Skip to content
Back to blog
9 min readBy The dialque Team

Per-user vs per-minute pricing models in cloud telephony: which is actually cheaper

Per-user looks predictable but penalises low-utilisation seats. Per-minute looks fair but punishes growth. Per-active-user is the third option that very few vendors offer. Here is the math.

PricingBuyer guide

Cloud telephony pricing in India clusters around three models — per-user, per-minute, and per-active-user — and most teams pick the wrong one because the vendor's pitch deck shows the favourable number, not the realistic total. This post is the math, the hidden costs, and the questions to ask before signing.

The three pricing models

Per-user (per-seat) pricing

₹1,500 - ₹3,000 per agent per month, regardless of how many calls they make. Common on platforms that bundle voice + IVR + recording + reporting into one SKU.

Best for: Teams with predictable, consistent utilisation across all seats. Worst for: Teams where some agents make 50 calls/day and others make 5.

Per-minute pricing

₹0.80 - ₹1.40 per minute of voice traffic, with usually a small monthly platform fee on top. The minute rate covers the platform, the carrier markup, and the vendor margin.

Best for: Teams with bursty / seasonal call volume. Worst for: High-volume steady operations — the per-minute rate compounds fast.

Per-active-user pricing

A newer model: ₹1,000 - ₹1,500 per agent per month, but only for agents who actually logged in and placed at least one call that month. Inactive seats, agents on leave, churned hires — not billed.

Best for: Teams with variable headcount (hiring waves, attrition cycles, seasonal scale-up). Worst for: Nobody, really — it is per-user pricing with the unfair edge cases removed.

The hidden charge underneath all three: per-minute carrier cost

Whatever the platform pricing model, the carrier minutes are billed separately or bundled. Three common patterns:

Bundled (no separate minute billing). The platform fee includes "unlimited minutes" up to a fair-use cap. Looks attractive but the cap is usually low (~200 minutes/agent/day) and overages are at a markup.

Marked-up minutes. The platform charges per minute of voice (₹0.90 - ₹1.40 in India), of which ₹0.30 - ₹0.60 is the actual carrier cost. The 2-3x markup funds the platform.

Passthrough minutes. The platform does not touch the carrier billing. You contract directly with Vilpower / Tata / Airtel Business, pay ₹0.30 - ₹0.60 per minute, and pay the platform a separate (typically smaller) per-user fee on top.

Most teams underestimate how much of their total telephony cost is minutes. For an outbound SDR team running 4 hours of talk-time per agent per day, minutes are 60-75% of the all-in cost. Choosing passthrough over marked-up minutes routinely saves more money than any platform-feature trade-off.

Modelled cost across team sizes

Same workload — agents doing 4 hours of outbound talk-time per day, 22 working days/month — under all three pricing models. Carrier rate is ₹0.50/min if passthrough, ₹1.10/min if marked-up.

10 agents, ~110,000 outbound minutes/month

| Model | Calculation | Total | |---|---|---| | Per-user (bundled minutes, with overage) | 10 × ₹2,500 = ₹25,000 + overage on first 30k min × ₹1.10 = ₹33,000 | ₹58,000 | | Per-minute (marked-up) | Platform ₹15,000 + 110k × ₹1.10 = ₹121,000 | ₹136,000 | | Per-active-user (passthrough) | 10 × ₹1,200 = ₹12,000 + 110k × ₹0.50 = ₹55,000 | ₹67,000 |

At 10 agents, per-active-user with passthrough is 40% cheaper than per-minute marked-up.

30 agents, ~330,000 outbound minutes/month

| Model | Calculation | Total | |---|---|---| | Per-user (bundled, overage hits hard) | 30 × ₹2,500 + 270k min overage × ₹1.10 = ₹297,000 | ₹372,000 | | Per-minute (marked-up) | Platform ₹45,000 + 330k × ₹1.10 = ₹363,000 | ₹408,000 | | Per-active-user (passthrough) | 30 × ₹1,200 + 330k × ₹0.50 = ₹201,000 | ₹201,000 |

At 30 agents, per-active-user + passthrough saves ~₹2 lakh/month vs the bundled-minute alternative. Over a year, that is enough for an extra senior agent or two.

100 agents, ~1.1M outbound minutes/month

| Model | Total | |---|---| | Per-user (bundled, large overage) | ~₹13.5 lakh/month | | Per-minute (marked-up) | ~₹13.6 lakh/month | | Per-active-user (passthrough) | ~₹6.7 lakh/month |

Halved. The pricing model matters more than any feature decision at this scale.

Where each model breaks down

Per-user, with no real utilisation enforcement

You hire 30 agents in March for a campaign. By June, 10 of them are on leave / underperforming / parked on other tasks. You are still paying for 30 seats. The vendor has no incentive to make this visible — they invoice you for 30 seats regardless.

Fix: insist on per-active-user pricing in the contract, or get a clause that lets you reduce seat count quarterly.

