COMPARISON · 01
The manual desk vs the AI quote desk.
The manual quoting desk is not broken — it is bounded. It works exactly as well as your best pricing head's capacity, memory and waking hours. This is the honest line-by-line of where those bounds sit, and what changes when software carries the desk.
| Dimension | Manual desk | AI quote desk |
|---|---|---|
| Intake | Re-keyed from email/Excel/WhatsApp by hand | Parsed to structured lines in seconds; coordinator verifies |
| Coverage | RFQs silently dropped under load (industry: most never answered) | Every RFQ read; 95%+ answered |
| Response time | Hours on a good day; industry median 57 hrs | Draft in ~21s; sent in minutes after approval |
| Hours | Office hours, one timezone | The 02:14 RFQ is drafted at 02:14 |
| Surcharge completeness | Depends who's quoting and how busy they are | Checklist runs identically every time |
| Margin discipline | In the pricing head's judgment (and mood) | Floor enforced; customer pattern surfaced on every draft |
| Rate freshness | Portal logins per enquiry; validity in memory | All five rate sources read live, source shown |
| Follow-up | When someone remembers (industry: 8%) | Every quote chased; replies classified |
| Key-person risk | The desk IS a person; leave and resignation hurt | Knowledge persists in the system |
| Scale | Linear: more quotes = more hires | 225+/day on the same team is a live number |
| The send | Human | Still human — the gate never leaves supervised mode |
The cost-per-quote arithmetic
A pricing coordinator producing 25 quotes a day at a fully-loaded monthly cost of, say, ₹80,000 (or AED 8,000) prices each quote at roughly ₹150–160 of labour — before the cost of the ones never answered, which industry data suggests is the larger number. The manual desk’s real cost lives in unanswered RFQs and missed surcharge lines, not salaries.
- Forwarder gross margins run 10–20% with net margins near 3–4% (industry benchmarks) — one missing THC line can erase a shipment’s profit.
- Every unanswered RFQ is a shipment your competitor won uncontested.
- The follow-up nobody has time for is statistically where win rate hides.
What the manual desk keeps forever
Two things do not automate: the relationship and the commitment. Knowing that this customer's "urgent" means Thursday, that this lane's consignee rejects partial shipments — that context makes your people better approvers. And the send itself stays a human decision at the gate, because the quote carries your company's name.
If your desk sees fewer than ~10 RFQs a day and one person handles them with time to spare, a disciplined manual desk with good templates is defensible. Above that, the bounds show: coverage slips first, then speed, then margin discipline. The AI desk doesn’t replace your pricing head — it removes the ceiling over them.
Run your own numbers.
The calculator models coverage and speed against your volumes — conservative math, your inputs.
Open the calculator