Two quarters missed. Two spend requests on the table in the same conversation.
We need 5 more people.
We already invested ₹7.8L on AI to accelerate sales.
Before we add anything — what did the ₹7.8L change?
[Silence]
Two spend requests on the table, and no measurement on what the last spend produced
The 20-minute check before you approve anything further
Two separate requests were being evaluated in that room — an AI tool spend and a headcount request. Each deserves its own 20-minute test before either gets approved.
On the AI spend
Name what the ₹7.8L was supposed to improve, in two measurable numbers, before and after. Common candidates: meetings booked per salesperson, or conversion rate from proposal to closed-won.
If neither number moved, understand why before investing in more tools. A tool that has not changed either metric will not change them by being joined by a second tool. The gap is not the AI category — it is that nobody defined, in advance, what success from that spend would look like.
On the five new hires
Calculate investment per hire against median output, not average output. Average is skewed upward by outliers — the one or two outperformers who are not representative of what a typical new hire will produce in their first months.
The Real Math
5 new hires
Not 5 potential outperformers. Five people who will most likely perform at or near the team's median, at least initially.
Median monthly output
Not average. The median reflects what a typical hire actually produces — the average is inflated by whoever is carrying a disproportionate share of the number.
Realistic months at full productivity
New hires typically reach median output only after three to six months. Multiply by the months genuinely remaining in the fiscal year — not the months since the hire started.
Run that calculation before approving the headcount. In most cases, the answer is smaller than the founder expects — and the fiscal year is often over before the new hires reach full productivity at all.
Hiring and AI may both still be right decisions
Neither request is automatically wrong. Additional headcount and additional tooling both have legitimate cases. But they answer a specific question — what capacity will be needed next — and that is a different question from why the capacity already paid for is not producing enough right now.
Hire now to produce more in Q4 and next fiscal year. Manage better what you have already invested in for this one
The two questions are not in conflict. But approving new spend without measuring the old spend means the same unmeasured pattern repeats at a larger scale next quarter — and the underlying constraint, whatever it actually is, never gets named.