Case: Veloria is 38% of Teams revenue

A headline like "Veloria is 38 % of revenue" is the start of an analysis, not its result. First check it against the data, and check what period it describes.

In the revenue file, March has two rows: Veloria with 300 lessons and 240,000 UAH, and the seven small clients with 490 lessons and 392,000. Total: 790 lessons, 632,000 UAH. Veloria's share: 240,000 ÷ 632,000 = 37.97 %.

Because every client pays the same 800 a lesson, the share of lessons and the share of revenue are the same number: 300 ÷ 790 gives 37.97 % too. And because every lesson brings the same 339 of contribution, it is also Veloria's share of contribution. With discounts or different service costs those three would differ, and you would compute each.

So Leo's 38 % is right, for March. That last part is what the rest of the case is about.

A share tells you how big a client is; it doesn't tell you what its loss would do. For that, turn it into money and take it all the way down to what is left for the team.

March contribution: 790 × 339 = 267,810 UAH. Veloria's part: 300 × 339 = 101,700. Teams' own payroll is about 150,000 a month (Hanna's figure, taken as given).

• With Veloria: 267,810 − 150,000 = 117,810 left after payroll.
• Without Veloria: the small clients' 490 × 339 = 166,110, minus 150,000 = 16,110.

That is the number Leo actually cares about. A client at 38 % of revenue takes away about seven eighths of what remains after paying the team. Shares understate this, because fixed costs don't shrink when a client leaves.

The common slip is to stop at contribution (166,110 without Veloria looks comfortable) or to subtract Veloria's revenue instead of its contribution.

37.97 % describes one month. Before acting on it, look at the months ahead in the plan.

In April Cobalt starts with 450 lessons. Teams then has 300 + 450 + 490 = 1,240 lessons a month. Veloria falls to 300 ÷ 1,240 = 24.19 %, and Cobalt is 450 ÷ 1,240 = 36.29 %: from April, Cobalt is the biggest client, not Veloria. Nothing about Veloria changed; the total did.

That is the snapshot trap: judging concentration from one period, usually the one that produced the headline. The question "is Veloria too big?" can quietly become the wrong question, because by next month another client is bigger.

So always read a share with the plan beside it, month by month, and ask which clients carry Teams over the year, not which one carried it in March. Whether the plan holds is a separate question: Cobalt's April start is contracted; later starts may not be.

Follow the plan further. From June, if Orchid signs, Teams has 1,640 lessons a month: Veloria 300, Cobalt 450, Orchid 400 and the small clients 490. Each big client's share falls: Veloria 18.29 %, Cobalt 27.44 %, Orchid 24.39 %.

But add them up: (300 + 450 + 400) ÷ 1,640 = 1,150 ÷ 1,640 = 70.12 %. Three clients carry more than two thirds of Teams. Individual shares fall while the weight of the big three together grows.

So the honest answer to Leo is not "Veloria is too big". It is "a few big clients carry most of Teams, and Veloria is one of them". That shifts the remedy from one account to the whole base.

And mark what is plan and what is fact. Orchid's contract is still with its lawyers. Without it, June has 1,240 lessons, and Cobalt stays at 36.29 % all year.

Once shares start moving, two misreadings are tempting.

"The risk shrinks as Veloria's share falls." The share falls; the loss doesn't. Losing Veloria in June costs the same 300 × 339 = 101,700 UAH of contribution a month as in March. A smaller percentage makes a fixed risk look smaller.

"June is what will happen." The June shares depend on Orchid signing, and its contract is still with its lawyers. A plan row is a plan, not a fact.

What the data does support: from April Cobalt is a bigger share of lessons than Veloria (36.29 % against 24.19 %); over 2027 as planned Veloria is 3,000 of 15,730 lessons, about a fifth (19.07 %), and Cobalt 25.75 %.

And look at cash, not only lessons. Veloria's first invoice, 720,000 UAH for March to May, goes out on 31 May and is due on 30 July. Until then Teams teaches Veloria on credit, which is worth knowing before anyone promises it longer terms.

Concentration has a second side: the small clients. They are what spreads the risk. Lose them, and every big client's share goes up, not down.

So check the clients file for contracts ending within 120 days of today, the point where the renewal calendar starts. On 1 April 2027 two qualify: the law firm (30 June, 90 days, 80 lessons a month) and the marketing agency (31 May, 60 days, 60 lessons). Together: 140 lessons and 140 × 339 = 47,460 UAH of contribution a month.

The agency also shows warning signals: attendance 54 % and payments in 81 days on average. It is both the more urgent renewal and the more fragile one, so it comes first: a conversation this week to find out what's getting in the way. The law firm, with healthy attendance, gets its year review this month.

When the big accounts dominate the attention, the small renewals are the ones that slip, and each one lost makes the dependence on the big three worse.

Pavlo asks whether Veloria's customs brokerage team could start in September, about 100 lessons a month. With a concentration worry on the table, refusing can feel prudent. It isn't.

The expansion adds 100 × 339 = 33,900 UAH of contribution a month. In September, with Orchid signed, Teams would have 1,740 lessons: Veloria 400 ÷ 1,740 = 22.99 %, Cobalt 450 ÷ 1,740 = 25.86 %. Veloria grows, and still stays below Cobalt.

Turning down a good client's growth to fix a ratio loses margin and goodwill and does nothing for the rest of the base. Growth with a big client is fine while the rest grows too: new clients in other segments, more lessons with the small ones, the two renewals kept.

Treat the expansion as the small new sale it is: ask who approves the new team's budget, and make sure the night-shift complaint is closed before you put a proposal on the table.

Marta's working rule: no client above a quarter of Teams' monthly lessons. The quarter is her choice, not a law, but it turns a vague worry into a number you can plan towards.

Apply it to the biggest client, Cobalt, at 450 lessons. For Cobalt to be at most a quarter, Teams needs at least 450 × 4 = 1,800 lessons a month. With Orchid signed, June has 1,640, so Teams needs 160 more lessons a month from other clients by June. If Orchid falls through, June has 1,240, and the gap is 560.

The rule is met by growing the base, never by cutting a good client: new clients in other segments, the small clients' renewals and expansions.

How it fits together: a headline share is checked, turned into money, read against the plan, and answered with growth elsewhere and a sturdier account. That closes the course: from mapping a buying committee to keeping key accounts and handing selling to a team, every step ends in a decision someone else can act on.