What is the worst investment the fund has made? Write the post-mortem on Bazaarvik Social Commerce, written off after the fund put in Rs 45 crore over three rounds while buyers grew 6x on incentives of up to Rs 140 an order. Which signals were visible at each round, and what rule would have stopped the follow-ons?

The decisive signal was visible before the second cheque: in the test city without incentives only 16% of buyers stayed to month three, against 41% reported with them. Buyers grew 6x while incentives rose from Rs 20 to Rs 140 an order, so retention was being paid for. A rule requiring 25% month-three retention without incentives, and positive contribution before incentives, would have stopped the Series B and C cheques and saved Rs 35 of the Rs 45 crore.

After a bad cricket season, a good coach replays the matches and asks at what point the problem was visible and what rule would have caught it, not who was unlucky. A venture post-mortem does the same: for each cheque, list what the fund knew, find the first number that pointed to the outcome, and turn it into a rule the fund can use next time. Interviewers who ask about the worst investment are testing whether you can do this honestly, without blaming the market.

At the Series A, 38% month-three retention with a small Rs 20 incentive was a reasonable start. At the Series B, buyers had tripled, but the incentive had risen to Rs 90 an order, the loss per order had widened from Rs 90 to Rs 150, and the test city showed only 16% retention without incentives. At the Series C, buyers were 12 lakh and reported retention 44%, but incentives were Rs 140 an order and the test city still showed 15%. Monthly losses on orders had grown from about Rs 2.2 crore to Rs 25.9 crore.

Follow one buyer. Over three months at 1.2 orders a month and Rs 140 an order, the company spends about Rs 504 on incentives. If only 14% of those buyers stay once incentives stop, each lasting buyer cost about Rs 3,600 in incentives alone, and even that buyer loses Rs 40 an order before incentives, so there is nothing to pay the Rs 3,600 back. Incentives can be a sound way to start habits if the buyer later earns a profit. Here contribution before incentives was negative at every round, so growth made losses bigger, not smaller.

No follow-on unless buyers acquired or kept without incentives clear 25% retention at month three, and contribution per order before incentives is positive or on a dated path to positive. At the Series B the test city showed 16% and contribution before incentives was minus Rs 60, so the rule fails and the fund keeps Rs 35 crore of reserves for other companies. The honest part of the post-mortem is that the data was in the room: the fund looked at reported retention and buyer growth because those were the numbers the company led with. The limit: one test city can be unrepresentative, so the rule should ask for a clean test, not just any number.