12 Questions to Ask a Wholesale CLI Provider

Routing and CLI delivery

1. How do you make sure my CLI is delivered on every call?

Why it matters: many providers mix CLI and non-CLI carriers on the same destination to save cost. A good answer names a fixed set of carrier partners per destination, describes how often CLI is sampled with real handsets, and explains what happens when a partner fails. "We have direct routes" with no detail is not an answer. See What Is a CLI Route?.

2. What happens to my traffic when your primary partner fails?

Why it matters: failover is where CLI is most often lost, because the backup is cheaper and less tested. A good answer explains that backup partners are also CLI-tested, that failover is visible in the portal, and that you can choose to fail rather than fall back to a non-CLI path.

3. Is this route suitable for my traffic type?

Why it matters: a route built for conversational calls may be restricted the day you start predictive dialing. A good answer asks about your traffic first: agent or IVR, average call length, short-call ratio, CPS. Read Short-Duration Traffic and CC Routes and IVR & Press-1 Routes Explained.

Quality and visibility

4. Can I see ASR, ACD, PDD and NER per destination in real time?

Why it matters: problems you can see within minutes cost far less than problems you find in a monthly report. A good answer is a portal with live metrics per destination and per endpoint, plus CDR export. Learn what each metric means in ASR, ACD, PDD and NER.

5. How do you detect and handle false answer supervision?

Why it matters: FAS inflates both your bill and your ASR. A good answer describes automated ring-out and invalid-number tests, media analysis, removal of offending partners and a clear credit policy. See False Answer Supervision.

6. Will you support a structured acceptance test?

Why it matters: a few test calls do not show how a route behaves at volume, across the day, over a week. A good answer is "yes, here is a test allocation and an engineer to review the results with you". Our 7-day route acceptance test plan is a template you can use with any provider.

Compliance and stability

7. What do you check during onboarding?

Why it matters: this sounds like a question about hurdles, but it is a question about route stability. Providers that accept any traffic from anyone lose interconnects quickly, and your traffic goes down with them. A good answer covers identity checks, CLI ownership verification, an acceptable use policy and campaign review for automated traffic. Read our compliance approach.

8. How are my US and Canadian calls attested under STIR/SHAKEN?

Why it matters: attestation level shapes how your calls display. A good answer names who signs your calls, at what level and on what basis, and confirms every provider in the path is listed in the Robocall Mitigation Database. See STIR/SHAKEN for Call Centers Outside the US.

9. Do you know the CLI rules in my destinations?

Why it matters: presenting the wrong number type in some countries leads to blocked or relabelled calls. A good answer includes country-specific guidance, like the summaries in CLI Rules by Country.

Capacity, billing and support

10. What channel and CPS limits apply, and how quickly can they be raised?

Why it matters: limits you did not know about look like route failures. A good answer states per-account and per-destination limits, how to raise them, and how bursts are handled. See SIP Channels & CPS Sizing.

11. How exactly is billing calculated?

Why it matters: billing increment (60/60, 30/6, 6/6, 1/1) changes the effective cost of short calls dramatically, and some providers charge for attempts or apply surcharges for short-call ratios. A good answer shows increments per destination, explains any surcharges up front, and provides per-call CDRs that reconcile with your own. Our How Billing Works page sets this out for CLIRoutes.

12. Who do I talk to at 3 a.m. when a destination fails?

Why it matters: dialers run when your account manager sleeps. A good answer is a 24/7 NOC reachable by ticket and email, with engineers who can read a SIP trace, and target response times you can hold them to.

Red flags

• Offers to let you present any number without checking ownership.
• No onboarding, no AUP, or promises that "anything goes" on a route.
• Unwillingness to share quality data or support a test.
• ASR claims quoted without reference to traffic type.
• Rates far below everyone else's on high-demand destinations, with no explanation of the path.

Good answers and weak answers, side by side

A provider that hesitates or gives a weak answer is not necessarily dishonest, but it is telling you that it has not built the processes that protect your traffic. Ask for evidence: a sample quality report, a sanitised SIP trace, the AUP, the onboarding checklist.

Comparing cost per contact

To compare providers fairly, normalise to the outcome you pay agents for: a live conversation. For each candidate, divide the total cost of a test period (minutes billed at the provider's increment, plus any attempt or surcharge fees) by the number of conversations longer than your qualifying threshold, for example 30 seconds. A route with a higher per-minute rate but better CLI delivery and no false answers often wins comfortably.

Run the comparison on the same traffic split between providers at the same time, not in consecutive weeks. Lists, seasons and even weather change answer rates; a simultaneous split removes those variables.

Asking for references

Ask for one or two references with a similar traffic profile to yours: the same type of campaign, similar volumes and overlapping destinations. When you speak to them, ask how quickly problems were acknowledged and fixed, whether routes have ever been withdrawn mid-campaign, how accurate the CDRs have been against their own records, and whether they would still choose the provider today. A provider confident in its service will arrange this; reluctance to connect you with any customer at all is worth noting.

Then trust your own test results above anything you are told. References and answers narrow the field; a structured, parallel test on real traffic makes the decision.

A simple scorecard

Score each answer from 0 to 3 and weight the areas. Then run the acceptance test on your top two candidates with the same traffic. Ready to test CLIRoutes? Open an account; your rate deck is available in the portal after sign-up.