Per-call answering services vs flat-rate AI

The pricing model is the quiet variable in answering services — and it changes behaviour more than the headline rate. Per-minute/per-call vs flat-rate creates opposite incentives; understanding it predicts your real bill.

How per-call pricing actually plays out

Traditional answering services charge per call or per minute because humans have real per-unit costs — which creates two dynamics:

The bill is a moving target. Quiet month: £80. Busy month: £400. Success (more calls) costs more — a tax on growth.

It changes your behaviour. Owners on per-call pricing start rationing — "do I really need them answering weekends?" The coverage they bought gets dialled back to control cost, which re-opens the missed-call problem they paid to close.

The flat-rate dynamic

Flat subscription pricing inverts it:

• Volume is free — the 200th call costs the same as the 1st. Growth is pure benefit.
• Budget certainty — the bill's the same every month, forever.
• Full coverage stays on — no incentive to turn it off evenings or weekends; the whole point is complete coverage.

The honest maths at real volumes

For a business averaging 150 calls/month at ~3 minutes each:

• Per-minute service at £1.50/min: ~£675/month.
• Per-call bundle at £2/call: ~£300/month.
• AI flat rate: £29–£99/month.

The per-unit models' economics are real — they reflect human costs — but they make small businesses pay enterprise-style prices for capacity they need.

When per-call still makes sense

• Very low volume (<20 calls/month) where a minimal bundle beats flat rate.
• Where human answering is genuinely required for every call.
• As a human fallback behind AI — using it only for escalated calls keeps per-call costs low.

The strategic point

Answering cost should be a fixed cost like rent, not a variable cost like fuel — because missed-call losses scale the same way. The business that can afford complete coverage is the one with flat-rate pricing; the one metered by the minute will always ration the very thing it bought.