Overtrading: why fast growth eats cash

Leo's argument: every family pays back its 4,000 UAH in five months, so growth pays for itself. The first half is true. To see why the second half doesn't follow, start with one family.

Hanna averaged past cohorts into a planning curve: what one acquired family leaves in each month of its life. It starts at 1,200 UAH and falls as families leave: 1,040, 900, 800, 720, 650, 590, 540, 490, 450, 410, 380.

The CAC of 4,000 UAH leaves in the family's first month, all at once. Contribution trickles back. Add the curve until it reaches CAC:

1,200 → 2,240 → 3,140 → 3,940 (month 4, still short) → 4,660 in month 5.

So one family pays back in month 5. Over twelve months it leaves 8,170 UAH, about twice its CAC.

The timing is lopsided: the cost comes in one month, the money back over many. That asymmetry is harmless for a single family and decisive for a growth plan. The slip is to stop at month 4 because 3,940 is "almost" 4,000.

A growth plan is built by stacking cohorts. Each month you add a new cohort and move every older one one step along the curve.

For month m of the plan:

• contribution = Σ over all cohorts of (families in the cohort × curve value for the cohort's month of life);
• ad spend = new families that month × CAC;
• net cash = contribution − ad spend;
• cumulative cash = last month's cumulative + net cash.

Steady plan, month 2: the January cohort is in its second month (60 × 1,040 = 62,400) and the new one in its first (60 × 1,200 = 72,000). Contribution 134,400 UAH; ad spend 240,000; net −105,600; cumulative −168,000 − 105,600 = −273,600.

In month 1 there is only one cohort: 72,000 − 240,000 = −168,000.

The table in this card repeats this for twelve months and two plans. The slip is to give every cohort the first-month value (1,200), as if all families were new: that overstates contribution once families start leaving.

Follow the cumulative cash of a growth plan month by month: the running total of contribution minus ad spend. It falls while new cohorts cost more than all cohorts bring back, then climbs. The lowest point is the cash Klasna must have before the plan starts.

Steady plan, 60 new families every month at 4,000 each:

• month 1: −168,000;
• month 2: −273,600;
• month 3: −325,200;
• month 4: −328,800, the low point (net only −3,600 that month);
• month 5: −289,200, because net cash turns positive (+39,600) when the first cohort pays back.

From there the plan climbs: zero is passed in month 8 (+49,800), and by month 12 the steady plan is 906,000 UAH ahead.

So the steady plan needs about 330,000 UAH of cash to get going, before fixed costs. The slip is to read the first month (−168,000) as the cash needed, or the last month (+906,000) as proof that no cash is needed.

Leo's fast plan acquires 60 families in month 1 and 20 more every month after: 80, 100, 120… up to 280 in month 12.

Each month the newest cohort is also the biggest. It costs 4,000 per family up front and leaves only 1,200 per family that month. The older cohorts are smaller and already on the falling part of the curve. As long as new spend grows faster than what older cohorts bring back, the hole gets deeper.

Fast plan, net cash by month: −168,000, −161,600, −142,800, −112,000, −70,000, −17,800 (month 6), then +43,800 in month 7. The cumulative line bottoms out at −672,200 UAH in month 6.

Compared with the steady plan (−328,800 in month 4), the hole is about twice as deep and comes two months later, although every single family still pays back in month 5.

The slip is to look for the low point where one cohort pays back. The plan's low point is where the monthly net turns positive, and a ramp pushes that later.

At month 12 the two plans look odd side by side: the fast plan has 795,200 UAH of cumulative cash, less than the steady plan's 906,000, after spending 8,160,000 on ads instead of 2,880,000. But it also has 2,040 families instead of 720.

Both facts are true, and they don't mean the fast plan is worse. Growth turns cash today into customers who pay later. The fast plan's newest cohorts were acquired in months 10–12; they have paid only a fraction of their 8,170 UAH of twelve-month contribution so far. A snapshot at month 12 catches them in the middle of their payback.

So the right question is not "does it pay?" (both plans do, because every family pays back) but "can Klasna carry the hole until it does?". The fast plan needs about 672,200 UAH at its low point, roughly twice the steady plan, and later.

The slip goes both ways: dismissing fast growth because month 12 shows less cash, or praising it for the number of families without asking who funds the hole.

Overtrading is growing faster than your cash can fund. Every customer is profitable, the plan pays in the end, and the business still runs out of money in the middle, because the hole was deeper than the cash it had.

To check a plan for it, compare two numbers:

• the depth of the hole: the lowest cumulative cash of the plan;
• the cash available for it: what's in the bank after fixed costs and known dips, such as Klasna's summer.

Klasna's plans, before fixed costs:

• steady: low point −328,800 UAH in month 4;
• fast: low point −672,200 UAH in month 6, turning cash-positive only in month 7;
• extra cash the fast plan needs: 672,200 − 328,800 = 343,400 UAH.

If Klasna can set aside about 700,000 on top of salaries and the summer, the fast plan is affordable. If not, it's overtrading, however good the unit economics. A condition the board can check is better than a yes or no: "we start the fast plan if we hold X in cash and CAC stays at 4,000".

"Every family pays back in five months" is a property of one cohort. A growth plan is a stack of cohorts at different ages, and its cash depends on the mix.

• In a steady plan every month looks alike: one new cohort costs 240,000, and from month 5 on the older ones bring back more than that. Net cash turns positive in month 5, together with the first cohort.
• In an accelerating plan the youngest, most expensive cohorts dominate. In month 5 the fast plan buys 140 families for 560,000 UAH, while all cohorts together leave 490,000: net −70,000. Only in month 7 does it turn positive (+43,800).

If Leo kept accelerating after month 12, the hole would stay open longer still. A plan that accelerates without end never stops needing cash.

The practical rule: never answer "can we afford this plan?" with a per-customer payback. Build the stack month by month and read the plan's own net cash.

The hole is what new customers cost up front, minus what they and the older cohorts have brought back so far. Three things set its depth:

• CAC per customer. A higher CAC raises the up-front part. A cheaper channel, such as school partnerships, lowers it.
• The payback curve. Less contribution in the first months means less coming back while the costs pile up. Keeping more families past their first months, or packages paid up front, bring cash earlier.
• The ramp. Adding new customers faster each month makes the newest, most expensive cohorts dominate for longer.

So the levers that make growth cheaper to fund are the same ones that shorten payback: lower CAC, better early retention, earlier cash.

And the dataset leaves out fixed costs. The team, the office and Klasna's summer dip need cash at the same time as the hole. The real question is whether the hole fits on top of them. The slip is to read the dataset's low point as the full cash need.