Low ASR on a VoIP route comes either from a change on your side (numbers, timing, caller ID reputation, traffic mix) or from the route itself getting worse (congestion, blocking, a changed supplier further down the chain). Check NER and the failure codes first: they tell you which of the two you are dealing with.
First question: is it the route or the traffic?
ASR is answered calls divided by attempts, so it moves when people stop answering as well as when the network fails. NER separates the two because it counts busy and unanswered calls as successes for the network. The method is in NER vs ASR.
- NER stable, ASR down: calls reach phones, people do not answer. Look at your side.
- NER and ASR down together: calls are failing in the network. Look at the route.
A quick extra check: send the same traffic through a second route for an hour. If ASR falls on both, the cause is almost certainly your traffic.
Causes on your side
| Cause | What to check |
|---|---|
| New number list | Share of 404, 484 and 410 responses; run an HLR or validity check on the list |
| Calling hours | ASR by hour of day in the destination's local time |
| Caller ID flagged as spam | Rotate test calls with a fresh number and compare; see how to check CLI delivery |
| Caller ID not delivered | People answer unknown numbers less; test CLI on the route |
| Traffic mix changed | Share of mobile versus fixed, or a new campaign type |
| Your own routing | Failover timers too short, wrong prefix, CPS limits hit on your switch |
Causes on the route
- Congestion: many 503 or 486 responses at peak hours, or the carrier's channel or CPS limits reached. See channels and CPS.
- A change further down the chain: your carrier switched its own supplier. The giveaway is a change in PDD, CLI delivery or audio at the same time as ASR.
- Blocking at the destination: operators may block traffic arriving from grey routes or with suspicious caller IDs, for example during a crackdown on SIM box bypass. Failures concentrate on one operator.
- Traffic type mismatch: a conversational route that starts rejecting dialer traffic, often with 403 or 503. See dialer traffic and VoIP routes.
- Portability errors: calls to ported numbers misrouted, see number portability and routing.
- Long post-dial delay: callers or dialers hang up before the phone rings, see post-dial delay causes.
A checklist to work through
- Confirm the drop is real: enough attempts, same hours compared, no reporting gap.
- Split ASR by route, destination breakout and customer to find where it is concentrated.
- Compare NER and the top SIP codes before and after the drop.
- Check PDD and ACD for the same period.
- Place test calls and capture SIP traces of failures.
- Send the carrier specific examples (time, called number, code) and ask what changed.
- Move part of the traffic to an alternative route while you wait.
Steps 2 to 4 are much faster if your CDRs are already aggregated per route, as described in monitoring route quality with CDRs.
When a high ASR is suspicious
The opposite can also be a problem. If a route suddenly answers far more calls than any other, while ACD falls, check for false answer supervision before you move more traffic to it.
If a route does not recover, VoIPMatcher can source an alternative VoIP termination route for the same destination and traffic type. Request a VoIP route.
Frequently asked questions
Which SIP codes point to a route problem?
Codes such as 503 Service Unavailable, 502 Bad Gateway and 408 Request Timeout usually point to the network or the carrier, especially when they rise suddenly. Codes such as 404 Not Found, 484 Address Incomplete and 410 Gone point to the numbers you dial. 486 Busy Here and 480 after ringing are normally user results.
Can a spam label lower ASR?
Yes. When call analytics or the handset labels your number as spam or likely scam, many people stop answering, so ASR falls while NER stays stable. Use numbers you own, register them where possible, follow consent rules and, for US calls, ask about attestation. Rotating numbers to hide from labels is not a fix.
How big an ASR drop is significant?
It depends on volume. With a few dozen calls, a swing of 10 points can be noise; with thousands of calls, a few points matter. Set a minimum sample before alerting, for example 50 or 100 attempts, and compare the same hours on previous days rather than a fixed target, since normal ASR differs by destination.
Does time of day affect ASR?
Yes, strongly. People answer less early in the morning, late in the evening and on local holidays, so ASR changes by hour in the destination's local time. Before blaming the route, compare ASR by hour with the same hours on earlier days, and check whether your own dialing schedule moved.
Need a route?
Send the destination, route type and target rate. We reply with what we can deliver.