How to test a new wholesale route before scaling

A new route can look perfect on twenty test calls and fall apart at two thousand. Congestion, answer-supervision tricks and CLI rewriting often only appear under real volume. The fix is not more test calls — it is a structured trial that increases volume in steps and compares results against a known baseline.

Record a baseline first

Pull ASR, ACD and failure counts for the destination from your current route over the same hours and weekdays you plan to test. Without a baseline, you cannot tell whether a 38% ASR is good or bad for that market.

Run manual checks

• Call handsets you control on the main mobile networks
• Confirm two-way audio and natural ringback
• Check caller ID presentation if the route is sold as CLI
• Verify the call appears in the CDR with the correct duration and cost

Split traffic in steps

Send 5% of live traffic for a day, then 20%, then 50%. At each step compare ASR, ACD and failure causes against the baseline for the same hours. Increase only when the new route matches or beats it.

Watch for red flags

• ACD suspiciously close to a round number — possible false answer supervision
• Answered calls with silence or robotic audio
• CLI replaced by a random local number
• Failure causes shifting from no-answer to congestion as volume rises

Decide and document

At the end of the trial, write down the metrics, the step at which any issue appeared and your decision. If the route passes, set it as primary with your old supplier as failover. If it fails, share the CDR evidence with the carrier — a good one will investigate rather than argue.