standard.
• What does it do to the money? Hanna: 60,000 for the preview isn't in any line, and the contingency reserve is for named risks, not new features.
• Bringing it together on one page: the lead.

When the sponsor goes straight to an owner, these roles blur and work starts before anyone has decided.

The shared buffer, held by the lead, is meant to absorb tasks on the chain that run long. It is not meant to fund work nobody approved, yet that is exactly where a sponsor's direct requests drain it.

At Klasna, after the 23 December status, 8 of the 11 buffer days were left. Leo then asked Danylo directly for a mobile preview, and Danylo spent two days on mock-ups. His team also owns the dry run and booking, both on the chain, so those days came out of the buffer: 8 − 2 = 6. Nobody decided that; it just happened.

Make such costs visible in days against the buffer and, where relevant, money against the ceiling and the reserves. A cost the sponsor can see is a trade-off he can choose; a cost hidden in the plan is a surprise in January.

Leo messaged Danylo directly, and Danylo spent two days on mock-ups for the preview. Those days came out of the shared buffer, but nobody decided that: it just happened.

The response has three parts. Pause without blame: Danylo did what the CEO asked; telling him off teaches the team to hide work next time. Record the cost: two buffer days, visible in the next status. Fix the process: ask Leo for one channel, so new requests come to the lead in writing and their cost is shown before anyone starts.

Two tempting moves make it worse. Letting Danylo continue "because the CEO asked" lets the buffer drain on a feature nobody approved. Quietly adding two days to every remaining task brings padding back and hides what happened. Vira, by contrast, hadn't started on groups: she asked first, which is exactly the habit the channel makes normal.

A sponsor can only decide a change well if its price is on one page, in the two currencies the project runs on.

Days, against the buffer. Danylo's rough estimate for the preview is 8 to 12 working days of his team, and the same team owns the dry run and booking, both on the chain to opening. With 6 buffer days left, even the low end eats more than the whole buffer unless something else moves.

Money, against the ceiling and the reserves. About 60,000 UAH for a contractor isn't in any line. The contingency reserve is for named risks, not new features, so new scope needs its own money decision, with Hanna copied.

Group classes have a price too: 10 to 15 days of Vira's time in January, exactly when she onboards tutors.

Leave out what isn't a cost, such as how good the mock-ups look, and don't pretend a reserve covers new scope.

A steady project isn't one where nothing changes; it's one where every change is a visible decision. The lead's role isn't to turn ideas down but to lay out the options for each: add it with a trade-off (what leaves the scope or which reserve or date absorbs it), do it later, or drop it. The sponsor picks, in writing.

Leo pushes back: "so everything I suggest is a change request now? That sounds like bureaucracy." The answer is in the size of the process: one short message per idea, not a committee. He gets the cost and the options, decides in two days, and the team stops guessing which priority is current.

A lead who simply refuses ends up fighting the sponsor and usually losing. A lead who simply accepts lets the scope grow until the date breaks. Showing the price is the path between them.

Not every good idea belongs in this launch. Group classes might well help Klasna later, but they need their own pacing, their own demo standard and 10 to 15 days of Vira's time in January, exactly when she onboards tutors. Added now, they put the 1 February date at risk.

The answer doesn't have to be yes or no. "Not now", with a date, is a real answer: keep a short list of later ideas, and review it at a fixed point, here the results check on 1 March. A group format could then become its own small project after opening, with its own brief.

Two things make "not now" credible. It is written down with the review date, so the idea doesn't come back every week. And nothing is promised to families in the meantime, as Vira insists: a "coming soon" line on the landing page would turn a postponed idea into an obligation.

The lightest process that keeps a project steady is one channel for new ideas: they come to the lead, in writing, and each gets one short message back with its cost in days and money and the options. The sponsor decides within the agreed two working days. No board, no forms, no meeting.

Sold that way, it answers Leo's fear of bureaucracy with what he gains: he keeps full say over the scope, sees the price before anyone starts, and the team stops guessing which priority is current, so his decisions actually stick.

The alternatives fail on one side or the other. A change board is the ceremony he fears. Letting him message the team directly is how two buffer days vanished without a decision. Asking him to stop suggesting things shuts out the person who owns why the project exists.

How it fits together: roles, a stakeholder map, a written weekly rhythm and one channel for changes keep the people side of the launch steady. The next module turns to risk and quality, starting with a register of 40 items that misses the real risks.