Metervik APIs bills on usage, and six months of revenue swung between Rs 1.9 and Rs 3.0 crore a month. The founder quotes ARR of Rs 36 crore, and the top customer is 30% of revenue. What run-rate would you underwrite, and how do you treat the top customer?

Underwrite about Rs 33.6 crore, the trailing three months annualised, and call it run-rate revenue rather than ARR; then value the top customer's Rs 10.1 crore separately at a discount. The Rs 36 crore annualises the single best month. Treating the top customer's revenue as worth half of the rest gives a risk-adjusted base of about Rs 28.6 crore, 21% below the founder's figure, so 12x on Rs 36 crore is really about 15.1x on what you underwrite.

A shop owner who takes a festival week's sales and multiplies by 52 has not described her shop; she has described her best week. Usage revenue moves with the customer's volume, so any single month is part trend and part noise, and the best month is the most noise you can find. Metervik's six months swing from Rs 1.9 to Rs 3.0 crore; month to month it moved +14%, -21%, +47%, +7%, -13%. The peak month times twelve is Rs 36.0 crore. The last month alone gives Rs 31.2 crore, which is too exposed to one quiet month. The trailing three months average Rs 2.8 crore, Rs 33.6 crore a year, which keeps the recent level and smooths one month's noise.

Because the business has moved. The first three months total Rs 6.4 crore and the last three Rs 8.4 crore, 31% higher. A six-month average of Rs 29.6 crore blends in a level the company has already left, so it is conservative for the wrong reason. The trailing quarter keeps the step up and still smooths the noise. Say the word that matters to the partner: this is run-rate revenue, not ARR. No customer has contracted to pay it, and any of them can cut calls next month without breaking a contract.

Picture a tailor whose biggest client is one wedding planner sending a third of the work. The tailor's shop is worth less than one with the same sales spread over a hundred customers, because one phone call can take a third of it away. Split the Rs 33.6 crore into Rs 23.5 crore from everyone else and Rs 10.1 crore from the lending app, and value the second part at a discount: at half, the base you underwrite is Rs 28.6 crore. That is not a forecast that the customer leaves; it is a price for the chance that it builds the API itself, switches to a cheaper vendor or simply lends less.

Now the price. At 12x the founder's Rs 36 crore the round values Metervik at Rs 432 crore. On the Rs 28.6 crore you underwrite, the same Rs 432 crore is 15.1x, a different deal. The judgement: underwrite Rs 33.6 crore of run-rate, price on about Rs 28.6 crore, and ask for the evidence that would close the gap: the top customer's contract and notice terms, its monthly volume trend, and whether month five's peak came from one-off traffic. The limitation is that the 50% haircut is a judgement; a signed minimum commitment from that customer would justify a smaller one.