Per-minute, when call volumes grow

A growing team feels per-minute pricing as a tax on success. Doubling team size doubles minutes doubles spend. Forecasting cuts both ways — you cannot project budget if your call volume varies 30% month over month.

Fix: re-negotiate at every renewal once volume is predictable, or move to per-active-user.

Per-active-user, when seats are continually rotating

The model assumes activity is binary (logged in + made a call vs not). If an agent logs in for one call by mistake, they count as active. Most vendors handle this with a minimum-activity threshold (e.g. "must have made 5+ calls in a month"). Verify the definition before signing.

Hidden costs nobody tells you about

1. Setup / onboarding fees

₹15,000-50,000 one-time, often only on annual contracts. Negotiable.

2. Number-rental on top of platform fees

₹400-800 per DID per month. A team running 50 numbers is ₹20-40k/month before they place a single call. Usually NOT included in the headline "per-user" price.

3. API access tier

Some platforms gate API features to higher plans. If you need webhooks, calendar integration, custom call routing — confirm what tier you need.

4. Recording storage

First 30 days usually included. After that, ₹2-10 per GB per month. A 10-agent team accumulates ~50GB/month — small individually, but storage costs hit at retention horizons (3-7 years for BFSI).

5. Contract renewal mechanics

"6-month auto-renew" clauses are common and rarely caught at signing. Read the termination terms.

6. Currency / payment surcharges

Some platforms invoice in USD; small fluctuations cost lakhs/year for large customers. India-incorporated vendors invoicing in INR avoid this.

7. Premium support tier

Default support is usually email + business hours. 24×7 phone is a separate paid tier. If you need it, factor it in.

The questions to ask vendors

Send these in the procurement email — the answers tell you what model you are actually buying:

  1. Is the per-user fee for provisioned seats or active users? What is the definition of "active"?
  2. Are voice minutes bundled, marked-up, or passthrough? At what rate per minute?
  3. What is the markup over the underlying carrier cost? (Most will not tell you this; the answer itself is informative.)
  4. Can I bring my own carrier (BYOC)? At what platform-side fee?
  5. What is the recording storage policy? When does it become billable?
  6. What is the all-in cost for a representative month of my team's volume? (Get this in writing.)
  7. What is the contract term and renewal mechanic?
  8. What are the early-termination fees?
  9. Are setup / training / integration fees one-time, in writing, capped?
  10. Is the API on the base plan or a paid tier?

Why most teams pick the wrong model

Three patterns we see repeatedly:

Pattern 1: optimising for the lowest sticker price The "₹999/user" plan looks cheaper than the "₹1,200/user" plan. But the ₹999 plan has marked-up minutes (₹1.10/min) and the ₹1,200 plan has passthrough (₹0.50/min). The ₹1,200 plan is cheaper after a few thousand minutes.

Pattern 2: not modelling the team six months out You sign at 10 agents. In nine months you are at 25. The plan that was cost-competitive at 10 is brutal at 25. Always model the team you are headed towards, not the team you are.

Pattern 3: ignoring the carrier cost Vendors love this — they pitch "all-inclusive ₹2,499/user" because the minute cost is hidden. Always separate the platform cost from the minute cost in your evaluation.

What dialque does

For transparency: dialque uses per-active-user pricing (₹1,200/month per agent who logged in + made at least one call that month) with carrier passthrough. You pick your own SIP trunk; dialque does not mark up minutes. This is intentional — at any team size, the math is cleaner and the vendor incentive is aligned (we want you to actually use the platform, not pay for inactive seats).

If your team is on a per-user or per-minute marked-up plan today and the renewal is approaching, the per-active-user + passthrough math typically saves 30-50% at 10+ agents.

Frequently asked questions

What is the rough threshold where per-minute becomes too expensive?

Above ~50,000 outbound minutes/month, per-minute marked-up plans become structurally expensive. Switch to either per-user (with low markup) or per-active-user + passthrough.

Does per-active-user mean unlimited calls per active agent?

Per-active-user covers the platform fee. Minutes are billed separately (either bundled or passthrough). Verify what is included in the base fee.

How does WhatsApp / SMS pricing typically work?

Almost always per-message, on top of voice pricing. WhatsApp via the BSP is ₹0.50-1.10 per conversation (Meta + BSP markup). SMS is ₹0.15-0.25 per message in bulk. Both are billed separately from voice.

What about pay-as-you-go (no commitment) pricing?

A few vendors offer it. The per-unit rates are higher (20-40%) but there is no contract. Good for short pilots; expensive at scale.

Should I sign annual or month-to-month?

Annual usually gets a 15-25% discount but you are locked in. Negotiate a "termination for material breach within 30 days" clause to give you an out if the platform underdelivers.

The pricing-model decision is more financially consequential than any feature decision past 10 agents. Model the all-in cost — platform + minutes + DIDs + storage — at your projected team size before signing, and the right model usually picks itself.