Compare an AI model company burning cash with a mature software company. Which multiple fits each, and what revenue multiple gap is justified?

Value Setubandh on earnings and Nabhra on what it can earn once it stops burning; on that basis Setubandh deserves about 5.4x revenue and Nabhra about 3.7x, so 10x for Nabhra is not defensible. Setubandh keeps Rs 85 of every Rs 100 before operating cost and earns Rs 30 of EBIT; Nabhra keeps Rs 45 and loses Rs 30. Growth adds rupees of revenue; gross margin decides what each is worth. On gross profit Nabhra does earn the higher multiple, 8.2x against 6.4x, and that is the honest way to pay for its growth.

Put both on Rs 100 of revenue. Nabhra pays Rs 55 for compute and delivery before it has sold anything, then Rs 75 of operating cost, so EBIT is minus Rs 30. Setubandh pays Rs 15 to deliver, Rs 55 to run, and keeps Rs 30. Nabhra's cost of delivering grows with every customer, like a restaurant that buys ingredients for every plate; Setubandh's grows hardly at all, like a cookbook that is written once and sold many times. That is what a gross marginRevenue less the direct cost of delivering it, as a share of revenue. For Nabhra the direct cost is GPU time per query; for Setubandh it is hosting and support. of 45% against 85% means, and it is structural until Nabhra owns cheaper compute or charges more per query.

Setubandh has stable earnings, so use EV to EBIT. At 18x, a reasonable mark for 12% growth with 30% margins, its enterprise value is Rs 32,400 crore, 5.4x revenue and 6.4x gross profit. Nabhra has no earnings, so a revenue multiple is used because it is the only line that is positive, which is a reason to be careful, not a reason to pay. Forecast it to the point where it earns: revenue growth of 80%, 50% and 30% takes revenue to Rs 7,020 crore in year 3, and if operating cost falls from 75% of revenue to 35%, the EBIT margin reaches 10%, Rs 702 crore. At the same 18x that is Rs 12,636 crore in year 3; discount three years at 15% for the risk and subtract the burn of Rs 720 crore in year 1 and Rs 378 crore in year 2 that someone must fund, and the value today is about Rs 7,396 crore, 3.7x revenue and 8.2x gross profit.

The defensible position is Setubandh at about 5.4x revenue and Nabhra at about 3.7x, with Nabhra on the higher gross profit multiple, 8.2x against 6.4x, which is where its growth is honestly paid for. The fund's 10x revenue, Rs 20,000 crore, is the Setubandh multiple applied to a business that keeps half as much of each rupee; to justify it Nabhra would need a mature EBIT margin near 20%, which gives about Rs 15,705 crore, 7.9x, and even that falls short. The growth drivers to look at are the ones that move gross margin: own data centres or cheaper chips, pricing per query rising with model quality, and the mix shifting to enterprise contracts with committed volumes. Setubandh's drivers are duller and more reliable: price increases, seat growth and new modules sold to the same customers.

Say the limitation. Every Nabhra number after today is an assumption: the growth path, the opex path, the 18x and the 15%. A 15% EBIT margin in year 3 instead of 10% adds half as much again to the value. The honest answer is a range with the assumptions on the table, and the one sentence to leave with the client is that a revenue multiple borrowed from a software peer assumes the peer's gross margin, which Nabhra does not have.