How to Choose a Wholesale VoIP Provider: A Buyer's Checklist
Wholesale VoIP provider selection is the process of evaluating carriers that sell bulk voice termination, judged on route quality and effective cost rather than on the advertised per-minute rate. The headline rate is the least informative number on any rate sheet, and buying on it is the most common and most expensive mistake in this market.
This guide is written for carriers, resellers and high-volume dialers comparing termination suppliers. If you are connecting a single organisation’s phone system to the public network, you want SIP trunking instead — different commercial model, different evaluation criteria.
Why is the cheapest rate usually not the cheapest route?
Because you pay for failed calls too, just not on the invoice.
A route quoted at a very low per-minute rate with a poor answer-seizure ratio costs more in practice than a higher-priced route that connects reliably. Every failed attempt consumes dialer capacity, agent availability and a contact opportunity. If a cheap route answers at 40% where a dearer one answers at 65%, the cheap route is more expensive per connected call long before you account for the lost contact.
Three mechanisms turn a cheap quote into an expensive one:
| Mechanism | What it looks like | What it costs |
|---|---|---|
| Low ASR | Calls fail to connect, often silently | Wasted dialer capacity and lost contacts |
| Minute rounding | A 20-second call billed as 60 seconds | Up to 3× on short-duration traffic |
| Stripped CLI | Recipient sees an unfamiliar or blank number | Materially lower answer rates |
Compare carriers on effective cost per connected minute on your own traffic profile. That single change in method removes most bad wholesale decisions.
What is ASR and why does it matter more than price?
Answer-seizure ratio (ASR) is the percentage of call attempts that are answered. Average call duration (ACD) is the mean length of those answered calls. Together they describe route quality far better than either does alone.
The combination matters because of a specific failure mode: false answer supervision. A route reporting high ASR with an unusually low ACD is often signalling “answered” when the call has not genuinely connected, which starts billing early. High ASR paired with a 6-second ACD is not a good route; it is a route that is charging you for calls nobody took.
Ask for both figures, per destination, with the sample size attached. A percentage without its denominator is not evidence. A carrier unwilling to break down performance on your top five destinations is telling you something.
How should I compare billing increments?
Per-second billing versus rounding up to the minute changes your effective cost more than most rate differences do.
Consider 10,000 calls averaging 25 seconds:
- Per-second billing: 4,167 billable minutes
- Rounded to the minute: 10,000 billable minutes
That is 2.4× the cost on identical traffic. A carrier quoting a rate 20% lower with minute rounding is substantially more expensive than one quoting higher with per-second billing — and the rate sheet alone will not tell you.
Short-duration traffic is where this bites hardest: verification calls, notifications, abandoned dials. Get the increment confirmed in writing before comparing anything.
What is CLI delivery and why does it affect price?
CLI (Calling Line Identification) is the caller ID presented to the recipient. Routes that preserve CLI end to end cost more per minute than routes that strip or replace it, and the difference is usually worth paying.
Two reasons. First, answer rates: people are measurably less likely to answer a call showing no number or an unfamiliar one. Second, regulation — CLI integrity is a legal requirement in a growing number of jurisdictions, and in the US, STIR/SHAKEN attestation depends on it.
Cheap “grey” routes frequently achieve their price by stripping CLI or routing through intermediaries that do. Test this explicitly during a trial rather than accepting an assurance, because it is easy to verify and easy to claim.
What should a rate sheet actually contain?
A usable rate deck is dated and destination-specific. Interconnect pricing moves continuously, so any permanently published public A-Z price list is either stale or describes a single route.
Expect to see:
- Destination, with the specific prefix or breakout — not just “Mexico” but mobile versus landline versus specific carriers, which can differ by an order of magnitude
- Rate per minute
- Billing increment
- Whether CLI is guaranteed on that route
- Effective date and notice period for changes
If a rate sheet lacks prefix-level breakouts, you cannot predict your bill. Average country rates hide the mobile destinations that will dominate your spend.
How do I evaluate failover behaviour?
Ask one question: what happens to in-progress and subsequent traffic when a primary route degrades?
The answers vary more than you would expect. Some carriers requalify automatically to an alternative route within seconds. Some requalify only on complete failure, not on quality degradation — which means you keep sending traffic down a route that is technically up and practically unusable. Some require you to notice and reroute manually.
Quality-based routing that reacts to falling ASR, rather than only to a dead route, is the meaningful differentiator. Ask how degradation is detected, what threshold triggers requalification, and how long it takes.
Should I use one carrier or several?
Several, once volume justifies it. Single-sourcing termination concentrates risk: one carrier’s outage or quality incident becomes yours entirely, and you lose pricing leverage at renewal.
The practical pattern is a primary carrier for most traffic with at least one qualified alternative already configured and periodically carrying live traffic. A failover route you have never actually used is an assumption, not a backup.
Balance this against operational overhead. Two or three carriers is manageable; ten fragments your volume commitments and your attention without adding resilience.
How do I run a pilot properly?
Quoted figures are marketing. Measured figures are evidence.
Send a small but representative share of real traffic for two to four weeks — representative matters, because a pilot sent only to easy destinations proves nothing about the routes that will cause problems. Measure ASR, ACD, PDD and CLI delivery yourself rather than relying on the carrier’s reporting.
Compare measured results against what was quoted. A carrier whose real performance matches its rate sheet is worth more than one quoting better numbers it cannot reproduce.
Frequently asked questions
What is a good ASR for wholesale VoIP?
It depends heavily on destination and traffic type, so a single target figure is misleading. What matters is comparing carriers on the same destinations with your own traffic, and investigating any route where ASR is far below your other carriers on that destination.
How often do wholesale rates change?
Frequently — carrier interconnect pricing moves continuously. Ask what notice you receive before a rate change and whether committed volume locks a rate for a defined period.
Is per-second billing standard?
It is common but not universal, and it is one of the largest hidden differences between quotes. Always confirm the increment in writing and recalculate effective cost on your own call duration profile.
What is false answer supervision?
Signalling a call as answered when it has not genuinely connected, which starts billing early. The signature is high ASR with an unusually low ACD. Monitor both metrics together to detect it.
Do I need a minimum volume commitment?
Many carriers price against committed volume, but terms vary. Be cautious about large commitments before a pilot has validated real-world quality on your destinations.
Salamtalk provides wholesale VoIP termination with per-second billing, guaranteed CLI delivery and quality-based routing. Rate decks are issued per account and dated. See carrier infrastructure for the technical detail.