Classic scaling pain: payroll swells, reps answer in hours, quotes have errors, conversion drops. An AI employee took the routine leads — people stayed on the hard deals.
- sales reps on staff
- 15 → 4
- instead of 5 hours to reply
- 3–5 sec
- effect per year
- ₽89.7m
What broke as they grew
- Payroll: 15 sales reps.
- Time to reply — up to 5 hours.
- Quote errors from manual maths.
- Conversion around 10–15%.
- Service quality fell as lead volume rose.
What we shipped
The AI employee replies 24/7, does not “round” figures, handles standard objections and writes straight into CRM. Reps only pick up what needs a human.
Savings
- Headcount 15 → 4.
- Errors down 98%, plus ₽2.2m/year on returns.
- Payroll: ₽6.5m/year (11 roles × ₽60,000 × 12).
Sales lift
- Reply time: 5 hours → 3–5 seconds.
- Conversion ×2: 10% → 20%.
- Extra revenue around ₽81m/year from speed and volume.
- The lead queue no longer waits on how many people are on shift.
Overall economic effect — over ₽89.7m a year. The company scales without hiring another sales wave.
An AI employee does not fire the desk. It takes nights, the queue and the arithmetic — so four reps close real deals instead of pasting the same answer.