Decision framework
Turn a provider search into a route decision.
Searching for a wholesale VoIP termination provider often produces similar claims about global reach, competitive rates and quality. Those statements do not show whether a route will fit your destination mix, call behavior, caller ID requirements or operational model.
A useful comparison starts with the same written traffic profile for every provider. It then measures representative test traffic, normalizes commercial terms and confirms who owns monitoring, fraud controls and escalation. That process gives commercial and technical teams a shared decision record instead of a collection of unrelated rate decks.
01 / Traffic fit
Build the traffic profile before requesting rates.
The same route can behave differently for short-duration contact center traffic, conversational business calls and application-generated notifications. Give each provider the same description of what will be sent so proposals and test results can be compared on equal terms.
Destination mix
List countries, fixed and mobile destinations, expected percentage by market and any priority prefixes. A broad A-Z request is less useful than a ranked traffic mix.
Call shape
Include monthly minutes, busy-hour concurrency, calls per second, average call duration, answer behavior and expected growth or campaign peaks.
Use and caller identity
Describe the business use, originating platform, caller ID source, number rights and whether the traffic supports customers, sales, applications or another lawful workflow.
Practical check: Send one versioned traffic brief to every shortlisted provider and record any assumptions they add to the proposal.
02 / Quality evidence
Use representative testing instead of a generic quality claim.
A short test to one easy destination does not represent a multi-country traffic profile. Agree on the destinations, time window, expected call attempts and acceptance method before testing. Keep the originating platform and caller ID behavior consistent with the intended production setup.
ASR, ACD and PDD are useful indicators, but none is a universal score. Customer answer behavior, voicemail, invalid numbers, time zones and destination networks can change the result. Read the metrics with SIP responses, audio checks and the known traffic sample.
Representative sample
Include priority destinations, fixed and mobile calls, realistic call duration and the busy periods that matter to the operation.
Shared measurement
Review ASR, ACD, PDD, SIP response distribution, post-dial audio, one-way audio and call completion using the same definitions.
Failure analysis
Ask how failed or degraded calls will be traced, what evidence is required for a ticket and who decides whether a route change is appropriate.
Practical check: Keep call detail records and timestamps for agreed test calls so both teams can investigate the same events.
03 / Commercial comparison
Normalize the rate deck and billing terms.
A lower headline rate can be offset by a different billing increment, rounding method, surcharge, currency basis or minimum commitment. Compare the complete commercial model rather than one destination price.
Rate governance
Confirm the rate-deck version, effective date, currency, notice process, prefix format and how additions or price changes are communicated.
Billing mechanics
Document billing increments, rounding, minimum call charges, payment terms, credit limits, taxes or surcharges and the dispute process.
Volume assumptions
Record any destination commitment, overall minimum, quality tier or traffic-balance assumption attached to the quoted commercial terms.
Practical check: Model the proposal against a recent traffic sample so commercial differences are calculated using your own destination distribution.
04 / Live operation
Confirm technical controls, compliance and escalation ownership.
Production readiness includes more than a reachable SIP endpoint. Both sides should know the permitted source IPs, authentication method, codecs, DTMF behavior, CPS and concurrency limits, fraud response and support path.
The provider should also understand the customer, traffic purpose and caller ID plan. Business verification and use-case review protect the service relationship and make expectations clear before volume increases.
Connection controls
Confirm signaling and media IPs, authentication, codecs, DTMF, SIP timers, capacity limits and any maintenance or change process.
Monitoring and fraud response
Define who watches spend, unusual destinations, call attempts, quality shifts and account compromise, plus the authority to block suspicious traffic.
Escalation path
Record support hours, severity definitions, required ticket evidence, named ownership and how commercial and technical issues are coordinated.
Practical check: Do not move full production volume until technical contacts, spend controls and an incident path are confirmed in writing.
Ready-to-send brief
Wholesale termination request checklist
Include these inputs in the first request so a provider can evaluate fit and prepare a focused commercial and test response.
- Company name, website, service purpose and authorized contact
- Originating country, SIP platform and interconnection method
- Destination list with fixed, mobile or priority prefixes identified
- Expected monthly minutes and percentage by key destination
- Busy-hour concurrency, calls per second and average call duration
- Traffic type, caller ID source and number authorization model
- Required codecs, DTMF method and signaling constraints
- Preferred test window, sample size and acceptance measures
- Commercial currency, billing and payment requirements
- Technical, fraud and billing escalation contacts
Common questions
Wholesale provider selection questions
Should I choose the lowest wholesale VoIP rate?
The lowest listed rate is not enough to determine fit. Compare the full billing model, route behavior on representative traffic, capacity limits, support, fraud controls and the provider response when a destination degrades.
How long should a termination test run?
The useful duration depends on the destination mix and call volume. The sample should include enough representative calls and busy periods to review completion, SIP responses, audio and operational behavior without treating a very small test as conclusive.
Which quality metrics should be reviewed?
Common inputs include ASR, ACD, PDD, SIP response codes, audio behavior and destination-specific completion. Interpret them with the call sample, customer answer behavior and known invalid or unanswered traffic.
What should be agreed before production traffic?
Confirm commercial terms, permitted traffic, caller ID rules, connection details, CPS and concurrency limits, spend and fraud controls, monitoring responsibility and technical and billing escalation contacts.